Based on the concept of cloud water resource (CWR) and the cloud microphysical scheme developed by the Chinese Academy of Meteorological Sciences (CAMS), a coupled mesoscale and cloud-resolving model system is developed in the study for CWR numerical quantification (CWR-NQ) in North China for 2017. The results show that (1) the model system is stable and capable for performing 1-yr continuous simulation with a water budget error of less than 0.2%, which indicates a good water balance. (2) Compared with the observational data, it is confirmed that the simulating capability of the CWR-NQ approach is decent for the spatial distribution of yearly cumulative precipitation, daily precipitation intensity, yearly average spatial distribution of water vapor. (3) Compared with the CWR diagnostic quantification (CWR-DQ), the results from the CWR-NQ differ mainly in cloud condensation and cloud evaporation. However, the deviation of the net condensation (condensation minus evaporation) between the two methods is less than 1%. For other composition variables, such as water vapor advection, surface evaporation, precipitation, cloud condensation, and total atmospheric water substances, the relative differences between the CWR-NQ and the CWR-DQ are less than 5%. (4) The spatiotemporal features of the CWR in North China are also studied. The positive correlation between water vapor convergence and precipitation on monthly and seasonal scales, and the lag of precipitation relative to water vapor convergence on hourly and daily scales are analyzed in detail, indicating the significance of the state term on hourly and daily scales. The effects of different spatial scales on the state term, ad- vection term, source–sink term, and total amount are analyzed. It is shown that the advective term varies greatly at different spatiotemporal scales, which leads to differences at different spatiotemporal scales in CWR and related characteristic quantities.

Aerosol indirect effects, including the Twomey and Albrecht effects, remain a major uncertainty in current weather and climate studies. Such effects are found to vary with cloud types, and the vulnerability is often evaluated with a metric called susceptibility -- the derivatives of the measured variable to aerosol concentration. In this research, the Weather Research and Forecasting model v4.3.1 combined with the NTU microphysical scheme is used to investigate the aerosol susceptibility of marine boundary-layer clouds over the NW Pacific Ocean during cold-air outbreak events. These clouds are often found in mixed-phase upstream and liquid-phase downstream of the cold air trajectory. We found distinctive features of aerosol susceptibility in both cloud types. In the liquid-phase stratocumulus clouds, most cloud properties (e.g., liquid water path, optical depth, and cloud albedo) have positive susceptibilities to the aerosol effects, consistent with earlier studies. However, the cloud fraction showed a negative susceptibility when drizzle is active. In the mixed-phase stratocumulus clouds, on the other hand, the cloud and ice water paths, as well as cloud fraction, are found to increase in the low-end and high-end aerosol concentrations but decrease when the aerosol concentration is between 10² and 10⁴ cm^-3. Such a nonlinearity is likely associated with the transition of precipitation formation from drizzle-dominated in low aerosol concentrations to ice-dominated at high aerosol concentrations.

The role of cloud subgrid-scale structure in modulating satellite views of clouds was investigated. This was realized by implementing a stochastic cloud generator into the CFMIP (Cloud Feedback Model Intercomparison Project) Observation Simulator Package together with surrogate clouds produced by a cloud- resolving model (CRM). The subgrid-scale structural parameters are decorrelation length for overlapping cloud fraction (Lcf), decorrelation length for overlapping cloud condensate (Lcw), and the shape parameter v for measuring cloud inhomogeneity. With the use of median values of Lcf, Lcw, and v derived from CRM, the simulated satellite views bear close resemblance to those using CRM-inherent clouds. Varying these parameters in the range of lower and upper quartiles leads to differences that are about one-fifth of those caused by changing cloud microphysics in CRM. While Lcf influences clouds throughout the whole troposphere, Lcw and
v result in changes mostly within the upper layers. Increasing (decreasing) cloud inhomogeneity or overlapping degree leads to decreased (increased) occurrence of clouds, except for high-topped clouds viewed by Multiangle Imaging Spectroradiometer. Care must then be exercised when interpreting model biases in comparison with different instruments. Sensitivity tests show changing condensate distribution from gamma to lognormal makes little impact on final results. Although the differences induced by any of the parameters alone are much limited, they are getting comparable to those seen between models and observations when all parameters are synergistically altered. This brings encouraging results to the modeling community that simulator-diagnosed clouds can be potentially improved by tuning cloud subgrid-scale parameters.

The top-hat approximation, which is widely used in the mass flux type convection, is verified at different horizontal scales, especially those at the limit of vanishing cloud fraction, with the aid of large eddy model simulations (LES). Three shallow convection cases in the Global Energy and Water Cycle Experiment (GEWEX) Cloud System Study (GCSS) programs are conducted, including BOMEX, RICO and ATEX. Results show that convective cloud fraction increases with the increase of horizontal resolution, consequently resulting in errors of cloud component in the decomposition of scalar flux increase. These errors are however largely compensated by the decrease of errors in the environment, leading to the error of top-hat approximation almost unchanged. For either cloud or environment component, there exists an inversed relationship between convective fraction and the covariance between vertical velocity and conserved tracer. This brings encouraging results to the modeling community that the mass-flux method in parameterizing convection still works at high resolutions. However, the closure remains a big problem at the limit of vanishing cloud fraction, which is not covered in this study.

Convective parameterization in numerical models remains an important source of model uncertainty, as evidenced in the deficiency in simulations of the diurnal cycle of precipitation (DCP). In this study, the behaviors of the phase and amplitude of DCP in China in three commonly used reanalysis products, including ERA-5, JRA-55 and MERRA-2, are compared against GPM (Global Precipitation Measurement) and CMORPH (Climate Prediction Center Morphing Technique). Results show while JRA-55 and ERA-5 produce DCP that are closely consistent with observation, MERRA-2 shows 1- to 3-hour shifts in phase and significantly underestimates the amplitude. Further analysis is performed by comparing 3-D cloud fraction, apparent heating (Q1) and apparent moisture sink (Q2) against cloud-resolving model (CRM) simulations. Only ERA-5 captures the evolution of clouds in association with convective precipitation as in CRM simulations, while the other two fail to reproduce such characteristics. This implies the need to link DCP with subgrid-scale processes besides convection and to treat physical parameterizations as an integrated system.

The Mei-Yu front heavy rainfall event occurred in northern Taiwan on 2 June 2017. The largest daily accumulated rainfall was 645.5 mm at the north tip of Taiwan. The frontal system becomes quasi-stationary in northern Taiwan, and it lasts for 10 hours. A strong barrier jet over the northwest coast of Taiwan is present when the Mei-Yu front approaches northern Taiwan. The strong low-level jet brings abundant moisture to Taiwan and causes strong convergence along the frontal zone. The strong convection produces more than 600 mm of rainfall when the system becomes stationary. In order to examine how the barrier jet over northwestern Taiwan affects the movement of the Mei-Yu front, this study uses the WRF model to reproduce this heavy rain event. Meanwhile, three sensitivity tests are conducted in the numerical experiment. In the removing Taiwan terrain test, the front moves southward quickly. The accumulated rainfall was only 200 mm and the barrier jet is much weaker than the CTRL, showing that the orographic can obstruct the frontal movement. In the replacing southern Taiwan mountain test, northern Taiwan terrain provides part of the blocking effect, but the front still moves southward quickly without a strong barrier jet. In the enhanced barrier jet test with the same topography as CTRL, the barrier jet is slightly stronger than CTRL. The position of the stationary Mei-Yu front is further north than the CTRL. The strong southerly wind also enhanced the convergence, resulting in more rainfall, however, the rainfall did not fall over the land. The results show that the orographic effect contributed to part of the blocking effect for the frontal movement in northern Taiwan, and caused the barrier jet to form in northwestern Taiwan. The slow movement of the Mei-Yu front is significantly influenced by the barrier jet.

The fine spatial characteristics of hail events over Beijing metropolitan region (BMR) is performed using direct observation from quality-controlled disaster information dataset during the year of 2011-2021. Hail is concentrated in urban and northeast mountain region which is highly related with both synoptic circulations and the underlying surface. Four synoptic circulation patterns: the northwest flow in front of ridge (NWP), straight westerly flow (SWP), cold vortex (CV), and pre-trough (PT) are investigated and the hail generation is found differs under these patterns. SWP favors local hail events with the largest convective available potential energy and PT favors systematic hail events with the highest vertical wind shear. With weak low level background flows under NWP and CV, hail events concentrate over the BMR’s plains with obvious surface warm center and wind convergence near urban region which favor storm initiation and enhancement. When the low-level background flow is larger in SWP and PT, the more dominated mountain-plain circulations lead to the hail events center changing to the BMR’s northeast mountains rather than the urban region. These results suggest the potential influences of urban environment and mountain-plain circulations on the distribution of hail events under different synoptic circulation patterns.

We explored the use of modern deep learning models of Convolutional Neural Networks (CNN) and Graph Neural Networks (GNN) for objective climate predictions of summertime air temperatures in South Korea. In order to design deep learning models more appropriate for climate predictions, the CUTMIX data augmentation technique was modified and applied to climate observations and APCC Multi-Model Ensemble (MME) data. The 3-dimensional CNN model performed much better with data augmentation for validation (accuracy > 0.6 for all summer months and folds) as well as test data for June (LT1) and July (LT2). Class activation maps were examined and the contributions of the data augmentation could be observed in some cases of the test data, e.g., in July 2018 and July 2021, when the northern Pacific are and northern polar region are activated and improved the predictions respectively. Graph Convolution Network models were developed for node classification (years as nodes) and graph classification (geographical grids as nodes and years as multiple graphs). Predictions for July (LT2) of the node classification model were improved with the month-agnostic approach, where all months of data are used for training. Predictions for June (LT1) of the graph classification model were improved with the month-agnostic approach (Heidke Skill Score > 0.35 for all folds) and predictions for July (LT2) were also improved with the aforementioned data augmentation. ※ This research was supported by APEC Climate Center.

Ocean surface albedo (OSA) is essential to the ocean and climate energy balance. It is usually considered as the constant or just a simple function of the solar zenith angle (SZA) in the climate model. However recent research suggests that the OSA can be significantly affected by low-level wind, and play an important role in climate simulation, especially in high-resolution simulation. In this work, we incorporated an improved OSA parameterization scheme into the First Institute of Oceanography-Earth System Model (FIO-ESM) V2.0. The revised parameterization scheme takes into account the sea surface roughness and whitecaps induced by surface wind and simplified water volume scattering. Numerical experiments indicate that the improved scheme leads to an increase in OSA of roughly 40% at the global scale. It is remarkable that the effects of foams or whitecaps are noticeable in areas with strong winds, such as the Southern Hemisphere westerly zone. The model bias in sea surface net shortwave radiation has been reduced by an average of 3 W/m², and over the subtropical and Antarctic oceans by up to 8 W/m² and 14 W/m². The enhanced OSA parameterization also reduces the global annual mean error of sea surface temperature by up to 0.87°C.

The Global Observing Satellite for Greenhouse gases and Water cycle(GOSAT-GW) satellite is planned to be launched in 2024 as the successor to the GOSAT-1 and GOSAT-2 greenhouse gas observation missions. The GOSAT-GW satellite challenges to simultaneously observe greenhouse gases and nitrogen dioxides (NO2), major air pollutants. The grating spectrometer is equipped, and more than three million points are observed per day.
The air mass factor (AMF) is used to convert from the NO2 slant column density to the vertical column density, and is one of the largest error sources in retrieving NO2 vertical column from the observation spectrum. High speed calculation of AMF is required because of large number of observation points of GOSAT-GW. The AMF values for all observation cases are pre-calculated by radiative transfer model and saved in the look-up table (LUT).
This study discusses the optimization of LUT, i.e, how to select the nodes of input variables in LUT. We used SCIATRAN version 4.6.1 for radiative transfer calculation. The node of each input variable, such as solar zenith angle, viewing zenith angle, albedo, and terrain height, are selected by AMF gradients for the variables. Our algorithm showed 45% lower error than conventionally-made LUT. In this presentation, we show the results of all LUTs used in the NO2 retrieval data processing of the GOSAT-GW.

Ambient NH₃ plays an important role in forming particulate matter (PMs), and therefore, it is crucial to comprehend NH₃'s properties in order to reduce PMs. However, it is not easy to achieve this goal due to the lack of monitoring data on ambient NH₃ concentrations in typical air quality stations. Nor are we aware of any study that has looked into NH₃ predictions as of yet. This study thus offers the first inquiry into applying machine learning and an auto hyperparameter optimization approach to estimate NH3 concentrations. To obtain more crucial data about NH₃ concentration, we additionally created time lag and parcel tracking routines. To analyze feature importance, we applied the SHAP (SHapley Additive exPlanations) function. From 2016 to 2018, Taichung's hourly average NH₃ values were about 16.9 ppb. Such NH₃ concentrations were predicted using an optimized extra trees model that has the potential to account for up to 96% of the total variance. Agriculture activity was the most significant factor (main source) to affect NH3 concentrations in Taichung among all the characteristics.

A significant relationship between the Indian Ocean Dipole (IOD) and the following year’ s El Niño-Southern Oscillation (ENSO) has been reported in recent decades. Nevertheless, uncertainty exists regarding the associated mechanisms. Based on the statistical analysis and numerical experiments, our study proposes that the spring southeastern Indian Ocean warming (SEIOW) plays a bridging role in the teleconnections of the IOD and subsequent ENSO. A positive IOD could induce a positive tendency of sea surface temperature (SST) in the southeastern Indian Ocean from autumn to winter, primarily through the cloud-radiation-SST feedback, forming an anomalous SEIOW in the subsequent spring. As a Gill-model response to this SEIOW, an anomalous anticyclone appears over the western North Pacific, accompanied by easterly wind anomalies in the western equatorial Pacific which generate eastward propagating upwelling Kelvin waves. As a result, anomalous cooling appears over central-eastern tropical Pacific in the following seasons, manifesting as a La Niña mode. The process is vice versa for the teleconnections of the negative IOD and following year’s El Niño. Additionally, role of the ocean channel (i.e., the Indonesian Throughflow) in connecting the spring southeastern Indian Ocean and the subsequent winter eastern Pacific SSTs is also discussed.

Mixed Rossby-Gravity (MRG) waves are westward propagating synoptic scale equatorial disturbances, which play a crucial role in the formation of tropical cyclones and tropical depressions. They constitute a significant part of various modes of tropical variability, like Madden-Julian Oscillation (MJO) and Quasi-Biennial Oscillation. This study investigates the trends and Inter Annual Variability (IAV) in the occurrence of upper tropospheric MRG events using ERA-I reanalysis data for the period 1979-2018. The MRG events are identified by projecting the 200hpa meridional winds onto the theoretical spatial structure of MRG waves. A steady increasing trend is observed in the MRG events, which is contributed by the MRG events associated with intrusion of extratropical disturbances. Possible factors governing the observed trend and IAV in MRG events are El-Nino Southern Oscillation (ENSO), MJO and extratropical forcing. The MRG events over the central and eastern Pacific contribute maximum to IAV. ENSO explains about 25% of IAV and exhibits a positive correlation with non-intrusion MRG events and a negative correlation with intrusion MRG events. These observations have been investigated by exploring the properties of the westerly duct at 200hPa and Outgoing Longwave Radiation during El-Nino and La-Nina years over the central-eastern Pacific. The convectively active state of MJO over the western Pacific explains 20% of IAV. The antisymmetric heating with respect to the equator, associated with MJO, enhances non-intrusion MRG events by forbidding the intrusion of extratropical disturbances through subtropical easterlies. The increasing trend in the intrusion of extratropical disturbances explains the observed trend in MRG events.

The global atmosphere NWP system – named the Korean Integrated Model (KIM) – developed by the Korea Institute of Atmospheric Prediction Systems (KIAPS) was made operational at the Korea Meteorological Administration (KMA) in April 2020. The global data assimilation (DA) system is based on a hybrid-4DEnVar system consisting of KVAR (KIM VARiational data assimilation) and LETKF (Local Ensemble Transform Kalman Filter). Although it provides a good forecast skill within the performance range of the world’s leading NWP centers, the prediction has some systematic bias over East Asia. Forecast skill of the forecast system including DA system is sensitive to initial data and assimilated observation data. Thus, improving the model itself is important, but the performance of observation data used for DA is also important in order to improve forecast skill. An Observing System Simulation Experiment (OSSE) has been conducted to understand the model prediction sensitivity according to DA. If we consider a global gridded virtual Sonde observation network and assume that the ERA5 analysis is true of the OSSE system, simulated observations with globally 1-degree intervals can be generated from ERA5. In this study, the difference between forecasts fields assimilated with the existing observation data and the virtual Sonde data is examined for cases chosen when KIM’s East Asia forecast skills are significantly low. Through these experiments and analysis, we would like to examine the following two. First, this system will be used to investigate the effect of the initial condition and the virtual Sonde data as the observation data on the 5-day prediction accuracy of the East Asia. Second, we plan to conduct sensitivity experiments for diagnosing regions where affect the improvement of East Asia’s forecast performance to improve 5-day forecast skill.

Observing system simulation experiment (OSSE) provide a rigorous, cost-effective approach to evaluating the potential impact of new observing systems and alternate deployments of existing systems and to optimizing observing strategies (Hoffman and Atlas, 2016). One of most advantageous thing in OSSE is that researcher can verify the result of Numerical Weather Prediction (NWP) with respect to the truth (called as nature run (NR)). Furthermore, observation can be simulated for any proposed (for example, making new type of sensor, appling realistic forward operator and method, specifying the locations, and error characteristics of the observation). To use such advantage of OSSE, we developed the prototype of global OSSE system used Korea Integrated Model (KIM).
In this study, we will introduce the theoretical background of OSSE (OSSE is generally consist of the NR, simulated observation, DA and forecast, and calibration and verification), KIM forecast system, and what i done to develop the OSSE system used KIM, briefly. Finally, to verify the OSSE system used KIM, we performed the denial experiments for aircraft, not only in OSSE but also performed in OSE. In comparison result between OSE and OSSE aircraft denial experiments, OSE/OSSE showed similar performance in most atmospheric variables of analysis field. So, we evaluated the OSSE system used KIM is well made enough to simulate the real case.

The planetary boundary layer (PBL) is the lowest part of the atmosphere and interacts directly with the Earth's surface, playing an important role in weather, climate, and air quality. It also plays an important role in the exchange of aerosols, heat, humidity, and other atmospheric gases. Therefore, it is vital to understand the spatial and temporal variations of the PBL height (PBLH). Radiosonde launching has been used to determine the vertical profile of the PBL. The Korea Meteorological Administration(KMA) has been launching radiosondes twice a day (00 and 12 UTC or 9 and 21 LST) at ten different stations in Korea. In this study, the PBLH was calculated using the bulk-Richardson method with the KMA radiosonde data. The bulk Richardson method calculated the PBLH using the ratio of thermally generated turbulence generated by the vertical shear. The used variables were virtual potential temperature, wind speed, gravity, altitude, and humidity. When the bulk Richardson method reached the threshold, two types of thresholds, 0.25 and 0.5, were commonly used. In this study, 0.25 was used as the threshold. The altitude when the threshold of the bulk Richardson method calculated from the ground exceeded 0.25 was determined as the PBLH. The calculated PBLH was also compared with those determined based on the ceilometer measurements.

Currently, the operational Korean Integrated Model (KIM) Numerical Weather Prediction system at the Korea Meteorological Administration (KMA) uses 3DIAU to add deterministic analysis increments to the model. The weights are based on a Dolph filter, but with the first and last weights modified to remove discontinuities, producing a “modified Dolph filter”. Hybrid-4DEnVar scheme used for deterministic analyses produces a total of 7 analysis increments – one every hour within data assimilation window. Thus, it is possible to replace the 3DIAU scheme with a 4DIAU scheme that makes use of the hourly analysis increments. We have experimented with a 4DIAU scheme that adds each analysis increment over a 2-hour period, again using modified Dolph weights. For increments that are, the response is determined by the weighting function used for the individual increments, giving relatively light time-filtering. For increments that are not, such as the stationary increments related to the static covariance, the response is determined by the sum of the weights across all increments, giving relatively heavy time-filtering similar to that obtained with the 3DIAU scheme. Thus, for balanced increments that are consistent in time, such as increments associated with rapidly-moving features like tropical cyclones, 4DIAU should be better than 3DIAU at retaining the increments. We compared analysis and forecast performance of the new 4DIAU scheme with the original 3DIAU scheme. In analyses, there was relatively little impact on T and Q, but U and V were improved. In the Northern Hemisphere, overall forecast performance was improved, and the performance improvement of geopotential height was significant. The performance improvement in the Northern Hemisphere is particularly noticeable in Asia. In a tropical cyclone case experiment, the 4DIAU scheme gave a slightly deeper central pressure than the 3DIAU scheme, and expressed the change in tropical cyclone position over time better than 3DIAU.

NIMS works for meteorological science research and policy support for people’s safety and happiness and has been doing lots of research under the research project, called the “Research and Development for KMA Weather, Climate, and Earth System Services”.
NIMS has developed technologies needed to improve forecasting for KMA such as a medium-term forecast guideline and marine weather forecasting system. Initialization and ensembles of climate prediction system were improved for more accurate and reliable climate predictions. Due to climate change, damages caused by heavy rains in summer have increased and rainfall forecasting has become difficult. So, NIMS has developed diagnostic factors for heavy rainfall through synoptic analysis of severe weather.
There was challenged study using advanced observation equipment meteorological observation vehicles, meteorological aircraft, and meteorological drones. For understanding and analyzing of a meteorological phenomenon, NIMS observed heavy rainfall, black ice, and so on. These data are used to understand the mechanisms of severe weather. Black ice prediction based on these data has been developed to decrease winter traffic accidents. The technologies of weather modification are constantly evolving using advanced equipment. Also, Korea Cloud Physics Experimental Chamber (K-CPEC) was conducted in 2021. We expect to develop better technology and understand cloud physics.
To support national climate change response policy, NIMS produced new future projections based on New Green House Gases (GHG) and published climate change reports. Also, NIMS continuously monitors GHG. For expanding renewable energy, NIMS provides high-resolution meteorological resource maps.
In addition, NIMS has been producing convergence weather information such as airport weather prediction, health weather, and agricultural weather. More detailed results will be shown at the conference.

As a land-atmosphere coupling “hot spot”, the northern China climate transition zone has a sharp spatial gradient of hydrothermal conditions, which plays an essential role in shaping the spatial and temporal pattern of evapotranspiration-precipitation coupling, but whose mechanisms still remain unclear. This study analyzes the spatial and temporal variation of land-atmosphere coupling strength (CS) in the climate transitional zone of northern China and its relationship with soil moisture and air temperature. Results show that CS gradually transitions from strong positive in the northwest to negative in the southeast and northeast corners. The spatial distribution of CS is closely related to climatic hydrothermal conditions, where soil moisture plays a more dominant role: CS increases first, and then decreases with increasing soil moisture, with the threshold of soil moisture at 0.2; CS gradually transitions from positive to negative at soil moisture between 0.25 and 0.35; CS shows an exponential decreasing trend with increasing temperature. In terms of temporal variation, CS is strongest in spring and weakens sequentially in summer, autumn, and winter, and has significant interdecadal fluctuations. The trend of CS shifts gradually from significantly negative in the west to a non-significant positive in the east. Soil moisture variability dominates the intra-annual variability of CS in the study regions, and determines the interannual variation of CS in arid and semi-arid areas. Moreover, the main reason for the positive and negative spatial differences in CS in the study area is the different driving regime of evapotranspiration (ET). ET is energy-limited in the southern part of the study area, leading to a positive correlation between ET and lifting condensation level (LCL), while in most of the northern part, ET is water-limited and is negatively correlated with LCL.

Air temperature and precipitation are two important meteorological factors affecting the earth’s energy exchange and hydrological process. High quality temperature and precipitation forcing datasets are of great significance to agro-meteorology and disaster monitoring. In this study, the accuracy of air temperature and precipitation of the fifth generation of atmospheric reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ERA5) and High-Resolution China Meteorological Administration Land Data Assimilation System (HRCLDAS) datasets are compared and evaluated from multiple spatial–temporal perspectives based on the ground meteorological station observations over major land areas of China in 2018. Concurrently, the applicability to the monitoring of high temperatures and rainstorms is also distinguished. The results show that (1) although both forcing datasets can capture the broad features of spatial distribution and seasonal variation in air temperature and precipitation, HRCLDAS shows more detailed features, especially in areas with complex underlying surfaces; (2) compared with the ground observations, it can be found that the air temperature and precipitation of HRCLDAS perform better than ERA5. The root-mean-square error (RMSE) of mean air temperature are 1.3 ◦C for HRCLDAS and 2.3 ◦C for ERA5, and the RMSE of precipitation are 2.4 mm for HRCLDAS and 5.4 mm for ERA5; (3) in the monitoring of important weather processes, the two forcing datasets can well reproduce the high temperature, rainstorm and heavy rainstorm events from June to August in 2018. HRCLDAS is more accurate in the area and magnitude of high temperature and rainstorm due to its high spatial and temporal resolution. The evaluation results can help researchers to understand the superiority and drawbacks of these two forcing datasets and select datasets reasonably in the study of climate change, agro-meteorological modeling, extreme weather research, hydrological processes and sustainable development.

Due to concentrated buildings, low vegetation, and road paving materials, cities are likely vulnerable to weather environments such as heat waves and floods. In addition, in complex urban structures, it is difficult to predict wind changes caused by high-rise buildings, making it challenging to analyze the thermal environment and pedestrians' thermal comfort in the city.
This study attempts to simulate and analyze wind changes in complex urban environments using a PALM(Parallelized Large-Eddy Simulation Model). Haeundae Marine City in Busan, Korea, located on the coast, was selected as the study target area. This area is where concentrated high-rise residential and commercial buildings are and where light reflection by the outer walls of the building glass and damage by strong winds occur.
Simulate the actual urban structure of Marine City with a large-eddy simulation (LES)-based PALM model to identify changes in surrounding winds due to the height difference of the high-rise buildings inside the city and compare environmental changes such as pedestrian thermal comfort.

In this study, the predictability of the Western Pacific (WP) pattern is evaluated using five seasonal prediction models the Asia-Pacific Economic Cooperation (APEC) Climate Center (APCC) multi-model ensemble (MME) for the winters from 1982/1983 to 2021/2022. The temporal correlation coefficient (TCC) between the observed and MME-predicted WP indices was 0.61 (0.37–0.54 for individual models) for the entire series. However, when only three Super El Niño (SEN) years (Niño3.4 ≥ 2.0) out of the 40-year series were excluded, the TCC dropped down to 0.54 (0.27–0.42). During the SEN years, the WP was strongly affected by the SEN-excited anomalies via the PNA. In observations from non-SEN years, the WP pattern was strongly related to the dipole pattern in Northwestern Pacific SST (TCC = 0.8), for the description of which we suggested a Northwestern Pacific (NWP) index, and it was significantly weakly related to the ENSO and IOD, whereas in the model simulations, the main role was played by the ENSO (TCC = 0.6). The NWP index was well predictable in MME (TCC = 0.73) and individual models (0.56–0.71). We showed that the prediction of the WP index polarity is reliable when both predicted WP and NWP anomalies are significant and indicate the same WP sign that has implications for the seasonal forecasting. Acknowledgment: This work was carried out with the support of the Research Fund of Research Institute for Basic Sciences, Pusan National University, Korea.

In this study we have conducted a survey of Mixed Rossby-Gravity (MRG) wave events in the upper troposphere and quantified their association with the intrusions of extratropical disturbances for the period 1979-2019. MRG events are identified by projecting the equatorial meridional winds at 200 hPa onto the meridional structure of theoretical MRG waves. 2390 MRG events are identified and majority (61%) of them occurred during May-October months, and 65% of the total MRG events occurred over the central-east Pacific and Atlantic Ocean domains. Not only the frequency of occurrence but also the amplitude, wavenumber and trapping scale of the MRG events are found to exhibit a clear seasonality. MRG events associated with intrusions of extratropical disturbances are identified as when the potential vorticity on the 350K isentropic surface at 15° latitude exceeded 1 PVU in the vicinity of the MRG events. We find that 37% of the MRG events are intrusion MRG events and a large majority (88%) of such events occurred over the central-east Pacific and Atlantic Ocean domains. It is also noteworthy that nearly 70% of such intrusions occurred in the winter Hemisphere where the westerly wind ducts are well developed. Over the central-east Pacific during Northern Hemispheric (NH) winter, it is observed that the amplitude of intrusion MRG events are larger and have a larger meridional extent compared to non-intrusion MRG events. They also exhibit a similar spatial scale as the extratropical disturbances implying that resonant interactions may be a primary mechanism for the genesis of MRG events. During NH summer, on the other hand, MRG events are primarily triggered by convective processes and the extratropical disturbances may be instrumental in amplifying their amplitude.

Korean Integrated Model (KIM), developed by Korea Institute of Atmospheric Prediction System (KIAPS), has been in operation at Korea Meteorological Administration (KMA) since April 2020. It has been continuously improved its performance through five updates (v3.5a (2020.06), v3.6a (2021.4), and v3.7 (2021.12), v3.8 (2023.2)) including data assimilation and physics.
One of the major systematic errors of the KIM is the cold biases in the lower Arctic atmosphere in summer, which degrades KIM’s prediction performance in the northern hemisphere. The radiative cooling effect by clouds, especially, over-estimated cloud ice, is analyzed as one of the main factors.
In this study, we will examine the possibility of improving the above-mentioned systematic errors through sensitivity test to changes in parameters related to ice crystals of WSM5, cloud microphysics scheme adopted by KIM.

The single-scattering albedo (SSA) of atmospheric aerosols is a key parameter that controls aerosol radiative effects. The variation of SSA is thought to be mainly regulated by aerosol absorption in the Himalayas and South Asia, but observations contradict this idea. In situ field campaigns conducted over two Himalayan sites revealed that SSA was strongly dependent on scattering but weakly correlated with absorption. Observational results combined with the Mie theory further illustrated that SSA was primarily modulated by size distribution rather than absorption. Aerosol Robotic Network (AERONET) data showed similar impacts of size distribution on SSA and that aerosol radiative forcing efficiencies were significantly dependent on SSA. Aerosol size distribution therefore considerably affects radiative forcing by modulating aerosol SSA over the Himalayas. This study highlighted the influence of aerosol size distribution on radiative forcing over the Himalayas, which has important implications for understanding aerosol radiative effects globally.

Polarization characteristics of the atmospheric scattering is important and not to be ignored in radiative transfer simulation. A new polarized radiative transfer method is developed for visible and near-infrared spectra, which is suited for use in remote sensing applications and can calculate the polarized radiation emerging from an atmosphere. The single-layer polarized radiative transfer equation and inhomogeneous multi-layer connection are solved by the discrete ordinates method and adding method, respectively. Monte Carlo model (MYSTIC, as the benchmark) and PolRadtran/RT3 are used to evaluate the new method in both accuracy and computational efficiency under different atmospheric conditions and view angles. Judging from the results, the accuracy of Stokes vector (I-, Q-, U-, V-component) calculated by the new method is a good agreement with the results by PolRadtran/RT3 except where near solar incident zenith angle and anti-incident zenith angle. The relative root mean square errors (RMSE) of Stokes vector for test cases between MYSTIC and the new method or RT3 can also prove the good accuracy of new method. Meanwhile, the new method has a higher computational efficiency compared to RT3, especially for the atmosphere with large scattering optical depth. As differ from RT3, the computing time of the new method cannot increase with increasing optical depth.

In this study, the Extinction coefficient(α) was calculated using lidar data installed at Seoul National University and divided by the mass concentration of fine particles to confirm the Mass Extinction Efficiency(MEE). This study used lidar data from January 2015 to June 2020 for analysis and compared the mass concentration of fine particle at the AirKorea Gwanak-gu station and relative humidity data(RH) at the Metropolitan Meteorological Administration. RH was divided into 7 sections, and we checked the change in the particle's characteristics. As a result, as the RH increased, MEE tended to increase, and it was confirmed that the PM2.5/PM10 ratio also increased. However, the α ratio observed by lidar appears different from the mass concentration. The α Fine/Total ratio rather decreased as the RH increased. Conversely, the α Coarse/Total Ratio showed an increase. We thought to be caused by differences between observation equipment. Airkorea's station method is a Beta-radiation attenuation monitor method that uses a heater to remove some moisture to lower the RH. However, in the case of LiDAR, the effect of RH is not removed by directly observing particles distributed in the atmosphere. Particles corresponding to PM2.5 in a dry atmosphere can increase to a particle size larger than PM2.5 as the RH increases, so they can be classified as coarse particles in LiDAR. However, in mass concentration measurement, the effect of humidity is partially removed, and It is classified as PM2.5. This work was supported by a grant from the National Institute of Environment Research (NIER), funded by the Ministry of Environment (MOE) of the Republic of Korea (NIER-2023-01-02-084).

When atmospheric information is analyzed through LiDAR, backscattering coefficients and extinction coefficients are calculated through LiDAR signals. The LiDAR ratio represents the ratio of the extinction coefficient and the backscattering coefficient. The LiDAR ratio is a variable that varies depending on the size distribution of particles in the air, the refractive index of particles, etc. Still, in most studies, it is fixed as a constant, and the extinction coefficient is calculated from the backscattering coefficient, or the backscattering coefficient is calculated from the extinction coefficient. In this study, LiDAR using a 532 nm wavelength laser and a camera attached to the lidar were utilized. The relationship between the extinction coefficient and the backscattered signal was examined by calculating the backscattered signal proportional to the backscattered coefficient with the pixel value of the laser taken in the picture while obtaining the extinction coefficient by LiDAR. Using these two variables, it was confirmed that the optimized lidar ratio could be calculated according to each atmospheric condition. In addition, in the case of LIDAR, because of the field of view (FOV), a short-range signal of about 0.5 km cannot be calculated. However, since the camera can also obtain pixel values at a short distance, the near-field extinction coefficient can also be restored by using the previously obtained LiDAR ratio and the backscattered signal received from the photograph. Through this study, it was confirmed that if the pixel signal of the camera is used, it is possible to calculate the extinction coefficient more accurately by using the LiDAR ratio optimized for each atmospheric situation, and the overlapping LiDAR short-range signal can be restored.This work was supported by a grant from the National Institute of Environment Research (NIER), funded by the Ministry of Environment (MOE) of the Republic of Korea ( NIER-2023-01-02-084).

We tried to use scanning Light Detection and Range (LIDAR) to measure PM mass concentration as a three-dimension. The scanning Lidar system can detect PM mass concentration in an area with a radius of 5 km using 1064 nm and 532 nm wavelengths. Furthermore, the time resolution of the scanning LIDAR is as short as 10 sec, so it can apply to the timely monitoring of specific aerosol emissions. In the experiment for emissions of aerosols, we faced some problems; the laser was blocked at 2500 m by mountain, and the overlap distance of the 1064 nm laser was longer than that of the 532 nm laser. For the former problem, we neglected the signal from blocking and found the reference extinction coefficient through measured signals over 10 min to increase SNR. In the case of the overlap problem, the range-corrected signal contributed to increasing the slope, which made the reference signal overestimated. Therefore, we need to correct the LIDAR signal by the geometric factor in blind-zone, in which the backscattered signal cannot reach the field of view on telescope. For those reasons, 1) we have theoretically calculated geometric factor from experimental system parameters, such as laser beam divergence, FOV, distance between laser and telescope, and others. 2) We also experimentally considered the relationship signals between 1064 nm and 532 nm lasers to find the signal loss with the assumption of constant Angstrom Exponent. 3) We have corrected geometric factors iteratively from theory to experiment. This work was supported by a grant from the National Institute of Environment Research (NIER), funded by the Ministry of Environment (MOE) of the Republic of Korea ( NIER-2023-01-02-084).

A mixing layer height (MLH) is an important factor that controls air pollution concentrations. Thus, it is necessary to determine MLH accurately and to understand its spatial and temporal variability. One of the common methods for determining MLH is to calculate MLH using the vertical profiles of aerosol backscatter measured by a ceilometer. Originally, a ceilometer has been developed to detect the cloud base height. However, its measurement has also been used to determine the MLH, and several ceilometer measurement networks have been emerged across the world. The Korea Meteorological Administration (KMA) has installed ceilometers to observe cloud amount and cloud base height, and two different types of ceilometers, Vaisala CL31 and Eliasson CBME80B have been operating at 64 stations in Korea. Since 23 December 2022, the National Institute of Environmental Research (NIER) in Korea also has deployed nine Lufft CHM 15k ceilometers to determine the MLH and they are in operation now.
In this study, the MLH was determined based on a gradient method using the aerosol backscatter data measured by KMA and NIER ceilometers, and the spatial and temporal variability of MLH were quantified. The spatial and temporal variability of MLH was further examined as a function of the distance between ceilometers. All correlation analyses were also separated for daytime (06:00 LST to 18:00 LST) and night-time (18:00 LST to 06:00 LST), which had also compared each other. Based on these analyses, the impacts of temporal and spatial variability of MLH on the determined MLH were examined.

Fog is defined as a case where the visibility distance is less than 1 km by atmospheric water droplet, and it is known that fog is difficult to predict because it is caused by various causes depending on the seasons and locations. However, the extinction coefficient of water droplet at a given wavelength of 550 nm is exactly determined by the product of the volume concentration of a water droplet with a given refractive index of 1.33 and the volume extinction efficiency defined at the given size and wavelength. So the rapid change of visibility can be simply explained only by the rapid changes of the volume extinction efficiency by particle size or the rapid changes of water droplet concentration. The volume extinction efficiency sequentially reaches the maximum value following the wavelength for the constantly increasing particle size. In this study, we developed the algorithms on how to predict the occurrence of radiation fog by measuring the growth rate of particles using the relative change in extinction coefficient that occurs sequentially at the three wavelengths of RGB (449nm, 534nm, 597nm). Based on the particle growth theory, when particles grow in various ways, how the extinction coefficient changes at three wavelengths was studied, and how the relative value of the extinction coefficient at each wavelength changes with the growth of the particle size. Based on these theoretical results, we checked the changes of the extinction coefficient obtained from the 3 wavelengths at the radiation fog events and predicted the fog disappearing in the Daecheong area around Daecheong Lake where actual radiation fog often occurs. This work was supported by the “Graduate school of Particulate matter specialization.” of Korea Environment Industry & Technology Institute grant funded by the Ministry of Environment, Republic of Korea.

It is well known that tropical intraseasonal oscillation (ISO) modulates the weather in the tropics. The ISO’s importance in cyclogenesis, extreme precipitation events, and heatwave events is established in the literature. Hence understanding and simulating these oscillations can tremendously improve the extended-range predictions of weather extremes. With changing climate, the frequency and intensity of heatwaves are known to be increasing and becoming one of the primary weather catastrophes. In May 2015 southeastern states of India faced the deadliest heatwave in its history which claims to take 2500 lives. The atmospheric circulation and role of surface fluxes in forming this heatwave were studied in the recent past but the underlying physical mechanism is not known. In this study, we investigated the impact of the tropical intraseasonal oscillation (ISO) in forming the May 2015 heatwave over the southeastern regions of India. The observations show that the occurrence of a heatwave is attributed to north-eastward propagating ISO circulation which results in persistent high-pressure anomaly with anomalous downward motion favoring clear skies, adiabatic heating, and horizontal warm advection. The 2m maximum temperature anomaly shows that 60% to 75% contribution in maximum surface air temperature (SAT) during heatwave period is from ISO-related temperature anomalies. Further numerical analysis using the WRF model confirms that this heatwave event was caused by Propagating ISO. The model output shows a difference of 1.5 to 2 ℃ in maximum SAT between the control and sensitive experiment. Indicating the substantial role of ISO in the development and intensification of the heatwave. This analysis emphasizes that improving the forecasting skills of ISO may facilitate the sub-seasonal forecast of local heatwave events.

The weather observation data collected over a long period of time can be used as important data as the basis for showing the long-term climate characteristics and extreme weather of an area. Meanwhile, in the case of developing countries, it is difficult to collect high-quality weather observation data due to the lack of installation and management system of weather observation equipment. Through the Van-KIRAP (Vanuatu Klaemet Informesen blong Redy, Adapt mo Protekt, in Bislama) project, APCC (APEC Climate Center) wanted to provide observational data that could be useful in the Vanuatu region by adding the gap-filling of observational data to the OSCAR (tailored System of Climate services for AgRiculture) system. Through a comparative analysis conducted by applying various existing gap-filling algorithms to observation data, an algorithm technique suitable for precipitation and temperature, which are the target variables of this study, was selected. The performance of the gap-filling algorithm was compared using the standard statistical indicator CC (Correlation Coefficient) and RMSE (Root Mean Squared Error), and an algorithm that shows compliance performance was selected. It is expected to be used in various applied studies, including long-term climate analysis and extreme climate analysis, through correction observation data through the OSCAR system.

Long-term climate data is one of the main data used for climate analysis, but in developing countries, the development of observation data is insufficient compared to developed countries. We tried to build observation data for long-term climate analysis using satellite remote sensing as the data needed for climate analysis. Precipitation datasets were constructed using GridSat and GPM-IMERGE, and the temperature was constructed based on AQUA, TERRA on installed MODIS sensor and AIRS satellite products. The results of the 10km spatial resolution of the satellite output were developed using the IGISRM technique to construct the entire Vanuatu area with high-resolution grid datasets of 5km. It is expected that climate analysis of high-resolution grid data produced in this project will be used as data for agriculture and various applications.

Large-scale dynamics in the tropics and midlatitudes are governed by two different dominant physical processes. The tropics is governed by the weak temperature gradient system where temperature gradient is constrained to be moderate, whereas the midlatitude area is governed by the quasi-geostrophic system where the Coriolis force and pressure gradient force are nearly balanced. Presumably, for these two different governing equations to be simultaneously valid in large scales, the boundary between these two regions must be connected by phenomena with small spatial scales. Therefore, in this study, we investigate the atmospheric behavior at the tropics-extratropics boundary in the Northern Hemisphere.
The 5800 meter height line at 500 hPa is defined as the tropics-extratropics boundary. This line serves as a proxy for the northern edge of the tropical region. Next, we focus on the strong wind axis of the westerly jet stream, which moves meridionally at mid-latitudes, because the jet stream can supply vortices with small spatial scales. Then, we investigate the positional relationship between the jet stream and the tropical mid-latitude boundary.
By measuring the mean latitudinal distance between the jet stream and the boundary, it is shown that the jet stream flows near tropics-extratropics boundary in most seasons. However, only in seasons when a sea surface temperature (SST) front exists near the boundary, the westerly jet stream is anchored by the SST front and temporarily leaves tropics-extratropics boundary. In boreal spring and autumn, when the westerly jet stream is trapped above the SST front, the existence of mesoscale phenomena such as the Meiyu-Baiu front may be required to connect the tropical and midlatitude solutions in place of the westerly jet stream.

This research aims to assess the effectiveness of hydrological drought indicators in describing characteristics of drought events and to develop a composite hydrological drought index (CHDI) to better characterize the drought conditions in the northern watershed of Thailand, where hydrological droughts frequently occur as a result of climate change and land use. The study utilizes the WRF-CESM climate model and land cover data from MODIS (IGBP) as inputs for the VIC hydrological model to analyze the drought indicators. The VIC model includes surface soil moisture, runoff, precipitation, baseflow, first-layer soil moisture, evaporation, and second-layer soil moisture. The results showed that all parameters have moderate to high correlations with runoff gauge station data, respectively. The CHDI was further developed by combining the indicators using weights derived from the principal component analysis (PCA) technique. The CHDI showed a better correlation (r= 0.45-0.73) with the observed data and was closest to the low-volume water events declared by the Upper Northern Region Irrigation Hydrology Center.

The frequency and intensity of extreme high temperature (EHT) in the Northern Hemisphere exhibit remarkable low-frequency (LF) variations (longer than 10 years) in summer during 1951–2017. Five hotspots featuring large LF variations in EHT were identified, including western North America–Mexico, eastern Siberia, Europe, central Asia, and the Mongolian Plateau. The probability density functions show that the higher EHT occurrences over these hotspots in recent decades is consistent with the shifted average and increased variances in daily mean temperature. The common features of the LF variation in EHT frequency over all domains are the remarkable increasing trends and evident decadal to multidecadal variations. The component of decadal to multidecadal variations is the main contribution to the LF variations of temperature in the last century. Further analysis shows that the coherent variability of decadal to multidecadal temperature variations over western North America–Mexico, eastern Siberia, Europe, and the Mongolian Plateau are the footprints of a dominant natural internal signal: the Atlantic multidecadal oscillation. It contributes to the variations in temperature over these hotspots via barotropic circumglobal teleconnection, which imposes striking anomalous pressure over these regions. This study implies that natural internal variability plays an important role in making hotspots more vulnerable to EHT.

To investigate the recent change of tropical night in South Korea, we analyzed the duration and intensity of tropical nights (daily minimum temperature, Korea Meteorological Administration; KMA) for 40 years (1979-2018) quantitatively. Spatiotemporal analysis showed that the tropical nights in the Seoul metropolitan area were more intense and longer-lasting as compared to other South Korean regions. Specifically, the tropical nights over this region increased more prominently in intensity, frequency, and duration. The tropical night event in the metropolitan area was classified into pure-TN (no heatwave prior to tropical night) and HWTN (tropical night following heatwave). The composite analysis was conducted for two types of tropical night events, pure-TN, and HWTN, based on 40-year ERA5 reanalysis data. Pure-TN mostly occurred when the edge of western North Pacific subtropical high (WNPSH) was present over the Korean Peninsula with the southwesterly wind, leading to a positive temperature advection anomaly in the metropolitan area. Furthermore, a positive low cloud cover anomaly with enhanced downward longwave radiation also prevailed. On the other hand, HWTN mainly occurred when the WNPSH expanded northwestward until its center was located over the Korean Peninsula with a positive downward shortwave radiation anomaly. Moreover, a descending motion anomaly induced adiabatic heating over the metropolitan area was presented. The significant increasing trends in tropical night events (pure-TN: 0.143 day/year, HWTN: 0.077 day/year) were observed at 95% confidence level. To investigate these trends, the regression analysis was performed on the synoptic factors for three sub-analysis periods with 10 days (21-31 July;P1, 1-10 August;P2, 11-20 August;P3). As a result, the favorable atmospheric conditions for HWTN (pure-TN) have been frequently constructed during P2 (P1 and P3).

Understanding the changes in extreme rainfall has raised a lot of attention since it is one of the major exposures in terms of climate risk. In Taiwan, extreme rainfall usually occurs with the unique environmental condition such as tropical cyclone or Mei-yu front. While Extreme indices has been widely used for analyzing extreme rainfall, the data sample is based on each grid instead of the extreme rainfall event itself. Therefore, investigating the extreme rainfall from the event perspective can provide a new insight for the changing extremes. Here we apply the event-tracking method by using Depth-First Search algorithm on high-resolution gridded observation data to track the extreme rainfall events from 1980 to 2019 in Taiwan. Two different thresholds (80mm and 200mm) are then selected for analysis due to their potential threats to river flood and flash flood. Our results show that the extreme rainfall events have increased significantly in both frequency and intensity. Frequency changes in extreme rainfall events indicate a 16.84% increase for 200mm-event and a 10.95% increase for 80mm-event. Mean changes of total rainfall volume also show a larger increase for 75.40% in 200mm-event than 33.03% in 80mm-event. Within all the contributors to the change of total rainfall volume, mean affected area contributes the most even without the effect of increasing rainfall intensity to the threshold. Furthermore, the increasing frequency of extreme rainfall events is larger in southern Taiwan than in northern Taiwan.

Extreme precipitation events (EPEs) associated with intense extratropical cyclonic storms (Western Disturbances; WDs) during the winter season (December through March) have the potential to induce catastrophic damages to life, infrastructure, environment and agricultural sustainability over the western Himalayas (WH). However, a sparse in-situ observational network in this region conjugated with complex topography highlights the uncertainties in available coarse resolution precipitation products and emphasizes the necessity of finer grid scaled regional climate models for accurate simulation of precipitation variability, specifically EPEs with highly localized characteristics as well as underlying synoptic dynamics. The present study is an effort to investigate the performance of a ultra-resolution Weather Research and Forecasting (WRF) model through the evaluation of its sensitivity to various microphysical schemes and their ensemble for the simulation of intense episodes of WDs and associated extreme precipitation events over the WH. The sensitivity of precipitation extremes (identified using percentile approach) to different microphysics (~3 km resolution), in the WRF model configured on two two-way nested domains (9km and 3km), has been validated using multi-source rainfall datasets including gauge-based (IMD) and satellite (IMERG) observations as well as with the latest regional reanalysis dataset, IMDAA. Different rainfall skill scores (CSI, ETS, BSS, HSS, POD and FAR) provided insight into how well WRF mimicked intense WDs and the accompanying rainfall. Furthermore, the corresponding dynamics and physical processes in the model has been elucidated through a composite analysis for the simulated EPEs using the reanalyses, IMDAA and ERA5. Our findings underpin the sensitivity of model-simulated winter precipitation extremes associated with intense WDs over WH to different microphysical parameterizations. Detailed results will be discussed.

The realistic simulation and projection of the Amundsen Sea Low (ASL) are essential for understanding the Antarctic climate and global climate change. Using 14 models that participated in phase 6 of the Coupled Model Intercomparison Project (CMIP6), this study evaluates the climatological characteristics of ASL with comparison to the ERA5 reanalysis and their CMIP5 versions and assesses the future change of ASL under 1.5℃-4℃ global warming. The climatological spatial distribution of ASL is captured reasonably but with underestimated intensity by CMIP6 multi-model ensemble (MME). Among the CMIP6 models, EC-Earth3 has most accurate representation of ASL according to the pattern correlation and biases. The seasonal variation of the ASL depth and location are found to be reasonably reproduced by the CMIP6 models. Compared with CMIP5, CMIP6 MME exhibit evident reduced uncertainties and overall improvement in simulating absolute depth and location of the ASL center, which might be attributed to models’ capability of representing westerlies, while the biases in relative depth become even large in CMIP6 MME. In response to future warming from 1.5℃ to 4℃ above pre-industrial levels, the absolute depth of ASL will very likely deepen with larger amplitude in all seasons, while the relative depth might enhance only under high-level warmer world in austral autumn to winter. The CMIP6 MME also projects that the ASL will shift poleward constantly in austral summer and migrate southwestward during austral autumn with the rising global mean temperature. The enhancement and poleward movement of ASL could also be identified during the Ross Sea ice advance season under 1.5℃-4℃ global warming. The results reveal the potential of CMIP6 models in the ASL study and the impact of ASL on Antarctic climate under different global warming levels.

Using the lagged maximum covariance analysis (MCA), the present study investigates the interannual variability of the storm track in the Southern Hemisphere and the Antarctic sea ice throughout the year. The results show that the two are most tightly coupled in the austral cold seasons. Specifically, storm track anomalies in June and July are associated with a zonal dipole structure of the sea ice concentration (SIC) anomalies in the western Hemisphere, with centers in the Antarctic Peninsula and the Amundsen-Bellingshausen Seas. The storm track can modulate the large-scale atmospheric circulations, which induces anomalous meridional heat transport, downward longwave radiation, and mechanical forcing to further influence the SIC anomalies. The resultant SIC anomalies can last for several months and have the potential to feed back to the storm track. According to the MCA, the influence of the SIC anomalies to the storm track is most evident in August. The SIC dipole along with the SIC anomalies in the Indian Ocean sector have large impact on the storm track activities downstream. The SIC anomalies alters the near-surface temperature gradient and subsequently atmospheric baroclinicity. Further energetic analysis suggests that the enhanced atmospheric baroclinicity facilitates the baroclinic energy conversion from mean available potential energy to eddy available potential energy, and then to eddy kinetic energy, strengthening the storm track activities over the midlatitude Indian Ocean.

Atmospheric Rivers (AR) accounts for more than 90% of poleward moisture transport across midlatitudes of both hemispheres and has a crucial impact on hydrological cycle and surface mass balance over the Antarctic ice sheets. Here we investigate the seasonal and interannual variations of the poleward moisture transport by AR using 43-year MERRA-2 reanalysis data. Integrated water vapor transport (IVT) is used to detect AR with latitude dependent thresholds of IVT to better detect AR-like features in the polar regions. Based on 3-hourly AR statistics, four main AR genesis regions in the southern hemisphere (i.e., southern Africa and southwestern Indian Ocean, southern Pacific Ocean, South America monsoon region, and Australia) are identified. In this talk, the variations of AR frequency and intensity at each genesis regions and associated patterns of atmospheric circulations and its impact on the poleward moisture transports will be discussed on intraseasonal, seasonal, and interannual timescales.

In early February 2020, two consecutive extreme warming events occurred within three days of each other at the same location in the Antarctic Peninsula (AP). The second event on February 9, 2020, saw the temperature reach 15.5°C at Marambio station on Seymour Island, northeast of the AP, the second-highest recorded temperature. To understand the cause of the extreme warming event, observational data from Marambio station and the European Center for Medium-Range Weather Forecasts Reanalysis v5 (ERA5) data were analyzed to examine extreme warming events that had occurred on Seymour Island during February over the past 40 years. The analysis revealed that the extreme warming event on February 9, 2020 was caused by foehn winds and large-scale horizontal advection, which are associated with a strong blocking high in the upper and lower atmosphere. Unlike typical extreme warming events in February over the past 40 years, the event in 2020 occurred not only in the AP but also throughout the entire West Antarctica.

The El Niño–Southern Oscillation (ENSO) is well-known climate variability in the tropics that has the potential to affect the Antarctic sea ice. The sea ice concentration pattern induced by ENSO is primarily linked with an anticyclonic circulation anomaly in the Amundsen-Bellingshausen Sea (ABS) that modulates atmospheric poleward temperature advection and longwave radiation forcing via poleward moisture advection. In the current climate, the remote influences between the tropics and the Antarctic are strongest during Austral spring (September-October-November, SON). However, this seasonality can change in future conditions. Comparing the remote influences from the tropics to the Antarctic between the present (1982-2014) and the near future (2015-2050) using the CMIP6 historical and SSP585 experiments, the positive geopotential heights at 500 hPa in the ABS regressed onto the ENSO index becomes stronger in Austral winter (June-July-August, JJA). This results in a more extended melting period of sea ice by the remote forcing from the tropics. The stronger response in the upper-level geopotential height anomalies in JJA is suggested by the increased forcing in the tropics and strengthened jet stream. The global model experiments with a dynamical core only support well these changes in the dynamical mechanisms related to the Antarctic sea ice condition in the future.

The current paper investigates the weekly forecasts skills on precipitation in boreal summer over Maritime Continent (MC), which is an area featured by complex topography, warmest oceans, and characterized by great vulnerabilities to high-impact precipitation events, for the models of European Centre for Medium-Range Weather Forecasts (ECMWF) and China Meteorological Administration (CMA) derived from the S2S Project. Results indicate that the ECMWF model shows generally superior forecast performances than CMA, which is characterized by lower errors and higher correlations compared with the observations. Meanwhile, ECMWF tends to produce wet biases with increasing lead times, while the mean errors of CMA are revealed to be approximately constant throughout lead times of 2–4 weeks over most areas. Besides, the temporal correlations between model outputs and observations obviously decrease with growing lead times, with a high-low distribution presented from north to south. In order to detect the sources of predictability, the roles of large-scale drivers like ENSO and BSISO in modulating subseasonal precipitation forecast skills are also assessed in the models. Both ECMWF and CMA can reasonably capture the ENSO related precipitation anomalies for all lead times, while their capabilities of capturing BSISO related precipitation anomalies decrease with growing lead times, which is more obvious in CMA. The enhanced subseasonal precipitation forecast skills mainly respond to the BSISO associated precipitation variability. For most MC areas such as southern Indochina, western Indonesia, Philippines and the eastern ocean, the forecast skills of both ECMWF and CMA can be improved to a great extent by enhancing the capture of BSISO related precipitation anomalies, with the temporal correlations for both ECMWF and CMA increased by about 0.15 for lead times of 3–4 weeks. It provides an opportunity window for the models to improve precipitation forecasts on the subseasonal timescale.

The Madden-Julian Oscillation (MJO) provides an important source of global subseasonal-to-seasonal (S2S) predictability, while its prediction remains great challenges. Based on an atmosphere-ocean coupled model and the widely-used nudging method, suitable initialization and ensemble schemes are explored toward an improved MJO prediction. It is found that strong but not excessive relaxation strength for the divergence and vorticity, and a mild relaxation for the air temperature are appropriate to generate good atmospheric initial conditions. Additionally, the ensemble strategy with perturbed atmospheric nudging coefficients conduces to adequate ensemble spread and hence improves the prediction skill.Here, an 18-member ensemble subseasonal prediction system called NUIST CFS1.1 is developed. Skill evaluation indicates that the NUIST CFS1.1 can extend the MJO prediction to 24 days lead, which outperforms a majority of current models in the S2S project but is far from the estimated potential predictability (~47 days). The limited skill at longer lead times corresponds to forecast errors exhibiting slower propagation and weaker intensity, which are largely owing to the model’s shortcoming in representing MJO-related physical processes. The model underestimates the diabatic heating of enhanced convection and fails to reproduce the suppressed convection within the MJO structure, collaboratively weakening the Kelvin/Rossby waves. This causes weaker horizontal winds and ultimately reduces the horizontal moisture advection on the two flanks of MJO convection. Furthermore, the underestimated Kelvin wave induces insufficient planetary boundary layer (PBL) convergence and thereby results in poor simulation of PBL premoistening ahead of MJO convection. These biases limit the MJO prediction in the NUIST CFS1.1, prompting further efforts to improve the model physics.

The prediction of monsoonal precipitation during Indian summer monsoon (ISM) remains difficult. Due to the high correlation between the Central Indian Ocean (CIO) mode index and the ISM precipitation variability, the predictability limit of the CIO mode index is investigated by the non-linear local Lyapunov exponent (NLLE) method in observations. Results show that the predictability limit of the CIO mode index can reach 38 days during boreal summer (from June to September), which is close to the upper predictability limit of intraseasonal precipitation (up to 40 days), and higher than the predictability limits of dynamical monsoon indices (under 3 weeks) and boreal summer intraseasonal oscillation (BSISO) indices (around 30 days). Such high predictability limit of the CIO mode index is mainly attributable to the long predictability limits from the intraseasonal sea surface temperature (SST) and intraseasonal zonal wind, which are the components of the CIO mode. However, the prediction skill of the CIO mode in the subseasonal-to-seasonal (S2S) air–sea coupled models is still an issue. The ECMWF and UKMO models display significantly higher skills for up to about 2 and 3 weeks, respectively, which are longer than other S2S models. The decline of the CIO mode prediction skill is due to the reduced signal of subseasonal zonal winds at 850 hPa over the tropical central Indian Ocean (especially along the equator; 5°S–5°N, 70°E–85°E). Therefore, a better simulation of tropical subseasonal zonal winds is required to improve the CIO mode prediction in models, and the improvement will benefit a better MISO simulation and a higher prediction skill during the ISM.

S2S forecast has significant potential to aid decision-makers in mitigating extreme weather risks. Efficient S2S forecasts employ a central component between Weather and Climate predictions (for two weeks to less than a season). S2S models can yield adequate information based on the probability of the occurrence of extreme events within sufficiently larger areas through the prediction of complex, large-scale predictors. However, one of the significant challenges for S2S forecast is to develop a robust relationship for warm season predictability. A warming climate amplifies extreme events, making them a greater societal threat. Therefore, the current study develops an adaptive machine learning approach to predict the weekly frequency of intense warm days and 14-day standardized precipitation index (SPI) during the warm season for the Mahanadi River basin, India. The framework incorporates the multiple S2S global forecast models and ensembles across a continuum of S2S timescales to characterize the deterministic predictability of warm season atmospheric teleconnection patterns. The current skillful S2S forecasts of extreme heat and precipitation in warm season greatly benefit multiple sectors, including water management, public health, and agriculture, in mitigating the impact of extreme events.

As an important parameter to measure the exchange of surface fluxes and climate sensitivity, the surface temperature is used to reflect specific responses to climate changes. The global mean surface temperature (GMST) justifiably became a primary indicator of climate volatility on a global scale to evaluate the intensity of climate change. However, the GMST peaked in 2020 under the background of moderate La Niña and even beat the highest record of GMST in 2016 when a strong El Niño occurred. The phenomenon overturned the conventional idea that the highest GMST occurs with strong El Niño. Our work used several datasets of surface temperature and adopted the ensemble empirical mode decomposition (EEMD) algorithm to distinguish the sources of direct contribution to the highest GMST of 2016 and 2020 at different timescales, especially focusing on the subseasonal-to-interannual variations. The results show that high GMST in 2016 was controlled by secular trend (SCT) and annual variability (ANV). However, the dominator of the sharp GMST rise in 2020 was SCT alone because the ANVs in different seasons canceled each other out in 2020, contributing little to the annual mean GMST. We further revealed two prominent patterns of seasonally varying ANVs, one is induced by atmospheric internal variability and the other is related to ENSO. The two patterns mainly located in Eurasia, North America, the Arctic Ocean, and the tropical eastern Pacific Ocean. Dominance by surface temperatures over the four crucial regions on the subseasonal-to-seasonal (S2S) GMST variations was also observed in nearly 80% of the years during 1982-2021, indicating a potential opportunity to improve the S2S GMST forecast by ameliorating the forecast in the four areas. Whether the contribution of ANV will become weaker and weaker in a warmer future world will be the concern of follow-up work.

The PISSARO project focuses on atmospheric and oceanic forecasting at the subseasonal scale for applications over the southwest Indian Ocean basin. It is a collaborative academic research project, developed and conducted in partnership with stakeholders from Reunion and Seychelles and a panel of scientific experts in tropical forecasting. We have exploited the IFS monthly forecast ensemble for the development of future S2S-derived forecast products to provide probabilities of tropical cyclone trajectory scenarios, rainfall anomalies and to follow the ITCZ migration in the southwest Indian Ocean. The products which have been designed and calibrated with the S2S data are now tested in real time during the 2023 austral summer by the forecasters of the Seychelles Meteorological Authority and at the Regional Specialised Meteorological Centre - Tropical Cyclones of La Réunion. A co-creation approach between scientists, forecasters and humanitarian stakeholders has also enabled the development of products to anticipate the cyclone hazard in the southwest Indian Ocean region to support the tropical cyclone disaster risk management. The PIROI is testing this product under operational conditions for the 2022-2023 cyclone season. At the end of the TC season, all these prototypes will be evaluated and the best products will be selected for an operational implementation. We will provide an overview of these products at the workshop.

The heat transfer through atmospheric circulation is vital in determining local weather patterns. In the boreal summer, latent heat transport is the main contributor to the moist static energy transport from the equator to high latitudes. Using a global coupled atmosphere-ocean model (KIM-NEMO), we analyze the impact of ocean coupling on the subseasonal-to-seasonal (S2S) prediction of latent heat transport. The results showed that the highest latent heat transport in the boreal summer was along the eastern part of the continent. The simulations in July 2017 revealed the significant differences between KIM and KIM-NEMO, particularly over the Western North Pacific Ocean. These changes were triggered by the cold SST anomalies (SSTA) in KIM. The improvement of cold SSTA through ocean coupling influenced the regional wind patterns, leading to modifications in moisture transport. Consequently, latent heat transport is simulated more realistically, when considering air-sea interactions.

Situated in the deep tropics, Singapore experiences considerable intra- and interannual variability of rainfall. Therefore, improving rainfall predictions, including the 1-month rainfall outlook, is an important component of improving climate services for Singapore. Traditionally, seasonal prediction models were used as the basis for predicting rainfall for the next month and next season. These seasonal predictions were also combined with additional information, such as the status and outlook of climate drivers such as ENSO (El Nino Southern Oscillation) and MJO (Madden Julian Oscillation) to produce an expert-based outlook. However, for users who do not require outlooks prior to the end of the month, S2S predictions can be used instead, with S2S predictions issued towards the end of the month showing improved skill compared to the seasonal predictions for the 1-month rainfall outlook. This work compares the 1-month rainfall outlook for Singapore from various sources including seasonal models (single models and multi-model ensemble), the ECWMF extended range model as well as along with expert judgment for the period 2020 – 2022. The benefits of the various types of outlooks will also briefly be discussed as part of improving climate services for the country.

Teleconnection patterns associated with the Madden–Julian oscillation (MJO) and El Niño–Southern Oscillation (ENSO) impact weather and climate phenomena in the Pacific–North American region and beyond, and therefore accurately simulating these teleconnections is of importance for seasonal and subseasonal forecasts. Systematic biases in boreal midwinter ENSO and MJO teleconnections are found in eight subseasonal to seasonal (S2S) forecast models over the Pacific–North America region. All models simulate an anomalous 500-hPa geopotential height response that is too weak. This overly weak response is associated with overly weak subtropical upper-level convergence and a too-weak Rossby wave source in most models, and in several models there is also a biased subtropical Pacific jet, which affects the propagation of Rossby waves. In addition to this overly weak response, all models also simulate ENSO teleconnections that reach too far poleward toward Alaska and northeastern Russia. The net effect is that these models likely underestimate the impacts associated with the MJO and ENSO over western North America, and suffer from a reduction in skill from what could be achieved.

Weather and climate in Southeast Asia are heavily influenced by the Monsoon, with the Monsoon traditionally characterised by the seasonal change in wind directions along with seasonal changes in precipitation. Wind observations or reanalysis data are commonly used for defining the Monsoon period and therefore having wind forecast for the region can provide information on the upcoming onset and cessation of the Monsoon. A spatial approach to the verification of gridded S2S wind forecasts for Southeast Asia is investigated and analysed. The approach uses the fractional skill score (FSS), a neighbourhood spatial verification method commonly used for verifying high resolution precipitation forecast. This spatial approach verification method avoids the problems of conventional local grid-based verification metrics, such as double-penalty when predicting the correct pattern of wind but at the wrong position, and can highlight the spatial displacement of the wind patterns in the forecast. The score for the ECMWF Extended Range wind forecast for Southeast Asia region will be evaluated and compared with the conventional local grid-based verification metrics for the added information using the spatial approach. The possibility of the usage of sub-seasonal wind forecast for the Monsoon onset and cessation forecast in the region will also be discussed.

The Madden-Julian Oscillation is the dominant source of rainfall variability in the tropics on subseasonal timescales. As such, the MJO plays a significant role in forecasting tropical rainfall on 2-3 weeks timescales, and previous studies have shown an improvement in extreme rainfall prediction skill when an MJO event was present. Southeast Asia (SEA) is greatly affected by intraseasonal rainfall variability associated with the MJO, and improvement of subseasonal prediction skill of MJO-related rainfall could benefit both stakeholders and the vast population in the region. In this work, we assess the representation of MJO-related rainfall in subseasonal forecasting models. Specifically, we will explore how well the commonly used MJO monitoring indices can represent rainfall associated with MJO in SEA, both in observations and the models. The skill of the models to simulate MJO-related rainfall in sub-regions in SEA where different MJO phases show the largest subseasonal variability will be examined, as well as the seasonality in prediction skill of MJO-related rainfall in these sub-regions.

Meiyu shows substantial subseasonal variation at periods of 10–30 days and 30–60 days, which often leads to extreme precipitation and disastrous flooding over the Yangtze River Basin. Monitoring and prediction of the subseasonal variation of Meiyu is crucial for disaster prevention and mitigation. Here, we proposed two sets of novel indices for Meiyu real-time monitoring and prediction based on the compound zonal displacements of the South Asia high (SAH) and the western Pacific high (WPH) at 10–30-day and 30–60-day period. For the 10–30-day period of Meiyu, the zonal displacement of the SAH is associated with a mid-latitude Rossby wave train, whereas the WPH is related to the second mode of the boreal summer intraseasonal oscillation. On the 30–60-day timescale, the zonal displacement of the SAH and the WPH are both associated with the first mode of the boreal summer intraseasonal oscillation. The subtle differences in zonal displacement of the SAH and the WPH determine eight type configurations, corresponding to distinct influences on Meiyu. Meiyu subseasonal variation can be well reconstructed by using the relationship between these two indices and rainfall anomalies pattern over China. Given that the ECMWF S2S model is more skillful in forecasting upper- and lower-level circulation than that in directly forecasting precipitation, a hybrid dynamical–statistical model is conducted to predict the Meiyu subseasonal variation with the ECMWF model forecast indices. The hybrid model outperforms the ECMWF model in predicting the Meiyu subseasonal variation at 17–40-day lead times.

In the spring season, Korea suffers from the Asian dust storms (ADSs) sometimes originated from the Gobi Desert and Inner Mongolia. Since the ADSs are known to deteriorate respiratory or eye health as well as visibility, it is important to predict the occurrence and the intensity of ADSs through numerical prediction and satellite observations. In this study, we assimilate aerosol optical depth (AOD) retrieved from Geostationary Environmental Monitoring Spectrometer (GEMS) observing East Asia. The ADSs are affected by various atmospheric (e.g., wind, temperature, planetary boundary layer, etc.) and land surface conditions (e.g., land use, soil temperature, soil moisture, etc.). Therefore, we employ a coupled meteorology-chemistry model --- the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) --- in order to consider the complex interactions among the atmospheric and land surface variables simultaneously. The Maximum Likelihood Ensemble Filter (MLEF) is adopted to assimilate AOD. For an ADS case in March 2021, it is shown that the GEMS AOD assimilation improves the analysis and prediction of AOD and is useful for ADS prediction in the East Asia region.

In air quality prediction, initial conditions are essentially required for a coupled atmosphere-chemistry model which is based on both atmospheric and aerosol/chemistry observations. In general, a higher amount and quality of observations produce more accurate model results; however, forecast errors in a region of interest may grow from an initial error in a specific upstream region, primarily due to the lack of observations therein. Therefore, it is a milestone to find the upstream areas from which a small initial error can grow into large forecast errors in the region of interest. The conditional nonlinear optimal perturbation for initial conditions (CNOP-I) is a suitable tool for adaptive (targeted) observations. To calculate the CNOP-I, several variables are included in defining multiple energy equations (e.g., kinetic energy, dry energy, moist energy, etc.). In particular, the previous studies on improving the Asian dust storms (ADSs) prediction mainly considered the kinetic energy calculation, which included only two variables (u-wind and v-wind). The goal of this study is twofold: finding variables that strongly affect the ADSs and identifying sensitive areas for the adaptive observations. We collected and examined the ADS events that occurred during the last 32 years (1990—2021) over South Korea. Through the principal component analysis, we first recognized top variables—temperature, vertical velocity, divergence, specific humidity, ozone mass mixing ratio, and eastward wind—which affect origination and translocation of the ADSs; for example, strong downdraft and divergence make ADSs finally land on the Korean Peninsula. We are applying the CNOP-I to identify the sensitive areas for adaptive observations in air quality prediction. We expect to improve air quality forecasts by classifying the synoptic situations that brings about severe ADS outbreaks in South Korea and by identifying the upstream areas for adaptive observations to which we can enhance observations, potentially through international collaborations.

The Indo-Pacific warm pool (IPWP) is a region with one of the strongest atmospheric convections on Earth and deep convection over the IPWP plays a key role in global climate. In recent decades, it has experienced human-induced warming. In addition, it has been observed to have a non-linear relationship between Sea Surface Temperature (SST) and precipitation. Even though SST is increasing, the increase in precipitation is limited up to a specific SST, defined as saturation threshold SST (STT). Existing STT indicates that a completely different state occurs before and after STT showing a non-linear response of precipitation to SST. However, how warmer climates affect changes in the SST-precipitation relationship as well as STT is yet to be unexplored. To investigate future changes in the relationship between SST and precipitation, we used a joint distribution of the two variables with the historical data as well as three Shared Socioeconomic Pathway (SSP) scenarios (SSP2-4.5, SSP3-7.0, and SSP5-8.5) during the near future (2041-2060), and far future (2081- 2100). The results demonstrate that with the shift of mean state, the STT rises due to the involvement of atmospheric stratification in all three scenarios in both future periods, but the largest increase is observed in the SSP5-8.5 scenario during the far future. In the warming climate, greater warming in the upper troposphere compared to the surface leads to stabilization of the troposphere, supporting an increase in STT through Moist-Adiabatic Lapse Rate (MALR) adjustment. This study suggests a change in the SST-precipitation relationship and a possible mechanism for a limited increase in precipitation, and thus provides a background for a better understanding of a non-monotonic precipitation response to SST in the context of climate change.

Convectively-coupled Kelvin wave (CCKW) is one of the important large-scale oscillations in the equatorial area. It is an eastward-propagating mode that constantly causes fluctuations in pressure and winds, and thus, CCKW takes part in shaping the weather and climate through bringing countless rainfall events to countries in tropics. CCKW is relatively active in the Pacific and the Indian basins due to the absence of obstruction from topography, and yet it shows distinct properties over these two regions, including the intensity, active season, symmetry, and propagation speed. The vertical profiles also suggest the differences in tilting angle and the extent of low-level pre-moistening at the leading edge of the convection. It is hypothesized that the differences arise from the mechanisms in triggering the convection associated with CCKW. We conduct a detailed moist static energy (MSE) budget analysis to uncover the dominant factors in driving the energy recharge-discharge process of CCKW. The results indicate that the key factors might lie in the horizontal advection and moisture profile, especially within the boundary layer. The gross moist stability, a measure of development of moist convection, is also tested over the two regions. It is speculated that the background environmental conditions play a major role in determining the structures of CCKW.

In this work, the spatial and temporal distribution characteristics of 171 QLCS cases in North China identified by an objective identification method for the period 2013—2018 are analyzed. According to the statistics of intense weather produced by them, at least two types of QLCSs exist: one type with strong thunderstorm wind gusts and the other with extreme heavy rainfall. Furthermore, the characteristics of circulation pattern, environmental condition, terrain effect, and surface cold pool as a key type of mesoscale system are given for the two types of QLCSs. The findings are presented as follows. The spatial distribution of QLCSs in North China. There are significant differences in the month of occurrence, the spatial scale, the moving speed, the formation time and maintenance period between the two types of QLCSs. The circulation background, environmental conditions and cold pool are also obviously different. The atmospheric baroclinicity is relatively obvious for QLCSs with strong thunderstorm wind gusts. The large value area of BCAPE and DCAPE caused by the dry middle layer and large temperature reduction rate are important environmental conditions for the generation of strong convective wind gusts. The strong cold pool and vertical wind shear within the layer of 0—3 km altitude play an important role in the forward propagation of the QLCSs. Extreme precipitation caused by QLCSs with heavy rainfall is more prominent for this type of QLCSs than that for the previous type. The second type of QLCSs usually occur in weak synoptic-scale forcing systems with sufficient water vapor supply. The back propagation maintained by the interaction between the weak cold pool or the windward slope and the low level southerlies is the main mechanism for the development and slow movement of the QLCS, which is also directly responsible for extreme heavy rainfall.

The latest Multi-angle Imaging Spectro Radiometer (MISR) V23 aerosol optical depth (AOD) products were released, with an improved spatial resolution of 4.4 km, providing an unprecedented opportunity for the refined regional application. To ensure the reliability of their applications and build a scientific reference for the further optimization, it is imperative to conduct a comprehensive evaluation, especially for the unique size-resolved AOD products: small-size AOD (AOD_S, representing the contribution of fine-mode aerosols), medium-size AOD (AOD_M), and large-size AOD (AOD_L), and AOD_M+L represents the AOD part of coarse-mode aerosols. AErosol RObotic NETwork (AERONET) and MODerate-resolution Imaging Spectroradiometer (MODIS) C6.1 aerosol products from 2001 to 2020 are utilized for the validation, analysis, and inter-comparison, considering three spatial scales and four key factors. In general, MISR V23 aerosol products show a good accuracy compared with AERONET. The best performance for all AOD products appears in forest units (the highest R ~ 0.93, data percentage within Expected Error bounds, %EE > 93), related to the inactive human activity and dark underlying surface. Dependences of retrieval deviations illustrate that the performance of MISR AOD deteriorates as aerosol loading increases. Namely, with the increase of aerosols, total AOD (AOD_T) and AOD_S show increasing negative deviations, while increasing positive deviations are observed for AOD_M+L. This suggests that the Empirical Orthogonal Functions do not perform well in this situation, since numerous aerosol particles can obstruct the underlying reflection and reduce the surface spectral contrast. In addition, AOD_T and AOD_S often exhibit anomalous positive deviations in areas with low vegetation cover, such as deserts, revealing that MISR will overestimate aerosol content over bright surfaces and in environments dominated by coarse-mode particles. The above findings not only deepen the understanding of MISR aerosols products from multiple perspectives, but also provide useful information for the product improvement.

The global methane pledge paves a fresh, critical way toward Carbon Neutrality. However, it remains largely invisible and highly controversial due to the fact that planet-scale and plant-level methane retrievals have rarely been coordinated. This has never been more essential within a narrow window to reach the Paris target. Here we present a two-tiered spaceborne architecture to address this issue. Using this framework, we patrol the world, like the United States, China, the Middle East, and North Africa, and simultaneously uncover methane-abundant regions and plumes. These include new super-emitters, potential leakages, and unprecedented multiple plumes in a single source. More importantly, this framework is shown to challenge official emission reports that possibly mislead estimates from global, regional, to site scales, particularly by missing super-emitters. Our results show that, in principle, the above framework can be extended to be multi-tiered by adding upcoming stereoscopic measurements and suitable artificial intelligence, thus versatile for immediate and future monitoring of the global methane pledge.

This study used Weather Research and Forecasting model (WRF) 4.4 and Weather Research and Forecasting model data assimilation system (WRFDA) 3D-var assimilation system to analyze and predict two extreme precipitation events in North China during summer 2022. Based on the system, different satellite data were assimilated in the initial field to improve the accuracy of the model forecast. The satellite data used in this study are NOAA AMSU-A/B, FY-3D MWHS-II/MWTS-II, FY-4A AGRI and Himawari-8 AHI. Artificial intelligence algorithm was applied to these satellite data to retrieve atmospheric temperature and moisture profiles, and these results were used for data assimilation in the system. In this study we have designed three experiments: 1 control run and 2 experiments. The control run only assimilated conventional data to the initial field. Besides the conventional data, the first experiment assimilated AI retrieved atmospheric temperature and moisture profiles and the second experiment assimilated brightness temperature from these sensors. We conducted 72 hours forecast for all the experiments and evaluated the precipitation amount and area with observed data. The results showed that the AI algorithm that we have developed in this study performed well in atmospheric temperature and moisture profile retrieval and the temperature and moisture profiles had positive contributions to precipitation forecast. According to TS, FAR and POD sores, the second experiments showed high accuracy in 0-6 hour precipitation forecast.

Total precipitable water (TPW), a column of water vapor content in the atmosphere, is an important meteorological factor and play a critical role in the occurrence of precipitation and atmospheric river. Traditionally TPW has been calculated from using vertical profiles of temperature and humidity. But operational TPW derived from GK2A provides under only clear sky conditions and TPW calculated from radiosonde observations provides in several stations at 4 times in a day. To overcome temporal and spatial limitation of these TPW, this study proposes a retrieval algorithm based on convolutional neural network (CNN) model to retrieve TPW from geostationary Korea multi-purpose satellite 2A (GEO-KOMPSAT 2A, GK2A) equipped with the Advanced Meteorological Imager (AMI). The TPW is prepared by integrating the water vapor profiles from radiosonde observations over Korea. The algorithm performances are assessed using radiosonde observations for other period as the reference data. Through statistical verification, the results show that the CNN model performs with a root mean square error (RMSE) of 9.42 (5.32) mm, a bias of 0.02 (-0.76) mm, and a correlation coefficient of 0.88 (0.94) under all (clear) sky conditions. This study addressed the availability of satellite-derived TPW to resolve the limitations of in situ measurements and cloudy sky conditions. Both the quantity and the quality of data are important in machine learning approach. We will expand training data like as GNSS and numerical weather forecast model to improve retrieval of TPW.

The National Meteorological Satellite Center of the Korea Meteorological Administration has developed an algorithm to retrieve the atmospheric profiles for Geostationary Interferometric Infrared Sounder (GIIRS) onboard FengYun-4A/B (FY-4A/B). Hyperspectral infrared sounder with thousands of channels has a high vertical resolution, so it can estimate more accurate atmospheric temperature and humidity profiles and is useful for monitoring severe weather such as typhoons and heavy rainfall because it highly depends on atmospheric instability and water vapor information. To develop the algorithm for GIIRS of FY-4A/B, channel selection, and systematic bias correction were performed. The channel was chosen to consider the vertical distribution of the weighting function, except for the channel with strong trace gas absorption and channels with large observation error among 1650 channels of FY-4A GIIRS and 1682 channels of FY-4B GIIRS. The retrieved temperature and humidity profiles with GIIRS/FY-4 for the summer of 2022 were compared with radiosonde and the profiles of the GK2A AMI. The validated results of GIIRS/FY-4 were similar to the GK2A even though more channels were used. So it is needed various sensitivity tests to get more accurate temperature and humidity profiles using hyperspectral infrared sounder such as the number of channels and wavelength selection used by the algorithm. The detailed results and plans are presented at the conference. This work was funded by the Korea Meteorological Administration’s Research and Development Program “Technical Development on Weather Forecast Support and Convergence Service using Meteorological Satellites” under Grant (KMA2020-00120)

This study investigates the radial distribution of deep convective clouds 24 hr prior to the rapid intensification (RI) of the tropical cyclones (TCs) in the Western North Pacific. TCs are categorized according to the 24 hr future intensity (V_max) change (ΔV_max), as follows: non-RI (ΔV_max < 30 kt), short RI (RI-S, ΔV_max > 30 kt for less than a day), and long RI (RI-L, ΔV_max > 30 kt continuously for at least 1 day). RI-S TCs have the strongest intensity and shortest RMW within 24 hr prior to RI, whereas non-RI TCs have the weakest intensity and longest RMW within 24 hr prior to its lifetime maximum V_max. In all categories, highest DCC percentage (DCC-P) and coldest temperature (DCC-T) are found within 3 RMW from the storm’s center. Non-RI has the lowest DCC-P and warmest DCC-T anywhere within 10 RMW, whereas RI-L has the highest DCC-P and coldest DCC-T within 2 RMW especially at the onset of RI. DCC-P starts to increase 12 hr prior to RI with sharpest 3-hr increase of 17% at 2-3 RMW in RI-L. DCC-T within RMW is always colder than that within 2-3 RMW, 9 hr prior to RI. Sharpest DCC-T difference of 3^oC is found in RI-L, at least 3 hr prior to RI. Results in the study can be used to identify the TCs that are more likely to undergo prolonged RI even before the RI onset.

Alisov's climate classification was proposed in 1954, and it focuses on the January–July changes in large-scale air mass zones and their fronts. In this study, data clustering by machine learning was applied to global reanalysis data assimilated from some satellite data to quantitatively and objectively determine air mass zones, which were then used to classify the global climate. The differences in air mass zones between two half-year seasons were used to determine climatic zones, which were then subdivided into continental or maritime climatic regions or according to east–west climatic differences. This study began by questioning whether the global climate can be divided into four air mass zones as Alisov did in the 1950s. The results showed that Alisov's four air mass zones from the 1950s were supported from a modern data-driven perspective using high-quality global reanalysis data. In addition, the clustering technique accurately captured frontal precipitation between air mass zones in the mid-and high latitudes. This study, thus, renews Alisov's climate classification for the first time in almost 70 years and applies data-driven machine learning to establish a standard for air-mass-based climate classification. This paper has been peer-reviewed once in Progress in Earth and Planetary Science (PEPS) and a revised manuscript is in preparation.

Numerical prediction models cannot correctly predict the extended-range (10-30-day) precipitation amount and pattern over Taiwan due to the low predictability. However, demand for extended-range precipitation forecasts by users in different sectors has grown significantly. In this study, a statistical post-processing technique combining Analog Post-processing (AP) and Probability-Matched mean (PM), called APPM, is used to perform bias correction and downscaling for week 3-4 precipitation forecasts in Taiwan. The purpose is to provide users with more accurate Quantitative Precipitation Forecasts (QPF) and more reliable Probabilistic Quantitative Precipitation Forecasts (PQPF). Previous study shows that 1-14-day post-processed precipitation forecasts using APPM, including QPF and PQPF, are promising. Here we want to confirm that the APPM method also works for the week 3-4 forecast range, which was thought to be a “predictability desert” with little forecast skill. Forecast evaluation shows that the raw ensemble is under-dispersive, while the calibrated ensemble distribution well represents the observation variability. Compared to the raw forecasts, the AP-based probabilistic forecasts have better reliability and higher skill in discrimination. Users with a much wider spectrum of cost/loss ratio can benefit more from the calibrated forecasts in decision making as compared to the raw forecast. In addition, the calibrated QPF removes most bias and displays obviously higher correlation and reduced error.

This study examines the environmental conditions that influences the tropical cyclone (TC) occurrence frequency (TCOF) as well as the rain microphysics of TCs over the North Indian Ocean (NIO) during major monsoon stages (pre-monsoon, monsoon, and post-monsoon) in 2014-2021. Two regions in NIO domain are a particular focus, namely, the Arabian Sea (AS) and the Bay of Bengal (BOB). TCs that formed within the NIO domain and reached at least the tropical storm stage were selected. High TCOF is attributed to the combined presence of warm sea surface temperature and cyclonic wind or positive 850-hPa relative vorticity. Moreover, high reflectivity and heavy rainfall are found within the 50 km radius from the storm’s center. Monsoon season TC Vayu (2019) have the heaviest rainfall and highest reflectivity within the inner band region, as compared to pre-monsoon TC Fani (2019) and post-monsoon TC Kyaar (2019). The probability density functions of the rain microphysics parameters vary with the season (pre-monsoon, monsoon, and post-monsoon) and the mode of rain (total, stratiform, and convective). In all season, the breakup process is found to be dominant in total and stratiform precipitating clouds, whereas the break-up and coalescence processes are comparable in convective precipitating clouds.

The Mei-yu season over Taiwan, mainly associated with frontal systems, is a transition period (May and June) between the winter and summer seasons. The interactions of Mei-yu season frontal systems with the complex topography of Taiwan lead to heavy precipitations across the island. The present study investigates rain and microphysical attributes of the Mei-yu season over Taiwan using Global precipitation measurement mission dual-frequency precipitation radar (GPM DPR). To analyze the regional and intra-seasonal characteristics of the Mei-yu season, May and June are segregated into pre-Mei-Yu (5/15-5/31), mid-Mei-Yu (6/1-6/15), post-Mei-Yu (6/16-6/30). There are clear distinctions in the raindrop size distributions among pre-, mid-, and post-Mei-yu seasons, with abundant large drops in the post-Mei-yu. Furthermore, raindrop size distributions also showed apparent differences among four regions (north, south, central, and eastern part) of Taiwan, with relatively bigger drops in central Taiwan. The contoured frequency by altitude diagrams of rain rate, radar reflectivity, mass-weighted mean diameter, and normalized intercept parameters are used to understand the microphysical processes responsible for the regional and intra-seasonal variations. 

From the climatological views, 20-yr (1999-2018) June CFSR Dongsha wind vertical profiles are used to define the MBLJ days. It is found that when MBLJ occurs at Dongsha, the above 90/95-percentile extreme rainfall occurrence frequency over low land (< 1km) of western Taiwan and northern Taiwan is larger than that of NoTC days, with the peak values occur over southwestern slope of the Snow Mountains and the Central Mountain Range, and southwestern Taiwan. It is similar that the above 90/95-percentile extreme rainfall occurrence frequency over high land (> 1 km) of Taiwan is larger during MBLJ days than NoTC days, with peak values occur over southwestern slope of the Snow Mountains and the Central Mountain Range. During early summer rainy season (June) over the South China Sea, there are six CMBLJs (Coastal MBLJs) with occurrence frequency peaks (> 35 %). Their formation mechanisms are related to terrain. (1) and (2) BLJ-East and BLJ-West (East and West of Hainan Island over South China); (3) and (4) Orographic enhanced southwesterly (southwesterly flow impinging on southeastern tip of Vietnam and Taiwan); (5) Barrier jet off the northwestern Taiwan; (6) Gap wind flow through mountain gap over central eastern Vietnam.

Under climate change, the number of typhoons in the Northwest Pacific region will decrease and the proportion of strong typhoons will increase. In order to understand whether the interdecadal changes in the ocean and atmospheric environment affect the development of typhoon intensity. This study compares the changes in the ocean and atmospheric environment from June to November in the past two decades (1980-1999) and the current twenty years (2000-2019), and discusses the relationship between changes in the ocean and atmospheric environment and the typhoon intensity. Compared with the past, the sea surface temperature has increased by an average of 0.42°C, the ocean mixed layer depth has deepened by an average of 2.7 meters, and the ocean heat content (OHC) has increased by 9.89 kJ/cm². The results show that under climate change, the thermal environment of the upper ocean will be more conducive to the development of typhoon intensity. In terms of atmospheric environment, the Pacific subtropical high strengthened and extended westward, and the monsoon trough also tended to retreat westward compared with the past. Under climate change, atmospheric environment has both positive and negative contributions to the development of typhoon intensity, and changes in the Pacific subtropical high, monsoon trough, and vertical wind shear will be unfavorable for the formation of typhoons. Using the observed maximum wind speed of typhoon and potential intensity to analyze the relationship between ocean and atmospheric environment and typhoon intensity. The results show that compared with the atmospheric environment, the ocean environment has a more direct relationship with typhoon intensity. Compared with sea surface temperature, OHC has a higher relationship with typhoon intensity. Moreover, the correlation between OHC and stronger typhoon is even more higher than that of weak typhoon. In conclusion, among all variables, OHC has highest correlation with typhoon intensity.

The coronavirus disease 2019 (COVID-19) is responsible for the worldwide disorder in livelihood and enormous loss of human lives. To curtail the COVID-19 spread, the government of China imposed a lockdown over most of its provinces from 23 January to 25 March 2020. The effects of the COVID-19 lockdown on public health, transport, aviation, and meteorological parameters were well documented globally; however, its influence on rain microphysics is not fully understood. In the present study, we tried to report the distinction in precipitation microphysics, especially the raindrop size distribution between the COVID-19 lockdown (23 January to 25 March 2020) and non-lockdown (23 January to 25 March: 2015- 2029) periods over Taiwan. The raindrop size distributions between the lockdown and non-lockdown periods showed clear distinctions, with more mid- and large-size drops during the lockdown periods. Furthermore, possible microphysical differences between the lockdown and non-lockdown periods are explored using reanalysis and remote-sensing data sets.

Studying the glacial climate provides opportunities to examine the change in climate systems in response to external forcing, e.g., solar radiation, greenhouse gases and the abrupt change in cryosphere. In addition, the climate is one of its external conditions which impacts the development of safety assessment scenarios, such as glacial and interglacial conditions causing sea level variations and the surface system changes. For example, the change in sea level in glacial and interglacial periods strongly determines the groundwater system and the ecosystem. The groundwater system dominates the flow rate which influences the hydrogeological condition in the near field and the potential release path from underground facility to the ecosystem in the far field. Also, the landform controls the release location of radioactive and how it will influence the biosphere. To have a better picture for the processes above and the potential risks, we leverage a GCM for simulating the glacial climate and downscale to the regional states of Taiwan. Specifically, we use ISCA, an idealized global circulation modelling framework, to simulate the glacial climate since 12,000 B.P. (last glacial maximum, LGM) until the present day. Following the standard of Held–Suarez experiment, we modulate the global energy budget through changing TOA radiation, solar zenith angle and proxy surface temperature (due to changed terrain in Taiwan and adjacent region) to mimic different climate states. Our study demonstrates the potential for safety assessment of spent nuclear fuel disposal repository based on model outputs from the idealized GCM along with the regional model downscaling.

The afternoon thunderstorm event is the common nature disaster in Taiwan area. Because of the short life cycle and high variety, it is difficult to observe the microphysics feature for whole system, especially the Drop Size Distribution (DSD), one of the fundamental microphysical parameters. Using dual-polarimetric radar to retrieve DSD is useful to obtain the variation of microphysical processes for the entire weather system. In general, the retrieved method is based on gamma distribution to reduce the degree of freedom and get the DSD parameter, such as shape parameter (µ), slope parameter (Λ), and intercept parameter (N₀). However, the shape of DSD is incompatible with the gamma distribution if the specific microphysical processes are dominated. There is a new method with the moment-based operator to convert the dual-polarimetric variables to the moments of DSD which can avoid the model error. In this study, the new retrieved method is used to study DSD during different life period of the afternoon thunderstorm in Taipei area. Moreover, we try to categorize the feature of DSD under the influence of collisional coalescence, breakup, or drop settling processes. Base on the results of retrieved DSD, the evaporation and coalescence are the main microphysics processes in initial stage and mature stage respectively.

In Summer, high temperatures and humidity are suitable for forage planting, but long days of sunshine are needed for hay preparation in this period to cope with the lack of fresh forage in autumn and winter over Taiwan. Therefore, the livestock industry has a demand for probabilistic forecasts of consecutive days without rainfall. However, due to the complex topographic distribution in Taiwan, rainfall forecast has high uncertainty in space and time and it is very difficult to capture the fine information from model output with low resolution. The purpose of this study is to provide users with reliable and skillful forecasts, which help users obtain more economic benefits in decision making. In this study, Analog post-processing (AP) is applied in 20-year re-forecasts of the National Centers for Environmental Prediction (NCEP) Global Ensemble Forecast System version 12 (GEFS v12) to produce calibrated and downscaled probabilistic forecasts of consecutive 5 days without measurable rainfall over Taiwan. The evaluation results show that (1) The under-dispersion of the raw forecasts is effectively improved through the AP method. (2) The probabilistic forecasts calibrated by AP method have better reliability, brier skill score and discrimination (potential usefulness) than raw forecasts. (3) The calibrated probabilistic forecasts can provide higher relative economic values compared to the raw forecasts, and it indicates the enhancement of forecast value.

Radar echo extrapolation utilizes the observed composite reflectivity to estimate the motion fields of radar echoes and provide advection and rotation information for extrapolation. The uncertainty of motion fields is one of the major error in radar extrapolation. In addition, the weather system may have different moving directions at different heights. In this study, 3-Dimension motion fields is estimated by the entire volume scanned data and applied to MAPLE (McGill Algorithm for Precipitation nowcasting using Lagrangian Extrapolation) nowcasting system. With autumn precipitation events in the Yilan area and the Meiyu front events in Taiwan, the characteristic of 3D motion fields in space and time are analyzed. The results show that u-velocity changes a lot in both time and vertical levels in Mei-yu front event. As for autumn precipitation, the standard deviation and mean of u-velocity are similar to v-velocity. Then, the added value of 3D motion fields for the nowcasting is evaluated by continuous and categorical verification. It is found that when applying 3D motion fields in radar extrapolation nowcasting, it can capture more complete advection and rotation due to orographic effect over Taiwan area. Overall, the improvement of the nowcasting with 3D motion fields can be up to 3-h.

This study mainly discusses the synoptic environment, observed wind speed, wind direction and raindrop size distribution (DSD) when typhoon Maria (2018) and Lekima (2019) approach Taiwan. When typhoon Maria moves westward approaching northern Taiwan coastal ocean, the near surface wind speed strengthens to > 17 m/s which is observed by wind profiler at National Central University (NCU) over northwestern Taiwan. In addition, the observed wind direction using profiler gradually changes from northwesterly to southwesterly. The mass-weighted mean diameter in millimeters (Dm), observed by NCU impact-type Joss-Waldvogel disdrometer (JWD), gradually decrease to ~ 1.5 mm. On the other hand, typhoon Lekima moves northwestward from east of Taiwan. The NCU wind profiler observed near surface wind speed is smaller than 17 m/s and wind direction changes gradually from northeasterly to southwesterly. The Dm observed by NCU JWD changes irregularly as time processes. Research results show that wind speed and wind direction have impacts on DSD. Moreover, observed radar reflectivities show that the principal rainband of typhoon Maria passes northern Taiwan. At that time, the NCU JWD observations show that Dm is large (~ 2.5 mm) and the continuous rainfall > 12 hr. On the other hand, only the distant rainband of typhoon Lekima passes northern Taiwan and the observed Dm is small. The rainfall maintains less than 4 hr. Research results show that if principal rainband passes Taiwan or not influences DSD and the duration of rainfall.

From 0200 to 1000 LST 2 June 2017, the shallow, east–west-oriented mei-yu front (<1 km) cannot move over the Yang-Ming Mountains (with peaks ∼1120 m) when it first arrives. The postfrontal cold air at the surface is deflected by the Yang-Ming Mountains and moves through the Keelung River and Tamsui River valleys into the Taipei basin. The shallow northerly winds are anchored along the northern side of the Yang-Ming Mountains for 8 h. In addition, the southwesterly barrier jet with maximum winds in the 900–950-hPa layer brings in abundant moisture and converges with the northwesterly flow in the southwestern flank of the mei-yu frontal cyclone. Therefore, torrential rain (>600 mm) occurs over the northern side of the Yang-Ming Mountains. From 1100 to 1200 LST, with the gradual deepening of the postfrontal cold air, the front finally passes over the Yang-Ming Mountains and arrives at the Taipei basin, which results in an east–west-oriented rainband with the rainfall maxima over the northwestern coast and Taipei basin. From 1300 to 1400 LST, the frontal rainband continues to move southward with rainfall over the northwestern slopes of the Snow Mountains. In the prefrontal southwesterly flow, the orographic lifting of the moisture-laden low-level winds results in heavy rainfall on the southwestern slopes of the Snow Mountains and the Central Mountain Range. With the terrain of the Yang-Ming Mountains removed in the high-resolution model, the mei-yu front moves quickly southward without a rainfall maximum over the northern tip of Taiwan.

Global cloud-resolving climate simulation is believed to promote our understandings of the atmospheric phenomena with multi-scale nature and also improve reliability of their climate projections such as tropical cyclones and cloud feedback. With this background, the Nonhydrostatic ICosahedral Atmospheric Model (NICAM) has been developed and used for more than two decades. The first realistic global cloud-resolving (here, 3.5 km mesh) simulation was published in 2007, and its extension toward climate scale is now in sight with the supercomputer Fugaku, the latest flagship machine of Japan. The value of such climate simulation is guaranteed by a series of recent model improvements to better simulate both mean state and disturbances. We are also revealing the climatological performance of the ocean-coupled version of NICAM through a series of seasonal to decadal simulations. Other important challenges include a global O(10³) ensemble prediction experiments with 14 km mesh and a global 220 m experiments toward an era of global large-eddy simulation. The above challenges on the supercomputer Fugaku would be harder without an effective co-design activity with the application developers. In terms of post-processing, we recently modified our MPI-based remapping tool and confirm its applicability to a global 870 m mesh data.

Mesoscale convective systems (MCSs) contribute a large fraction of warm-season precipitation and generate hazardous weather with substantial socio-economic impacts. Uncertainties in convection parameterizations in climate models limit our understanding of MCS characteristics and reliability of future projection. We examine MCSs simulated by the global 14-km Nonhydrostatic ICosahedral Atmospheric Model (NICAM) without cumulus parameterization against satellite observation from Global Precipitation Measurement (GPM) during 2001-2008. We focus on MCSs over the central U.S. and eastern China where MCSs are prevalent from March to August. A process-oriented tracking method incorporating both cloud and precipitation criteria is used to identify and track MCSs. About 140/100 MCSs initiate in the central U.S./eastern China per warm season and most of them initiate east of high mountains and in coastal regions. The frequency distribution of MCS lifetime is well captured by NICAM. However the simulated MCSs have stronger precipitation, smaller precipitation area, and larger cold cloud system than observed in both regions, which may be caused by weak entrainment as it is not well resolved at 14 km resolution. The simulated MCS number is also underestimated in summer. By examining the climatological and MCS large-scale environments, the significant underestimation of MCS number in summer over the central U.S. may be attributed to the large climatological dry bias in the atmosphere. For China, mean moisture in summer is well simulated but deficiency in capturing the dynamic condition related to the coastal topography for triggering convection may have contributed to underestimation of MCS even in a sufficiently moist environment.

Machine learning technology has recently been applied to weather and climate simulations. Compared to atmospheric models that include various algorithms, data-driven models can reduce the amount of computation (= elapsed time) and achieve significant speedups. A data-driven model for weather forecasting is being developed using datasets such as weather reanalysis (e.g., ERA5) and multi-model intercomparison (e.g., CMIP6). However, for meteorological/climate researchers who want to elucidate processes and mechanisms, data-driven models shouldn't be used as complete black boxes. We need high-speed, on-demand "surrogate models" to imitate the physical models that algorithms are based on understanding processes and continuously developed.
We developed a coupling library h3-Open-UTIL/MP (Arakawa et al., 2022) to bridge legacy weather/climate models and modern machine learning tools. This library was initially designed to connect each component of the earth system model. We extract the necessary training data from the part of the model code written in Fortran and pass it to ML frameworks in Python on-the-fly. The proposed system enables the creation of the necessary surrogate models step-by-step using a large amount of simulation data. We will present an overview of the library and initial experimental results.

Nowadays, extreme precipitation has brought very serious losses. How to timely and accurately forecast precipitation has become the top priority of disaster prevention. The emergence of deep learning technology has provided a new idea for the precipitation nowcasting, and achieved good results in the short forecast period, but there is still a problem that as the forecast period becomes longer, the forecast effect becomes worse. On the basis of deep learning technology, this study attempts to introduce additional physical information and simulate the future information of numerical model forecast to improve the accuracy of precipitation nowcasting. This study uses precipitation data, and two-dimensional water vapor flux data in the input and forecast stages respectively. Because there are errors in the real forecast, random errors are introduced when additional future data is added into the forecast stage. The results show that, compared with the simple application of precipitation data, the introduction of more physical information in the input stage can improve the prediction effect by 4.75%, while the introduction of predicted future physical information in the forecast stage can improve the effect by 16.69%, showing a strong application prospect.

Urban expansion has been rapid in North China Plain (NCP) since 2000, and the metropolitan area of 2015 increased by 4.92% compared with 2000. The change in Land-Use and Land-Cover (LULC) directly affects the meteorological conditions, impacting the atmospheric pollutant concentrations. Therefore, based on the MERIS300 LULC product of 2015, this study simulates a continuous heavy pollution incident from 05 December 2015 to 02 January 2016 in NCP, to explore whether the urban expansion causes LULC changes that will worsen the atmospheric environment. The T2 increases by 0.19 ℃ while the relative humidity (RH) decreases by 1.28%. Urbanization aggravates the pollution of O₃ but reduces the other four concentrations of pollutants. The concentration of PM_2.5 increased significantly. In addition, the trends of PM_2.5 and O3 are significantly negatively correlated in vertical distribution. Therefore, it is necessary for PM_2.5 and O₃ to cooperate in emission reduction.

Changes in land use and land cover (LULC) influence meteorological fields and biogenic emissions, further affecting the atmospheric chemistry and air quality. Combing the satellite measurements and WRF-Chem model simulations, we evaluate impacts of the LULC change between 2001 and 2018 on the summertime ozone (O₃) formation in North China Plain and surrounding areas (NCPs). Satellite measurements have revealed that from Taihang to Yanshan Mountain, the fraction of broadleaf and needle forest coverage has increased by 5%-20% and the urban area has increased by up to 25% in the NCP. Additionally, the vegetation density has increased significantly in the NCPs except for urban areas. The LULC change generally enhances biogenic volatile compounds emissions in the NCPs, particularly over Taihang and Yanshan mountain, but the O₃ variation is divergent. The maximum daily 8-hours average (MDA8) O₃ concentrations are reduced by 1%-7% over Taihang and Yanshan Mountain because the raised vegetation density increases O₃ dry deposition velocity to accelerate the O₃ loss. The raised vegetation density enhances the evapotranspiration to decrease the near-surface temperature by 0.1 ℃ ~ 1.5 ℃, which also generates a divergence in the low-level atmosphere in the NCPs, causing secondary northerly or easterly winds in the NCP. The O₃ enhancement along the coastal areas of the NCP is attributed to the perturbation of wind fields and photolysis induced by the LULC change. The divergent variation of the MDA8 O₃ concentrations in the NCP is generally caused by the variations of biogenic emissions and photolysis.

Solar farms consist of arrays of solar panels to convert solar energy to electricity. Building solar farms has become an acknowledged approach to mitigate global climate change. Meanwhile, solar farm deployments change the surface radiative properties and, thus, can impact the local surface radiation budget and climate. So far, global modeling studies have primarily focused on simulating the impact caused by surface albedo changes due to large-scale solar farm deployments. Such modeling studies lack long-term observational constraints and do not consider complex atmosphere-panel-surface interactions. Using Moderate Resolution Imaging Spectroradiometers (MODIS) aboard the NASA Aqua satellite, we have quantified the change of surface spectral reflectances, spectral emissivities, and land surface temperature over six solar farms commissioned in the southwestern U.S. Results suggest a 20-25% reduction of surface reflectance over the seven MODIS shortwave bands due to the solar panel installation, leading to a ~23% decrease in the surface upward shortwave broadband flux and a ~14-18% decrease in the clear-sky reflected shortwave flux at the top of the atmosphere. Despite the uncertainty and disagreement in current surface retrieval algorithms, results show that the outgoing longwave radiances over the MODIS infrared window channels are reduced. Based on these observations, we will implement a more realistic representation of solar farms in global climate models. Specifically, we will put a translational layer in the Community Terrestrial System Model, the land component in CESM. This layer will modify the radiative and momentum fluxes given solar panels’ geometrical and radiative properties on top of the natural surfaces. The new scheme can output panel temperature to be compared to and explain in-situ and satellite observations. It will help us understand the potential impact on the climate system if an increasing number of solar farms are deployed in the future.

The critical zone (CZ), extending from the bottom of groundwater to the top of the lower atmosphere, plays a crucial role in shaping the environment and human well-being (Brantley 2007). The CZ and climate systems are affected by various natural and human-driven processes. In this study, we utilized a modified version of CESM2 that incorporated data on groundwater, irrigation, and human water use. The groundwater scheme has also been enhanced to simulate lateral flow and aquifer pumping. To assess the impact of irrigation on climate, we conducted two experiments spanning from 1901 to 2014: one with a fixed irrigation demand in 1901, and the other with a changing irrigation demand (Sibert 2015). Our results indicate that irrigation is concentrated in India, China, and North America. In India, the irrigation effect leads to increased soil moisture, latent heat, and decreased sensible heat, resulting in a cooling effect. Rainfall is also seen to increase, particularly in the Ganges river region. Conversely, the west coast of the USA experiences contrasting feedback with decreased rainfall and soil moisture. References: Brantley SL, Goldhaber MB, Ragnarsdottir KV. (2007). Crossing disciplines and scales to understand the Critical Zone. Elements 3: 307-314. Siebert S, Kummu M, Porkka M, et al. A global data set of the extent of irrigated land from 1900 to 2005[J]. Hydrology and Earth System Sciences, 2015, 19(3): 1521-1545.

Severe groundwater depletion due to intensive irrigation has begun to emerge and has been clearly detected by satellite-based estimates in major agricultural areas, particularly the North China Plain. The effects of groundwater overexploitation are not restricted locally, but an incursion of dormant groundwater into the active hydrological cycle can critically affect the hydroclimatology of remote places and the global water budget, which eventually leads to terrestrial contributions to rising sea levels. This study explores how much groundwater depletion as a result of intensive irrigation will contribute to accelerating global sea level rise. For this, the long-term simulations with/without groundwater pumping are performed using the fully coupled configuration in Community Earth System Model (CESM). The groundwater-fed irrigation is activated in the North China Plain, the entire China, and global scale to discern different responses to the locations of the irrigation forcing. A comparative analysis of a suite of sensitivity experiments can facilitate identifying the obscure effect of groundwater pumping from the background climates. In particular, the quantification of the contributions of groundwater pumping to sea level changes will be focused, which will strengthen our understanding of the nonlinear climate processes that occur when the atmosphere, land surface, and subsurface are perturbed by groundwater pumping. [Acknowledgements]This research was supported by project GRF16309719, which was funded by the Research Grants Council (RGC) of Hong Kong. 

Wildfire hazards have witnessed a substantial increase in frequency, extent, and severity globally. While wildfire outbreaks and spread are strongly affected by non-climatic factors such as campfires, forest management, and monitoring capacity, there is a growing scientific consensus that increasing temperatures and resultant dryness may be responsible for a favorable environment for wildfires. A potential escalation of wildfires under global warming could result in changes to the land cover and carbon emissions, which may in turn have an impact on regional climate. However, complex climate-soil-vegetation-interacting processes leave numerous uncertainties in wildfire projections and their impacts under the acceleration of global warming. In this study, Forest Fire Danger Index (FFDI) and Fire Weather Index are calculated based on the Coupled Model Intercomparison Project (CMIP) multi-model projections in order to investigate future changes in fire weather danger under different emission pathways. The regional hotspots which are expected to face a heightened risk of fire weather are then focused on to further comprehend the wildfire and land cover feedback. (Acknowledgements: This study was supported by the Special Research Support Scheme from the Hong Kong University of Science and Technology (R9055), which was funded with the donation from the Chau Hoi Shuen Foundation.)

In this study, the fast response of East Asian summer precipitation to COVID-induced aerosol emission reductions is examined using the Community Earth System Model version 2.2 (CESM2.2). The emission reductions decreased aerosol optical depth and cloud cover over northern China in June 2020. The troposphere became warmer, strengthening the land-sea thermal contrast and anomalous southerly winds. The subtropical westerly jet accelerated and shifted southwards, favoring the low-level convergence, upward air motions, and subsequent condensational heating over the Yangtze River Basin (YRB). The feedback of condensational heating in return strengthened the convergence and ascent. The western North Pacific Subtropical High was intensified, which further enhanced the moisture advection and convergence over the YRB. Both the enhanced moisture convergence and ascent increased precipitation over the YRB during June 2020. Furthermore, local and remote emission reductions show different impacts on convection and moisture transport over the YRB. The emission reductions over China caused stronger convective precipitation (1.15 mm/day versus 0.63 mm/day), but weaker larger-scale precipitation (1.17 mm/day versus 2.24 mm/day) than the emission reductions outside China did.

The presence of biological (airborne pollen) and non-biological (PM_2.5, PM₁₀, O₃, CO, NO₂, SO₂, etc.) particles in the atmosphere is associated with the incidence of adverse allergenic reactions affecting human health. Considering all the probable effects of air pollutants and airborne pollen on allergic reactions, the present study mainly examines the relationship between the daily concentrations of airborne pollen, PM_2.5, PM₁₀, O₃, CO, NO₂, SO₂, and the number of patients with allergic rhinitis (AR) at the same time in northern Beijing, China by statistical analysis. The results showed that the daily visiting number of AR patients was positively associated with the concentration of airborne pollen, as well as the pollutants NO₂ and PM₁₀ particles, while, it was negatively associated with the concentrations of O₃ and CO. Airborne pollen and part air particles (e.g., NO₂, PM₁₀) had a significant impact on the morbidity of AR patients. Thus, public health and clinical approaches are needed to anticipate and reduce the morbidity of allergic disease caused by airborne pollen and non-biological particles.

Deep convective clouds and associated anvils exert opposite radiative effects. The impact of different aerosol types on these two categories of clouds remains a major challenge in understanding aerosol-cloud interactions. Using 11-year satellite retrievals, we find that cloud top height (CTH) and ice cloud fraction of deep convective clouds and anvil cirrus identified by Cloud-Aerosol Lidar with Orthogonal Polarization increase with small aerosol loadings and level off or even decrease with further aerosol increase. Compared with continental aerosols, CTH affected by marine aerosols starts to decrease at smaller aerosol loadings. Moreover, cloud optical depth (COD) of deep convective clouds decreases with aerosol loadings. COD of anvil cirrus increases with increased loadings of most aerosol types but decreases with smoke aerosol. These relationships are mainly attributed to the aerosol effect rather than the meteorological effects. Our findings contribute to the development of models and better assessment of aerosol-cloud radiative forcing.

Interactions between aerosols and gases in the atmosphere have been the focus of an increasing number of studies in recent years. Here, we focus on aerosol effects on tropospheric ozone that involve meteorological feedbacks induced by aerosol–radiation interactions. Specifically, we study the effects that involve aerosol influences on the transport of gaseous pollutants and on atmospheric moisture, both of which can impact ozone chemistry. For this purpose, we use the UK Earth System Model (UKESM1), with which we performed sensitivity simulations including and excluding the aerosol direct radiative effect (ADE) on atmospheric chemistry, and focused our analysis on an area with a high aerosol presence, namely China. By comparing the simulations, we found that ADE reduced shortwave radiation by 11 % in China and consequently led to lower turbulent kinetic energy, weaker horizontal winds and a shallower boundary layer (with a maximum of 102.28 m reduction in north China). On the one hand, the suppressed boundary layer limited the export and diffusion of pollutants and increased the concentration of CO, SO₂, NO, NO₂, PM_2.5 and PM₁₀ in the aerosol-rich regions. The NO/NO₂ ratio generally increased and led to more ozone depletion. On the other hand, the boundary layer top acted as a barrier that trapped moisture at lower altitudes and reduced the moisture at higher altitudes (the specific humidity was reduced by 1.69 % at 1493 m on average in China). Due to reduced water vapour, fewer clouds were formed and more sunlight reached the surface, so the photochemical production of ozone increased. Under the combined effect of the two meteorology feedback methods, the annual average ozone concentration in China declined by 2.01 ppb (6.2 %), which was found to bring the model into closer agreement with surface ozone measurements from different parts of China.

Central Asia is prone to wildfires, but the relationship between wildfires and climatic factors in this area is still not clear. In this study, the spatiotemporal variation in wildfire activities across Central Asia during 1997–2016 in terms of the burned area (BA) was investigated with Global Fire Emission Database version 4s (GFED4s). The relationship between BA and climatic factors in the region was also analyzed. The results reveal that more than 90% of the BA across Central Asia is located in Kazakhstan. The peak BA occurs from June to September, and remarkable interannual variation in wildfire activities occurs in western central Kazakhstan (WCKZ). At the interannual scale, the BA is negatively correlated with precipitation (correlation coefficient r = –0.66), soil moisture (r = –0.68), and relative humidity (r = –0.65), while it is positively correlated with the frequency of hot days (r = 0.37) during the burning season (from June to September). Composite analysis suggests that the years in which the BA is higher are generally associated with positive geopotential height anomalies at 500 hPa over the WCKZ region, which lead to the strengthening of the downdraft at 500 hPa and the weakening of westerlies at 850 hPa over the region. The weakened westerlies suppress the transport of water vapor from the Atlantic Ocean to the WCKZ region, resulting in decreased precipitation, soil moisture, and relative humidity in the lower atmosphere over the WCKZ region; these conditions promote an increase in BA throughout the region. Moreover, the westerly circulation index is positively correlated (r = 0.53) with precipitation anomalies and negatively correlated (r = –0.37) with BA anomalies in the WCKZ region during the burning season, which further underscores that wildfires associated with atmospheric circulation systems are becoming an increasingly important component of the relationship between climate and wildfire.

This study examines the seasonal and diurnal variations of carbon dioxide and energy fluxes over three land cover types of Nepal by using the eddy covariance method from March to November 2016. The surface energy balance closures were moderate with the values of about 56%, 61%, and 64% closure at Kirtipur, Simara, and Tarahara sites respectively. The monthly average values of net radiation flux and latent heat flux peaked in August at Kirtipur and Tarahara sites whereas in June at the Simara site respectively. The maximum monthly average measured sensible heat flux was 37 W m⁻², 43.6 W m⁻², and 36.3 W m⁻² in April for all the sites whereas soil heat flux was 5.1 W m⁻² and 2.9 W m⁻² in April for Kirtipur and Simara sites and 6.2 W m⁻² in June for the Tarahara site. The magnitude of diurnal peak of net ecosystem CO2 exchange (NEE) reached up to 11.04 μmol m⁻² s⁻¹ at Kirtipur, 15.04 μmol m⁻² s⁻¹ at Simara, and 10.44 μmol m⁻² s⁻¹ at Tarahara sites respectively. Among the three study sites, the ecosystem at the Kirtipur site was a good carbon source; the ecosystems at Simara and Tarahara sites were low and good carbon sink in the growing season. In addition, all three different land cover ecosystem were carbon source when accounted for the measurement period.

In this study, we performed a synoptic analysis to investigate the reason for the spatial difference in PM_2.5 concentration despite a similar synoptic pattern. Synoptic pressure patterns associated with high PM_2.5 episodes (greater than 35 ㎍/㎥) were classified into three sub-groups related to high concentrations occurring only in Busan and Seoul metropolitan areas using K-means cluster analysis, based on the 900 hPa geopotential height of NCEP FNL data. Although the synoptic patterns of high PM_2.5 episodes that occur independently in Busan and Seoul metropolitan areas were similar, there was a difference in the intensity of pressure gradient and its direction, which tends to be an important factor determining the movement time of pollutants. The spatial difference in PM_2.5 concentration in the Korean Peninsula is due to the difference and direction of the atmospheric pressure gradient that develops from southwest to northeast direction.

With fast economic development, urban expansion and haze pollution are two main eco-environmental challenges confronting China. Though their individual impacts on urban climate and environment are investigated by multiple studies, the synthetic effect and intricate interaction of these two effects have not been fully understood. Here, based on in-situ measurements and numerical simulations at the densely populated and heavily polluted Beijing-Tianjin-Hebei (BTH) city cluster, the respective impact of aerosol and urbanization is quantitatively identified. Under polluted conditions, the aerosol radiative effect dominates with 3 ^oC cooling at ground surface and 1 ^oC warming in the upper boundary layer. In contrast, urban heat island (UHI) effect is more prominent during clean days, leading to over 1 ^oC warming in the lower atmosphere. Furthermore, aerosols are found to enhance UHI via confining the dispersion of anthropogenic heat and perturbing longwave radiation budget. Since the implement of China’s Clean Air Action in 2013, aerosol reduction contributes to a decreasing trend of both aerosol radiative effect and UHI intensity in BTH city cluster. The study highlights a comprehensive understanding of air pollution and urbanization as well as their interactions over megacities.

The Global Environmental Measurement and Monitoring (GEMM) Initiative is an international project of Optica and the American Geophysical Union seeking to provide precise and usable environmental data for local impact. The Initiative brings together science, technology, and policy stakeholders to address critical environmental challenges and provide solutions to inform policy decisions on greenhouse gases (GHGs) and air and water quality. GEMM Centers are currently established in Scotland, Canada, New Zealand, and the United States. These Centers represent partnerships with leading institutions that are actively working toward developing or deploying new measurement technology and improved climate models. Additional Centers are under development in India and Australia with plans to expand to Asia and Africa. In addition to establishing monitoring centers worldwide, GEMM actively engages with other sectors (including industry, standards organizations, and regional or national governments) to support the incorporation or adoption of these evidence-based approaches into decision making processes. For example, Glasgow, Scotland is piloting the GEMM Urban Air Project, deploying a low-cost, real-time, ground-based network of devices that continuously monitors GHGs and air pollutants at a neighborhood scale. The sensor network in Glasgow is increasing the precision of local models that can provide the city with information to assess current policies and support future action. Here we will share the progress and outputs of the GEMM Initiative to date and highlight paths forward to grow the network.

Antarctic sea ice plays an important role in polar ecosystems and global climate, while its variability is affected by many factors. Teleconnections between the tropical and high latitudes have profound impacts on Antarctic climate changes through the stationary Rossby wave mechanism. Recent studies have connected long-term Antarctic sea ice changes to multidecadal variabilities of the tropical ocean, including the Atlantic Multidecadal Oscillation and the Interdecadal Pacific Oscillation. On interannual timescales, whether an impact exists from teleconnection of the tropical Atlantic is not clear. Here we find an impact of sea surface temperature (SST) variability of the tropical Atlantic meridional dipole mode on Antarctic sea ice that is most prominent in austral autumn. The meridional dipole SST anomalies in the tropical Atlantic force deep convection anomalies locally and over the tropical Pacific, generating stationary Rossby wave trains propagating eastward and poleward, which induce atmospheric circulation anomalies affecting sea ice. Specifically, convective anomalies over the equatorial Atlantic and Pacific are opposite-signed, accompanied by anomalous wave sources over the subtropical Southern Hemisphere. The planetary-scale atmospheric response has significant impacts on sea ice concentration anomalies in the Ross Sea, near the Antarctic Peninsula, and east of the Weddell Sea.

The El Niño-Southern Oscillation (ENSO) influences West Antarctica via atmospheric teleconnection, but it is delayed by one season because the Amundsen Sea Low (ASL) response to the ENSO has the strongest intensity in May. However, the mechanism related to the delay has not been fully discovered yet. This study investigates the formation mechanism of the ENSO-driven teleconnection in each month by examining the kinetic energy conversion (CK), total wavenumber, and wave activity flux. The perturbation induced by the ENSO gains energy from the basic state as the ENSO appears, however it does reach high latitudes only in May. Although the ENSO becomes weak in May, the developed subtropical jet makes waves to move further south, resulting in the strong ASL anomaly. Numerical experiments considering the decay of the ENSO forcing also show that the ENSO-related teleconnection is the strongest in May. The results demonstrate that the basic state is essential for the ENSO teleconnection to extend to West Antarctica.

In November 2020, Taiwan implemented a 5-day intensive observation experiment on the Lanyang Plain in the northeast, which we call the Yilan Experiment for Severe Rainfall (YESRain). The reason for this project is that according to the statistics of the Taiwan Central Meteorological Bureau over the years, when the northeast monsoon occurs, the Lanyang Plain often receives the most precipitation in Taiwan, and this precipitation is even higher than the Keelung area on the windward side.In order to analyze the reasons for the rainfall in the Lanyang Plain under the strong northeast monsoon, this study analyzed and simulated the moveable rain belt, rainfall time and rainfall intensity caused by the strong northeast monsoon on November 23, 2020. In addition, various observation tools such as laser wind profiler are used to analyze the relationship between the wind field and the rain belt after the interaction between the bell mouth terrain in Yilan area and the wind field at different heights, and to verify each other with the WRF numerical simulation results. It was found that the northeastern part of Taiwan was affected by the strong northeast monsoon on the 23rd, and the severe rainfall in the Yilan area not only appeared on the south windward side, but also occurred in the Lanyang Plain and the outer sea. The precipitation distribution is highly related to the direction and speed of the wind field, while The surface flux has a considerable contribution to the severe rainfall in Yilan.

There are typical spatial rainfall patterns in the northeast Taiwan area (Yi-Lan) under the northeasterly in wintertime. Such rainfall features are related to the interaction of large-scale background mean flow and complex topography. According to the climatology of rainfall, most precipitation hot spots occur on the windward side of the terrain and decrease toward the plain area. However, the difference in large-scale environmental conditions can lead the variations in spatial rainfall characteristics. By analyzing the gridded rainfall data in the Yilan area in the past 60 years, we noticed that the average rainfall has an increasing trend. The increasing trend was more significant in the southern mountain region on the windward side. The rainfall increase in winter was more extensive than in autumn and spread to the plains in coastal areas. We have also noticed that there will still be heavy rain (HR, >80mm/day) events in the Yilan area in autumn and winter when Taiwan has less rainfall. The probability of HR events has increased significantly in the past 30 years, and HR events also occurred in plain areas. On the other hand, although the probability of occurrence was low, there still were torrential rain (TR, >200mm/day) events in the mountainous areas of Yilan. The frequency and extent of occurrence of these types of extreme rainfall events have significantly increased during 2016-2020 compared with 1961-2015. According to the reanalysis data, the background large-scale circulation field has also changed in the past 60 years. The change in the background circulation led to changes in the frequency of occurrence and hot-spot locations of convection initiation, which also caused variations in the spatial distribution of rainfall.

Mountain-waves, also known as Lee wave, can generate strong rotational forces or up and down air currents that can affect aircraft operations. If mountain-wave clouds exist during the calculation of atmospheric motion vectors using satellite data, the vector may become inaccurate. Therefore, it is important to analyze the atmospheric flow characteristics and generation mechanisms of mountain-wave formation. In this study, we present a satellite-based characteristics of mountain-wave events in the Taebaek Mountain range in Gangwon-do, Korea. Our analysis was based on detailed visual inspection of infrared satellite imagery in the ch13 (10.5㎛) of the GK-2A satellite during 2020-2022. Two analysts independently inspecting the same satellite images using visual inspection methods, and cross-verifying the results. If there are enough moisture in the incoming airstream, mountain-wave clouds are generated parallel to the mountain range from the ridge of mountain-wave that have the characteristics of normal waves. Therefore, we detected clouds whose positions did not change over time and observed the duration of their occurrence. Previous research used visible satellite imagery in 15-min intervals during daylight hours. This had limitations as it could not identify mountain-wave cloud that occurred or lasted at nighttime hours and the 15-min interval was too long to detect the exact duration. But, using the infrared imagery from the GK-2A satellite with 2-min intervals, we were able to overcome these limitations and achieve more accurate observations. The maximum number of mountain-wave events occurred in winter season, with the largest number in February. Conversely, the minimum number of mountain-wave events occurred in summer season, with the fewest number in July. The average duration of mountain-wave events was longest in October and shortest in June. The maximum duration is 2268-min, while the minimum is 48-min.

As the latitude of the continental cold high-pressure center was higher and gradually moves eastward to the sea, the environmental wind field in northern Taiwan was mostly northeasterly or easterly winds introduced by continental high-pressure peripheral circulation. Though the distance between the Taipei Basin and the Yilan of northern Taiwan were closed, the local wind fields were quite different. In order to understand the relationship between the environmental wind field (northeast monsoon), local circulation, and complex terrain effects, this study applied the Weather Research and Forecasting Model, simulating the change of local circulation in the Taipei Basin and Yilan under the northeast monsoon on October 25 and November 26, 2022.It was found that when the wind field in the Taipei Basin is weak northeasterly, local circulation will be generated after noon due to the influence of the terrain, and the wind field will gradually change to easterly or southeasterly easterly. With such airflow characteristics, and the weather tended to be stable. When the wind field of the same environment passed through the Lanyang Plain, it was affected by the terrain effect, and the area where the wind field convergence zone was formed in the Lanyang Plain has Nimbostratus and indirect precipitation. In addition, a northwest-southeast trending wind direction convergence zone was formed in the vicinity of Sansing. As the terrain blocking and local circulation became more obvious, the wind direction gradually turned to westerly and went out to ocean. When the northeast monsoon was strong, the large-scale ambient wind field would directly change the boundary layer structure of the Taipei Basin and the Lanyang Plain. However, under the same northeast monsoon environment, there would still be obvious differences in the boundary layer wind field characteristics between the Taipei Basin and the Lanyang Plain.

The rain gauges over Yilan observed the heavy rainfall during 15 Oct 2022 to 17 Oct 2022 with maximum accumulated rainfall is 972.5 mm within 40 hours. During 14 Oct 2022 to 18 Oct 2022, typhoon Nesat carrying abundant moisture moved westward, passing through the sea south of Taiwan. After Nesat moved into South China Sea, the northeasterly wind enhanced. In this case, the maximum moisture flux exceeded 300 g kg^-1 m s^-1 near 950-hPa level. Typhoon Nesat provided abundant moisture when it was on southeast and south of Taiwan. The northeasterly wind played an important role in moisture transport after Nesat passed by Taiwan. The wind profile radar at Dongshan observed that during 1600 UTC 15 Oct to 0800 UTC 16 Oct, the low-level wind over Yilan was consistent easterly wind with the maximum wind speed ~20 m s^-1. After 2000 UTC 16 Oct, the northeasterly wind advanced into Yilan, and the thickness of northeasterly wind was ~3 km. The maximum wind speed of northeasterly wind was near 1 km height, and the wind speed exceeded 20 m s^-1. After 1600 UTC 17 Oct, the northeasterly wind weakened, and the westerly winds were present in the planetary boundary layer. In this case, the characteristics of drop size distribution over southern Yilan plain showed two different types. One was affect by typhoon circulation mainly (during 1200 UTC 15 Oct to 1200 UTC 16 Oct), the rain rate was stronger with larger diameter (D_m) and smaller number concentration (N_w). The other type was under northeasterly wind, and the rain rate was smaller with smaller D_m and larger N_w.

The heavy rainfall caused by Mesoscale Convective Systems (MCS) has been increasing recently. As a result, there is a growing emphasis on studying MCS. The formation and development of MCS are associated with the thermodynamic and dynamical environment, which has been mostly investigated using a numerical model. There have been limited attempts to utilize in situ rawinsonde measurements, which provide an accurate and high-resolution thermodynamic and dynamic profile, to characterize the MCS over South Korea. In this study, the sounding-derived parameters of the MCS occurring on the Korean Peninsula were presented using the data obtained in the summer of four years (2018–2021).
We classified the MCS cases into four types: convective cells (CC), mesoscale convective complex (MCC), diagonal squall line (SLD), and parallel squall line (SLD). The results show that CC was generated by thermal instability and had the largest convective potential available energy at 470.8 m² s^-2. The total Richardson number had 27.0, exhibiting a distinctly large value compared to other types, but mechanical instability was the smallest in the MCS types. MCC had a considerable convective inhibitory energy of 174.5 m² s^-2 and a large low-layer shear of 7.8 m s^-1. SLD and SLP had a large vertical wind shear, with mechanical instability contributing to formation and development. Sounding-derived parameters during the formation, development, and extinction cycles of MCS suggested that the thermodynamic values gradually increased until they reached their highest values at the peak of convective system development, followed by decreasing trend during the dissipating stage. This study will present the favorable environmental condition for each MCS type to identify the key parameters in determining the MCS type and the development stage.

ACKNOWLEDGEMENT
This work was supported by the National Research Foundation of Korea grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).

Improving the predictability of precipitation requires a comprehensive understanding of precipitation growth and decay (GD). Investigating the precipitation GD has mostly relied on the Eulerian approach, which examines the physical characteristics in a fixed coordinate. On the other hand, Lagrangian approach follows the precipitation system motion and enables capturing its temporal evolution. This study quantitatively retrieves radar-based GD of precipitation in South Korea in a framework of semi-Lagrangian advection. This study is followed by the investigation of the dependence of GD on the following factors: flow direction, flow speed, and relative geographical location to the topography and land–ocean boundary. The results indicate that the flow direction determines the spatial distribution of GD. The growth generally tends to occur on the windward side of the mountain range and vice versa on the lee side. The flow speed affects the intensity of GD. The study further explores the diurnal variability of the GD. Under the strong orographic forcing on the diurnal variation, unique geographical features of South Korea created interesting diurnal patterns near Seoul by land–ocean breeze circulation and on the windward side of the mountain ranges (inland) by solar heating. The GD in South Korea shows two diurnal peaks of growth: the early morning peak over the ocean and the afternoon peak in the land areas. In addition, the GD in South Korea exhibits a monthly variation with the most intense growth occurring in August.

ACKNOWLEDGMENT
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).

The mountains have been getting warmer in recent years. Special interest was put on mountain snow cover change as their environment was accounted to the cooler alpine climate zone. Snow cover is expected to reduce in response to global warming and is most evident in mountainous regions. Using available reanalysis and observation data, a decline in snow cover and snow depth was detected over the last 43 years. Our study considers the relationship between global warming and changes in snow persistence. We found that the mean global warming rate is higher than mountain warming. In addition, polar amplification and global warming do not play a significant role in snow cover changes on individual snow persistence mountains. However, local processes can influence snow cover, and long-term changes in these radiation budgets are significant drivers of the changes being observed. We conclude that regional radiation imbalance is a major contributor to the decline in snow cover. The results reveal new insights into the course of snow cover changes and their implication on the mountain climate. Understanding the impact of changing snow cover patterns is critical for informing sustainable development and preparing for the potential consequences of these changes.

The Yeongdong region (the eastern mountainous region of Korea) is frequently vulnerable to heavy snowfall. By virtue of a lot of previous efforts on the snowfall study, snowfall forecast has been significantly improved, but the phase change of solid precipitation is still hard to predict since its time scale is very short (a couple of hours) and spatial dimension is very limited, which is largely attributable to Taeback mountains and East Sea effect. For instance, two ice pellet (IP) episode (1 March 2021 & 14 March 2018) occurred along with a rapid cooling by strong cold advection during the transition period of winter to spring, when a reversed S profile of temperature was identified in both rawinsonde soundings and model reanalysis. Another episode of phase change was observed on 19 March 2022 when dramatic change of precipitation occurred in the process of rain to snow. Precipitation started on 17 March in the rain phase and persisted 3 days when the Low-pressure system passed over the Korean peninsula. Phase changed from rain to aggregate of dendrites in the morning of 19 March, and melting snow and graupel around noon in the condition of low-level cooling together with surface warming explained by thermodynamic soundings. Later, the habit of snow crystal returned to aggregates. The dramatic change of phase and habit occurred within several hours, which was observed by multi-angle snowflake camera and PARSIVEL together. The results emphasize that cold air intrusion with barrier wind plays an important role in phase change of solid precipitation such as ice pellet (or freezing rain) with a help of cold air blocking by the mountains during the transition period of winter to spring. It is necessary to examine other phase change episodes, and to understand the various roles of cold air advection along the mountains in precipitation.

This study used Doppler radar observations, a dense rain gauge network, and disdrometers to investigate the kinematic and microphysical characteristics of orographic precipitation over Da-Tun (DT) Mountain of northern Taiwan associated with an outer rainband of Typhoon Chanthu (2021). The outer rainband, located about 100 km from the typhoon center, approached and made landfall on DT and brought a maximum rainfall of 57 mm h^-1 at 0600 UTC on 12 September. A significant precipitation enhancement of 25 mm h^-1 over DT was confirmed as compared to rain gauge observations over coastal and plain areas. The rainband was characterized by a broad area of strong reflectivities (> 40 dBZ) with a vertically extending feature, and its echo patterns were less organized with weak bright-band signatures. The upward motions (~1.5 m s^-1) and near-zero vertical motions existed at mid-levels and lower levels, respectively. The low-level horizontal winds impinging on the DT barrier intensified with time and could reach about 20 m s^-1, causing upslope lifting (~2.0 m s^-1) and a corresponding local reflectivity maximum over mountain slopes. Averaged polarimetric vertical profiles calculated over the DT area indicated a significant increase in Z_HH, Z_DR, and K_DP toward the ground below 3 km MSL, implying that the microphysical process contributing to the rainfall enhancement was primarily related to the collision-coalescence. These results were also supported by disdrometer observations, showing an increase in mass-weighted mean diameter and liquid water content during the passage of the rainband. In addition, as evident from the drop-size distribution (DSD) analysis, the observed heavy rainfall was mostly due to the presence of a numerous number concentration of midsize drops. This study demonstrated the important roles of orographic effects in modulating kinematic and microphysical characteristics and contributing to the heavy precipitation over DT.

A lidar data assimilation system was developed based on the Weather Research and Forecasting-Local Ensemble Transform Kalman Filter (WRF-LETKF) framework coupled with the Community Multiscale Air Quality (CMAQ) model. The objective was to investigate the impact of lidar data assimilation on PBL prediction and the subsequent influence on PM_2.5 prediction for a high-air-pollution event. The fine particulate matter (PM_2.5) profiles retrieved from two micropulse lidar observations were assimilated in the WRF-LETKF system. Three numerical experiments, BASE (with a nudging strategy), CTRL (with an ensemble framework), and LDA (with assimilation of lidar-retrieved PM_2.5 profiles), were conducted for a high-air-pollution episode. The BASE simulation overestimates the wind speed, leading to PM_2.5 underestimation. The CTRL and LDA simulations can improve the wind fields and enhance the PM_2.5 accumulation. With a strong error correlation between the lidar-retrieved PM_2.5 concentration and the wind fields, the LDA simulation effectively corrects the wind flow from the surface to the PBL top, further adjusting the PM_2.5 transport processes.

The Spectral Bin Model (SBM) is a 1-dimensional microphysics model that includes various thermodynamic processes such as melting, freezing, evaporation, and sublimation. The SBM determines precipitation type (PT) based on the simulated rainfall rate (RR), snowfall rate (SR), and input environmental profile. When the sounding observation is used as the input environmental profile, it has been shown that the PT produced by the SBM have relatively higher accuracy compared to different traditional prediction methods. The sounding, however, suffers from limited spatiotemporal resolution. The question that arises is whether the high-resolution thermodynamic fields of the operational numerical weather prediction model yield satisfactory results in determining the spatial distribution of PT. Local Data Assimilation and Prediction System (LDAPS) has been developed by the Korean Meteorological Administration (KMA) and has meteorological variables (Temperature [T], Relative humidity [RH], Pressure [P]) with a 1.5 km horizontal resolution and 70-71 layers. Prior to the joint utilization of SBM and LDAPS, we need to understand the impacts of errors in LDAPS on the performance of the SBM. The T, RH, and wet-bulb T (T_w) profiles of LDAPS will be compared with those of the sounding. Additionally, we will compare the accuracy of PT produced by the SBM when the input is either sounding or LDAPS.

Acknowledgment
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646). This work was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI2022-00310.

Different rainfall systems such as stratiform and convection evolve in different developing mechanisms. The characteristics study of precipitation types are important in order to understand their differences in terms of physical, dynamic, thermodynamical and micro-physical processes. Therefore, we investigated the whole properties of various rainfall systems after identifying precipitation types using newly developed classification algorithm. Categorization of precipitation types is performed using the combined algorithms composed of 1) the SL3D (storm labeling in three dimensions) and 2) classification algorithm based on feature parameters. The feature parameters are derived from VIL (vertical integrated liquid water contents) and mean reflectivity. The precipitation types are divided into 8 categories (precipitation stratiform, non-precipitation stratiform, convection, updraft, deep system, shallow system, low cloud and anvil). The vertical structures of different precipitation types are explored to compare the physical (dynamic) peculiarities using dual-polarization radar observation (WISSDOM attributes). In the addition, the generalized microphysical parameters (N₀’, D_m) and the latent heat (LH) acquired directly from GPM product are also investigate to analyze the microphysical and thermodynamical characteristics respectively. The N₀’ and D_m are obtained using polynomial regressions consisting of Z_DR and reflectivity. The bright band appeared at 4.5~5.5 km altitude in stratiform and deep system as shown in the vertical profiles of reflectivity and Z_DR. The convergence (divergence) patterns, one of WISSDOM attributes, are distinctive in lower (higher) atmospheric layers for convection and updraft regions. Both D_m and log (N₀’) of updraft are relatively higher than those of stratiform. Similar to several previous studies, the mean LH profile of convection is higher than one of stratiform at all heights. 

Acknowledgment: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).

The Walker circulation is the most prominent tropical zonal circulation on a planetary scale, which is influenced by sea Surface temperature fluctuations at a time scale longer than the intraseasonal time scale. Previous studies have theoretically suggested that equatorial waves accelerate or decelerate the large-scale zonal flow of the upper troposphere in the tropics. Therefore, this study investigates the relationship between the Walker circulation and equatorial waves to clarify which equatorial waves are dominant depending on the strength of the Walker circulation. By sorting the strength of the downward branch of the Walker Circulation above the western Indian Ocean, we make composite maps of outgoing longwave radiation. The results imply that prominent equatorial waves vary depending upon the seasonality. In boreal winter when the Walker circulation is strong, some signals presumably associated with mixed Rossby-gravity waves are hinted, whereas in boreal spring when the Walker circulation is weak, Kelvin wave-like features are exhibited. Interannual variability of the aforementioned relationship will also be discussed.

In this study, we contrast the distinct energy recharge-discharge cycle between propagating and eastward-decaying Madden-Julian Oscillation (MJO) events during the boreal winter season (November – April) from 1979 to 2018 using the moist static energy (MSE) budget and gross moist stability (GMS) plane analyses. A total of 41 MJO events are selected during the 40-year period, including 16 strong propagating (SP), 13 weak propagating (WP) and 12 eastward-decaying (ED) types of MJOs. The column-integrated MSE budget shows that, depending on the phases, vertical and horizontal advection terms take turns leading the role in discharging or recharging the MSE. In the phase when the net flux heating reaches the maximum, vertical advection plays the lead role in discharging the MSE. In the subsequent phases when the net flux heating gradually weakens, horizontal advection becomes the dominant factor in discharging the MSE. A similar situation occurs during the recharge phases to balance the net flux cooling. The SP MJO exhibits a stronger energy recharge-discharge cycle compared to the WP and ED MJOs, and the contrast significantly enlarges from the Indian Ocean to the Maritime Continent. The GMS plane analysis further reveals that four different types of convection attribute to the recharge-discharge process. During the wet phases, the top-heavy ascending motion and negative shallow convection jointly stabilize the atmosphere by exporting the MSE; while during the dry phases, the top-heavy descending motion and the positive shallow convection together destabilize the atmosphere through importing the MSE. Moreover, the recharge (discharge) phase leads the amplifying (decaying) phase by a quarter cycle is essential to maintain a robust energy recharge-discharge cycle to drive the eastward propagating MJOs.

Numerical weather prediction (NWP) models are running at finer grid spacings from 1 to 10 km with increasing computational resources. However, as the grid spacing becomes comparable to the length scales of convection, known as the grey zone, the turbulent eddies in the atmospheric boundary layer are partially resolved and parameterized. Whether a convection parameterization (CP) should be used in the grey zone remains controversial. Scale-aware CP schemes are designed to improve the representation of convective transport within the grey zone. For example, the multi-scale Kain-Fritsch (MSKF) scheme introduces modifications, such as tuning convective adjustment time scale and adding linear mixing of updraft vertical velocity with grid-scale vertical velocity so that it can be implemented successfully at grid resolution as high as 2 km. However, the scale awareness is empirical and, to some extent, remains tunable. In recent years, more and more machine learning (ML) models have been applied to different domains of atmospheric sciences, including using ML models as substitutions for physical parameterizations. In this work, we proposed to use bidirectional long short-term memory (Bi-LSTM) models to replace the scale-aware MSKF CP scheme. The Weather Research and Forecast (WRF) model is used to generate training and testing data over South China at a horizontal resolution of 4 km. Torrential rainfall often occurs along the coast of South China, with the background of a frontal rain band over northern South China or the Yangtze River basin. It is called warm sector heavy rainfall and is usually associated with convective, localized, torrential, and long-lasting; thus is challenging to predict. In addition, the Bi-LSTM based CP scheme is coupled with the WRF model. The preliminary results demonstrate that the Bi-LSTM model is able to achieve good accuracy, showing the potential substitution of the MSKF scheme by ML models in grey zone.

Monsoons are often considered as ITCZ migrations into the land. Monsoon theories also consider the main control on precipitation properties to be the moist static energy near the surface. This study attempts to further clarify the role of surface thermodynamic gradients in determining precipitation outcome.
We run a series of idealised ITCZ simulations at 30-km grid spacing with the WRF model, in an aquapatch domain from 63S to 63N. The model is forced by a meridional contrast of surface temperature, with comprehensive physics, rotation, and symmetric boundary conditions at the North and South boundaries.
Summer solstice equilibrium simulations capture the main general circulation features and the main spatial organisation of convection. Turning on or off the convective parameterization does not impact the mean precipitation latitudinal distribution, but it does impact the eddies at smaller spatiotemporal scales, which shows a non-negligible but limited influence of convection on the summer ITCZ properties. By varying the forcing, we show that monsoon intensity follows a highly non-linear (logarithmic) relationship with surface temperature gradient, and that a finite temperature contrast is necessary to generate any cross-equatorial flow.
Simulations with an additional seasonal cycle forcing capture the abrupt and delayed monsoon onset. We also compare seasonal cycle simulations over ocean and over an idealised land to tell apart the influence of land in the delayed ITCZ migration.

Agricultural activities and processes emit greenhouse gases like CO₂, CH₄ and N₂O, as well as NH₃. While greenhouse gas emissions affect the global climate, ammonia can be a toxic pollutant to human health and biosphere environments. To reduce the impact of livestock production and fertilizer on the environment, it is vital to quantify greenhouse gases as well as ammonia emissions. However, accurate monitoring of NH₃ emissions from barns, livestock, fertilization, and industry can be challenging at low levels (ppb to ppm) due to the high reactivity of ammonia and its tendency to adsorb to surfaces. Here, we present the new Picarro G2509 Cavity Ring-Down Spectroscopy (CRDS) analyzer that is optimized for continuous measurement of ammonia, while also reporting concentrations of greenhouse gases CO₂, CH₄ and N₂O. The measurement response time of NH₃ is sped up with low-reactivity internal materials and a high flow rate. The CH₄ concentration range of the analyzer has been extended to 800 ppm to account for large variations in barns. The G2509 analyzer adapts the excellence of the G2508 that is commonly used to study greenhouse gas fluxes from soils by implementing key design of the G2103 analyzer which is dedicated to low level ammonia monitoring at air quality stations. Researchers studied ammonia production from livestock in cattle barns using G2103, and G2509 has been proved to be optimized for ammonia abatement measurement at dairy barns. Methane and nitrous oxide emissions have great influence on crops production and impacts of temperature and moisture on carbon and nitrogen cycling has been investigated by G2508. The effect of nitrogen and carbon modification on soil and plants has also been widely studied among Asia regions. Therefore, it is worthwhile to understand the emissions, transport, fate, and impact of greenhouse gas and ammonia pollutants via continuous gas monitoring techniques.

Nitrous acid (HONO) is a key precursor of the highly reactive hydroxyl radical (OH). OH is a major oxidant to remove many gases emitted into the atmosphere and produces secondary air pollutants. Previous studies have revealed that the HONO emitted from soil is an important source of atmospheric HONO. As an agricultural country with a huge amount of fertilizer application, the effects of N fertilizers on HONO emissions have not been studied. In addition to HONO emissions introduced by fertilizers, emissions from unfertilized soils also accounted for a large part of total soil emissions due to their long duration. However, there has been limited research to quantify the HONO emissions for different soil types, thus lacking corresponding parameterization schemes to assess their impact on air quality. Here, we measured background (unfertilized) HONO and NO emission fluxes of soil samples collected from diverse soil types across China, and observed post-fertilization soil HONO emissions after applying three common fertilizers. The results show much higher emissions of HONO than those of NO, especially for samples from northern China. In terms of the fertilization effects, high HONO emissions from soils at 75%–95% WHC were founded, which contrasts with previous lower predictions at high soil moisture. Based on the results of laboratory experiments, the parameterization schemes for HONO emissions from fertilized and unfertilized soils were developed and implemented in a chemistry transport model. The inclusion of the soil HONO emissions improved simulations of atmospheric HONO and revealed a significant impact of this HONO source on air quality.

Wet deposition of reactive nitrogen in the form of NH₄⁺ and NO₃^- has become a major concern for scientists. The excess amount of such nitrogen species poses a threat to the environment and human health. NH₃ and NO₃^- which primarily represent a major fraction of Total Nitrogen (TN) are contributed from the emissions from transport, agricultural and industrial sectors. N containing fertilisers, N fixation and livestock are the major sources of NH₃ in the atmosphere while vehicular emissions and biomass burning contribute NO₂ and further NO₃^- in the atmosphere. This study reports the measurements of Total Nitrogen (TN), Dissolved Organic Carbon (DOC) and Inorganic Carbon (IC) in the wet deposition at a sub-urban site (Saharsa) and a rural site (Purnia) during July 2018 - September 2019 in the Mithila region of highly enriched alluvial plains of Bihar. Total 63 samples were collected from Saharsa and 29 samples were collected from Purnia using a polypropylene bottle-fennel assembly placed at a height of 20m from the surface of ground. The concentration of TN was determined by using Shimadzu TOC-TN analyser (Shimadzu model-TOC-LCPH E200 ROHS) The volume weighted mean of TN was comparatively high at the sub-urban site (1.72 mg/L) than the rural site (1.32mg/L) suggesting a greater contribution of NOx and NH₃ at the sub-urban site. The annual wet deposition fluxes of TN were calculated as 13.45 kg N ha⁻¹ yr⁻¹ at the sub-urban site and 5.73 kg N ha⁻¹ yr⁻¹ at the rural site. The airmass trajectory analysis carried out using NOAA data suggested that part from local sources, both the sites also had an influence of trans-boundary and long-range transport of air pollution. Keywords: Reactive Nitrogen, Total N, Wet deposition flux, Airmass trajectory.

This study examined the cause of a record torrential rain event over the western coast of Sumatra Island in March 2016. The influence of atmospheric equatorial waves (EWs) and the characteristics of the EWs were investigated. Analysis of the Japanese 55-year Reanalysis data (JRA-55) and precipitation data from the Global Precipitation Measurement (GPM) satellite showed that the event was caused by the combined effects of Kelvin waves, equatorial Rossby waves, and westward inertio-gravity (WIG) waves. An examination of the characteristics of the EWs revealed that the Kelvin waves had longitudinal scales of ~6,000 km, with a period of ~6 days and phase speed of ~12 m s^-1, which was typical of the convectively coupled Kelvin waves in this region. The WIG waves had a scale of ~2,500 km, with a period of 2.5 days and a relatively fast phase speed of 12~13 m s^-1. Heavy precipitation occurred when an eastward Kelvin wave from the Indian Ocean encountered a westward inertio-gravity (WIG) over Sumatra Island. It was concluded that along with the Kelvin and equatorial Rossby waves, the WIG waves might have played a major role in the formation of the extreme precipitation event.

Proxy-observational studies, and a sole model study, suggest that the Indian Ocean Dipole (IOD), an important global climate driver, exhibited multi-scale temporal variability during the Last Millennium (LM; CE 0851-1849, with relatively high number of strong positive IOD events during the Little Ice Age (LIA; CE 1550-1749), and strong negative IOD events during the Medieval Warm Period (MWP; CE 1000-1199). Using nine model simulations from the PMIP3, we study the IOD variability during the Last Millennium after due validation of the simulated current day (CE 1850-2005) IOD variability. Majority of the models simulate relatively higher number of positive IOD events during the MWP, and negative IOD events in the LIA, commensurate with simulated background conditions. However, higher number of strong positive IOD events are simulated relative to the negative IODs during the LIA, in agreement with proxy-observations, apparently owing to increased coupled feedback during positive IODs.

Using a variety of CloudSat/CALIPSO products, this study synergistically examines the performance of clouds and their radiative effects (CRE) for models participating in CMIP6. Results show virtually all models overestimate the net cooling effect of clouds, which is caused by the overestimation of shortwave CRE and the underestimation of longwave CRE. By dividing clouds into regimes jointly sorted by cloud water path and cloud cover, we found models commonly underestimate the relative frequency of occurrence (RFO) for clouds that are geometrically thick, and the bias of RFO is dominant over that of within-regime CRE in an error decomposition of total CRE. This results in underestimations of CRE in geometrically thick clouds, which are partially offset by overestimations in the remaining cloud regimes, leading to the globally averaged CRE being less biased. The consideration of regime-based CRE gives important information that could be used for correction of cloud parameterization in models.

This study aims to explore aerosol-cloud interaction over eastern China (EC) and its adjacent ocean (ECO) in boreal winter by coupling of a spectral-bin cloud microphysics (SBM) and an online aerosol module (MOSAIC) in WRF-Chem, with the support of four-dimensional data assimilation. The evaluation shows that assimilation has an overall positive impact on the simulation, and the coupling system reproduces the satellite-retrieved cloud parameters while exhibiting significantly improved simulation ability compared to the original SBM scheme as well as the bulk microphysical and MOSAIC coupling system. Differences in aerosol composition and physical processes lead to clear discrepancies in the aerosol-cloud interactions of EC and ECO during the simulation period. In EC with the gradual increase of aerosol number concentration (N_aero), cloud droplet number concentration (N_d) first increases then decreases and fluctuates around 800 cm^-3, while N_d in ECO increases faster initially, but soon its activation is suppressed by aerosol hygroscopicity and high activation threshold of numerous small particles, and almost no additional cloud droplets are produced. In terms of rapid adjustments, more bursty atmospheric supersaturation and lack of subsequent water cause cloud liquid water content (CLWC) in EC to increase explosively with N_d when there are few cloud droplets, but only maintains a low increase rate with further increasing N_d. ECO exhibits a fast increase in CLWC with N_d at high proportion of naturally emitted large aerosol particles, but its CLWC increase gradually stagnates as N_d increases. For non-precipitating clouds with less water content, CLWC in EC increases slowly with N_d, but can maintain a stable trend. While ECO, which relies mainly on large scale water and temperature variations to reach supersaturation, the increase in N_d leads to a decrease in CLWC.

We use the WRF-Chem model to simulate the extreme precipitation event in Guangzhou on May 7th, 2017, and investigate the effect of aerosols during this event. It is found that under polluted conditions, more latent heat is released through the condensational growth of droplets, which increases the updraft speed and ice-phase hydrometers. Precipitation is increased due to the melting of snow and graupel. However, in the latter stages of MCS development, the instantaneous extreme rain rate is lower in the polluted experiment, while the accumulated precipitation is higher. This study explains the evolution of extreme precipitation in the MCS and how the different microphysics, dynamics, and thermodynamics processes contribute to the evolution.

Since the late 1990s, NASA’s Earth Observing System constellation of satellites has provided continuous, long-term observations of atmospheric and surface processes on Earth, which has advanced our knowledge of our planet’s radiation budget. Imagery-based surface properties such as surface albedo and Bidirectional Reflectance Distribution Function (BRDF) are generated as an operational product using cloud-cleared, multi-angle surface reflectances from multiple overpasses over several days. The BRDF is central to imagery-based cloud and aerosol retrievals, while the surface albedo is a fundamental Earth energy budget parameter. Yet, this product is currently unavailable at higher latitudes such as the Arctic where (1) the low contrast between clouds and snow/ice poses a challenge for cloud detection and clearing, and (2) ice floes are drifting over time, which is currently not accounted for in the operational products. As a result, there is a significant gap in our understanding of the Arctic radiation budget and the influence of melt events on the ice-albedo feedback, cloud-radiative effects, etc. To address this gap, we propose the development of a BRDF/albedo product for moving sea ice floes and snow called the Sea Ice Floe and Snow Albedo Tracker (SIF-SAT). By leveraging multi-overpass, multi-angular satellite data, SIF-SAT will retrieve BRDF and albedo under low contrast and moving surface conditions. We combine existing cloud masks with machine learning (ML) models to produce cloud-cleared scenes in the Arctic. These scenes are then fed to a segmentation algorithm to identify individual moving sea ice floes and their reflectances are tracked over time to obtain BRDF and albedo. The focus of this presentation will be on the cloud-clearing model which has implications for radiation science in polar regions. SIF-SAT will enhance our capabilities in the challenging conditions of the Arctic and enable more accurate estimates of the cloud-radiative effect and ice-albedo feedback.

Precipitation is an important process that modulates the water cycle and also contributes to controlling the concentration of atmospheric pollutants by washing out particulate matters (PMs) (i.e., wet removal or below-cloud scavenging). Previous studies showed the spatial and temporal characteristics of the concentration of PMs or precipitation. Although several studies attempted to quantify the impact of the wet-removal mechanism on the concentration of surface PMs, they were limited to a few events-based case studies. Thus, this study determined the impact of the wet-removal mechanism on the concentration of surface PMs (i.e., PM2.5 and PM10) by many events (over 10 thousand) based on long-term measurements in Seoul, Korea. Twenty-one years (2001-2021) surface measurements of precipitation and PM10 made at 25 stations in Seoul were used, while PM2.5 measurements were only available for seven years (2015-2021). All measurements were made with a one-hour resolution. The statistical analyses (i.e., monthly, seasonal, and yearly averages) for the variability of precipitation and mass concentrations of PMs were performed. The precipitation events were identified when the threshold time interval (i.e., 2 hours) existed between hourly precipitation data. The characteristics of precipitation events were quantified based on the precipitation duration, precipitation intensity, and accumulated amount of precipitation. The variation in PM concentrations during the precipitation events and the rate of wet removal were calculated as functions of the characteristics of precipitation events. The optimal duration, intensity, and accumulated amount of precipitation for the maximum wet removal were determined and the influence of precipitation properties on the wet-removal mechanism was also evaluated.

Aerosols are one of the crucial factors that impact regional and global climate. Dust and anthropogenic aerosol from East Asia contribute to the change of radiative fluxes and climate by modifying the cloud regimes in Northwestern Pacific (NWP). In this study, we attempt to find the role of aerosols on cloud properties in various cloud regimes under the control of differential weather systems in NWP. Using the cloud products obtained from Himawari-8 satellite combined with the Moderate Resolution Imaging Spectroradiometer (MODIS) retrievals for comparison, we classified the cloud regimes in NWP using the "k-means" clustering algorithm. The characters of the cloud regimes are described by analyzing their spatio-temporal distributions of the occurrences. Using the principal component analysis in T mode (T-PCA) for classifying weather patterns, we discussed the association between cloud formation and weather systems. We assessed the sensitivity of cloud properties to the perturbation of aerosols of different types in the main cloud regimes under the dominant weather patterns. This study helps to improve our understanding of the aerosol-cloud-radiation interactions in NWP.

The typical areas of aerosol distribution in China include three mega-city regions (BTH, YRD, PRD) and the background area of the Qinghai-Tibet Plateau. To compare the distribution and source of aerosol particles in these regions which have large differences in natural conditions and population density. In this research, we observed the aerosol particle number concentration continuously from July 2017 to September 2020. The conclusions can be deduced: 1. Aerosol-sounding observations conducted on the Qinghai-Tibet Plateau from June to September 2020 found that there is a high concentration level ( > 165 #/cm³) of fine particle layer (N < 300nm) near 15km in the upper air. 2. Due to the Northwest airflow in the upper air on the Qinghai-Tibet Plateau in September, the concentration is the lowest value of the four times observations. 3. The Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model is used to analyze the 96h back trajectories of particles on the Qinghai-Tibet Plateau. It is found that in the late summer and early autumn, there are 29.3%~57.3% of the aerosols were transported from the southwest of the observation site. 4. According to the analysis of aerosol trajectories and particle size distribution characteristics, the source of aerosol in megacities is mainly attributable to human activities. We first investigate the aerosol differences under the influence of natural atmospheric conditions and human activities. But there are some deficiencies in the current research: further research is needed on the vertical distribution of aerosols and their climatic effects in megacities and the remote background region.

Anthropogenic nitrogen oxides may influence the cloud condensation nuclei (CCN) activity of biogenic secondary organic aerosols (SOA) in both daytime photooxidation and nighttime NO₃ oxidation, which has significant implications for the climatic impact of SOA. However, the influence remains unclear. In this study, we investigated the CCN activity of monoterpene-derived SOA formed from photo-oxidation of two monoterpenes (α-pinene and limonene) at different NO_x levels, as well as BSOA generated in the dark by reaction of limonene and NO₃. And we use the hygroscopicity parameter κ to characterize the CCN activity of MT-SOA. In daytime OH oxidation, NO_x had little influences on the CCN activity of MT-SOA. In nighttime NO₃ oxidation, MT-SOA had much lower CCN activity compared with those formed via OH or O₃ oxidation. And we report the κ of monoterpene-derived organic nitrates (0.029-0.052) for the first time to our knowledge, which may be used to improve model simulations of CCN concentrations.

This study researched the distribution characteristics of air pollution in Shuangliu District, Chengdu, the main persistent aerosol pollutants and ozone concentration in Shuangliu District from December 21, 2022 to February 5, 2023 were comprehensively analyzed by using the results of cruse observations. The sources of pollutants were analyzed by combining the meteorological data provided by Shuangliu Meteorological administration The Portable Optical Particle Profiler (POPS), ozone detector, hand-held weather station and wind lidar were integrated on the vehicle and made the cruse observation around the Baihe Park site, Yong'an Primary School site and their surrounding area in Shuangliu District. Three-dimensional spatial and temporal changes of areosol size distribution and ozone were obtained. The results showed that the average daily total particle number concentration was 5000~18000 #/m³/15000~18000#/m³ in haze weather, 5000~8000#/m³ after rain, respectively. coarse particles (272nm~473nm) accounted for 45%~60% in the whole observation period. Results showed that pollution was more likely to form the north wind, mainly due to industrial emissions and human activities in the upper wind direction region. The total particle number concentration decreased by 3%~8% when arriving at the park, but increased significantly, around 3%~10%, when passing by construction sites and driving close to trucks. Results of the cruse observation provide a scientific basis for particle and ozone control in Shuangliu District, which is conducive to the subsequent targeted pollution prevention and control of relevant departments, and can also provide reference for other urban areas.

The Greenland ice sheet (GrIS) has been losing mass at an accelerating rate in recent decades due to warming, and understanding the underlying mechanisms, such as the impacts of clouds, is essential. Using space-borne data, this study investigates the spatial distribution of ice clouds and liquid-bearing clouds (LBCs) over the GrIS and their surface radiative forcing effects during summer daytime from 2006 to 2017, along with their characteristics during the North Atlantic Oscillation (NAO) events. Due to the perennial high-albedo surface, both ice and LBCs have a less important shortwave radiative cooling effect than in other environments. Based on the spatial variation pattern of clouds with the NAO index, the GrIS can be divided into three regions: the western, central, and eastern GrIS. During the positive NAO, the westerly wind strengthens in the western region, which causes the fraction of both ice clouds and LBCs to increase, and the cloud radiative effect at the surface increases by 2.07 W/m²; the temperature decreases in the central region, the fraction of ice clouds increases, the fraction of LBCs decreases, and the net radiative forcing is −2.05 W/m²; and sinking airflow is generated in the eastern region, both ice cloud and LBCs decrease, and the net cloud radiative effect at the surface is −1.34 W/m². The spatial and temporal variations in clouds in different phases over the GrIS are closely related to the NAO, and the response of clouds to changes in the atmospheric circulation field during the NAO varies in different regions of the GrIS.

To explore the influence of dust aerosols on cloud formation via radiative effect, this study takes the Badain Jaran desert area as the research area where cloud formation is less disturbed by human activities compared to other densely populated areas such as the North China Plain. We first identify 469 cases with cloud formation after 03:00 UTC (11:00 Beijing time) by manual method (checking every cloud image to identify the temporal evolution of clouds) using the high temporal resolution Himawari-8 satellite data. Through the analysis of individual cases, we find that the cloud formation time is mainly dependent on the low tropospheric stability (LTS) when the total precipitable water (TPW) is less than about 18.5 kg/m², and greatly affected by the TPW when it is larger than 18.5 kg/m². Dust aerosols near the ground absorb shortwave solar radiation, making the atmospheric structure unstable (decreased LTS) and conducive to cloud formation. Statistically, for every 0.1 increase in AOD, the cloud formation time is advanced about 138 minutes. Further analysis shows that dust aerosols actually inhibit the formation of convective clouds through microphysical processes when the TPW is small, while promote the formation of convective clouds through radiative processes when the TPW is large. This study reveals the effects of dust aerosols on the time of cloud formation along with the underlying mechanisms, which provides observational support for improving cloud parameterization in numerical models.

The climate of Indonesia located under the equator is predominated by the monsoon of both the northern and southern hemispheres. Climate changes in Indonesia would affect agricultural production and the whole economy. The future climate changes were projected by GCM simulation, but the geographical features of Indonesia are very complicated. Therefore, various downscaling methods involving RCM simulations are necessary to find detailed changes in the monsoon climate over Indonesia. Our previous work showed that precipitation over Java Island generally decreased in the onset season and increased in the dry season deduced by the bias-corrected CMIP5-GCMs output. In this study, CMIP6-GCMs are added as boundary conditions, the time scale of the analysis is varied from daily, 10-day, and monthly scales, and a downscaling method is also applied to account for the monsoonal atmospheric field and its changes. Preliminary analysis results using two CMIP6 GCMs (MIROC6 and MRI-ESM2) indicated that in each model group (MIROC5:MIROC6 and MRI-CGCM3: MRI-ESM2-0) seasonal monsoonal precipitation shift and wind system of historical runs (1995-2014) is similar. That is the onset of the monsoon is relatively postponed and the total amount of precipitation is small in MRI-ESM2-0. Future precipitation changes in the onset season would also decrease in MRI-ESM2-0. Multiple bias correction and statistical-downscaling methods are also being applied to the CMIP6 dataset to further analyze the seasonality and model dependence of future changes in precipitation in Indonesia.

We investigate the projected climate change over Mainland Southeast Asia (also known as the Indochina Peninsula) and the Lancang-Mekong River basin, a region with complex topography and unique weather and climate systems, but limited availability of published high-resolution regional climate model (RCM) studies. The study is based on an unprecedented ensemble of 21st century projections with the RegCM4 RCM driven by five different general circulation models (GCMs) at a grid spacing of 25 km under the representative concentration pathways RCP4.5 and RCP8.5. We focus on mean temperature and precipitation in the dry season November–March (NDJFM), the wet season May–September (MJJAS), and the whole year. Intercomparison between the RegCM4 simulations with the driving GCMs is provided to illustrate the added value of the RegCM4 experiments. RegCM4 reproduces greater and more realistic spatial detail of the present day temperature and precipitation distribution compared to the driving GCMs, but some biases are found, such as an overestimation of precipitation over high topography regions. The spatial pattern of biases shows some consistencies across the GCMs and, for NDJFM the RegCM4, although weak correlation is found between the GCM and nested RegCM4 biases. A generally lower warming is projected in the future by the RegCM4 in different seasons and the whole year. For precipitation, while prevailing increases are found in the GCM projections, large areas of decrease occur in the RegCM ones, in particular during the wet season, possibly due to the more detailed topographical representation. The change patterns of precipitation show consistencies across the GCMs and the RegCM4, especially in MJJAS. The projected changes of extreme indices indicate a general decrease/increase of extreme cold/warm events. Drought events are projected to be more frequent over southwestern, while a general increase of heavy rain events prevails over most parts of the region.

Policies for carbon neutrality are being implemented around the world. To achieve carbon neutrality, renewable energy production must be increased. Solar energy is a representative renewable energy and is known to be significantly affected by climate change. In this study, based on high-resolution climate change scenarios, future Photovoltaics potential (PVpot) changes in South Korea was predicted. According to the four SSP scenarios (SSP1-2.6, SSP2-4.5, SSP3-7.0, SSP5-8.5), at the late 21st century, surface down-welling shortwave radiation, which has a major impact on PVpot, is expected to increase (by about +0.8 ~ +3.6 W/m2) compared to the present. However, PVpot is expected to increase (by about +1.3%) in the low-emissions scenario (SSP1-2.6) and decrease (by about -0.7 ~ -2.0%) in the high-emissions scenarios (SSP3-7.0, SSP5-8.5) at the end of the 21st century compared to the present. Because as the temperature increases significantly, the temperature of the solar cell panels increase, and performance ratio of PVpot decreases. Seasonally, PVpot is expected to increase (by about +2.3 ~ +4.0%) in summer and decrease (by about -3.6 ~ -7.1%) in winter.

Reanalysis metadata are a critical resource for examining meteorological changes and evaluating wind energy potential. Observed wind speed studies reveal a global decline in land surface wind speed since the 1960s, known as terrestrial stilling, but with a reversal around 2010. The extent to which the decline in trend and turning point of land surface wind speed has been captured by reanalysis products is currently unknown. To address the research gap, a systematic evaluation of climatological winds and trends in three reanalysis products (ERA5, JRA-55, and NCEP/NCAR) was conducted by comparing 10-m wind speed time series from grid cells with observational data from 32 in situ meteorological stations for the period 1973-2005. To measure the accuracy of the reanalysis products, a range of statistical metrics were employed, including standard deviation (STD), mean absolute error (MAE), percent bias (PBIAS), correlation coefficient (R), M-K test and Sens slope estimation. Based on the evaluation of climatological winds, JRA55 proves to be the most accurate reanalysis product compared to the observations. Despite this, considerable deviations were discovered between the simulated wind speeds and the observations. The JRA55 reanalysis product was found to reflect changes observed in the Subcontinent, with the NCEP/NCAR product following closely behind. The JRA55 reanalysis demonstrated agreement with observations m/s, with a high R of 0.93, a PBIAS of -10.70, and a low MAE of 0.28. further, NCEP/NCAR showed lower standard deviation levels of 0.54. The Subcontinent's wind energy industry is experiencing growth and expansion, with opportunities for the development of offshore wind farms and the expansion of existing land wind farms. This study aims to highlight the discrepancy between observed and simulated land surface wind speeds, emphasizing the importance of careful use of reanalysis products for wind assessment and forecasting in the subcontinent region.

In dynamical downscaling, biases in lateral boundary conditions obtained from coarser models can play an important role in the dynamically downscaled simulations. As a part of Singapore’s Third National Climate Change Study, dynamical downscaling has been carried out over Southeast Asia (SEA; 79E-160E;16S-24N), using the Singapore Regional Climate Model (SINGV-RCM; an adapted version of Singapore’s operational NWP model) using perfect boundary conditions (ERA-5) and four CMIP6 GCM (ACCESS, EC-Earth, MPI & UKESM) at 8km horizontal resolution. The SINGV-RCM is forced with ERA-5 reanalyses data for a 36-year period (1979-2014) at 8km resolution over SEA with regular update of the sea surface temperature at 6-hr interval; further, a similar simulation is carried out using four CMIP6 GCM (ACCESS, EC-Earth, MPI & UKESM) forcing data for a 60-year period (1955-2014). The last 20-year period (1995-2014) from the five simulations with reliable high-resolution observation (merged in-situ and satellite) is selected for evaluation. Rainfall characteristics including the diurnal cycle and extremes from the five simulations evaluated against observations will be presented.

Climate change due to global warming affects various areas such as agriculture, ecosystem, water resources, and forest in different ways. Climate change scenarios have been used as a scientific basis for establishing climate change adaptation policies to preemptively minimize damage by assessing future impacts due to climate change. In this study, downscaled data with 1 km resolution for the Korean Peninsula was produced using the SSP (Shared Socioeconomic Pathways) scenario used in the AR6 report. In order to consider the uncertainty of the future outlook that may occur depending on the selection of a specific Global Climate Model (GCM) when using only a single climate model, a total of 18 climate model data were used to analyze the future projections based on multiple model ensemble (MME). First, the downscaled data was produced using a non-parametric quartile mapping (QM) technique that does not use a distribution formula in estimating the bias between the observed data and the GCM data. Then, the characteristics of future fluctuations of 19 bioclimatic indicators were analyzed. The future projections were analyzed for each of the 30-year periods of the near-future (2011-2040), mid-future (2041-2070), and far-future (2071-2100) compared to the past 30-year period (1981-2010). In the case of annual precipitation and annual average temperature, the rate of change tended to be higher during the far-future compared to the near-future period. Within the same future period, SSP5-8.5 showed the highest change rate while SSP1-2.6 showed the lowest change rate. In the case of bioclimatic indicators, bio4 and bio15, which represent seasonal variability in precipitation and temperature, also showed the highest increase rates in the far-future and SSP5-8.5 scenarios.

Extreme precipitation events, causing tremendous socio-economic damage, are among the most disastrous phenomena. Nowadays, many studies address the localization of extreme events and their intensification throughout climate change; thus, the demand for high-resolution and reliable atmospheric data is increasing. Downscaling of model output is one of the efficient ways to produce detailed data, and it simultaneously draws attempts to increase its efficiency through deep learning (DL) methods. Although DL methods have the advantages of less demand for long-term observation data and lower calculation costs, there are limitations, too: in particular, the performance of downscaling through DL methods heavily depends on the amount of the training data set. The enormous amount of atmospheric data is suitable for training DL-based downscaling models; however, high-quality atmospheric data with appropriate resolution and reliability are restricted. Considering various features of regional climate, investigation of the domain-shifting problem is crucial in order to apply the learned DL model to other regions excluded from the training data. Accordingly, the performance differences among the DL-based downscaling models, trained with differently distributed data, are investigated by comparing the cross-applying results. In addition, the DL model with integrated learning data is also investigated. This study lays the foundation for the smooth application of a deep-learning model for detailed daily precipitation.

Episodic cold surges associated with the East Asia (EA) winter monsoon can penetrate deep into the South China Sea (SCS), enhance consequent tropical rainfall, and further strengthen the EA meridional overturning circulation (MOC). These cold surges can promote strong surface fluxes and lead to a deeper marine planetary boundary layer (PBL). In this study, we use high resolution radiosonde data of temperature and humidity profiles over Dongsha Island to identify the PBL height (PBLH), mixed layer height (MLH), cloud base, and cloud top for the period of December-January-February (DJF) from 2010 to 2020. We combined ERA-5 meteorological variables and surface fluxes, MERRA-2 cloud radiation data, and radiosonde-derived PBL parameters to perform an energy budget analysis and turbulent diagnostics from the mixed-layer model from Nicholls (1984). Here we show a strong turbulent flux convergence of both heat and moisture over the SCS during cold surges, which leads to a lifting of the MLH to ~1.0 km and PBLH to ~2.0 km and associated cloud development over Dongsha Island (116.69E, 20.70N). The cold and dry horizontal advection is balanced by this vertical turbulent flux convergence in the energy budget. Overall, at post-surge the PBL is stable but mixed layer is unstable, which contrasts with the pre-surge stage that features a stable mixed layer and a conditionally unstable PBL.

From December 13, 2017 to January 2, 2018, three tropical cyclones (TCs) successively made landfall in Mindanao and led to the most extreme wintertime (December - February) subseasonal peak rainfall event (SPRE) over the eastern Philippines (9°-14°N, 122°-127°E). The SPRE is defined by the maximum 15-day accumulated rainfall amount within the region during a time span of 90 days. The 15-day rainfall extremes over the eastern Philippines are associated with the environmental low-level cyclonic vorticity, which can be contributed to La Niña condition, Madden-Julian Oscillation (MJO), and equatorial Rossby (ER) waves. During the 15-day period of 2017/18 winter SPRE, the cyclonic vorticity phase of the westward propagation ER waves from the North Pacific occurred twice and the second episode occurred simultaneously with an MJO from the Indian Ocean across the Maritime Continent to the western Pacific. Favored by anomalous cyclonic vorticity and humidity produced by La Niña and MJO, the two ER waves enhanced and resulted in two TCs and led to the extreme SPRE. Based on the hindcast data in the S2S database, the 2017/18 SPRE and associated MJO and ER waves in ECMWF and NCEP models show skillful forecasts up to the extended range (11-day lead). For the SPRE in 2017/18 winter, the ECMWF model performs better than the model’s hindcast skill, whereas the NCEP model performs worse. The model with a better representation of SPRE timing and intensity has smaller phase and amplitude biases of tropical waves in the medium (5-10-day lead) and extended ranges. The forecast skill of the MJO and ER linked to ENSO phases will be discussed. The analysis procedure proposed in this study can be applied to monitor and predict the SPREs and their associated large-scale drivers in other regions.

In order to improve the precipitation forecast of the next-generation Global Prediction System with the Finite-Volume Cubed-Sphere Dynamical Core in Taiwan’s Central Weather Bureau (TFV3), this study modified the convective processes in New Simplified Arakawa-Schubert scheme (NSAS) based on the methodology of scale-aware parameterization developed in Kwon and Hong (2017) and investigated its impacts on a front event, which propagated across Taiwan and produced heavy rainfall in late May of 2020. Results show that the modified scale-aware parameterization has significantly improved the intensity and the spatial distribution of frontal precipitation forecasts due to the proper definition of convective updraft fraction. However, the synoptic-scale features perform a larger warm bias with the modified scale-aware parameterization. To understand the reason for the warm bias, the vertical profiles of zonal mean temperature, cloud cover and the heating rate from each physical parameterization were investigated to reduce the biases. The result of verification shows that the occurrence of a larger warm bias in low-to-mid levels is attributed to the decrease in subgrid-scale water vapor consumption by the deep convection process and the increase in grid-scale condensation heating resulting from the consumption of environmental instability by microphysical processes. Therefore, further modification of the scale-aware capability of convective cloud water detrainment is proposed to reduce the heating from microphysical processes and result in a better overall performance for the medium-range weather forecasts.

In the boreal warm season (May-September), tropical cyclone genesis (TCG) is active in the West North Pacific (WNP) region. TCG occurs in WNP monsoon trough that are influenced by active weather and climate oscillations. In addition, TCG tends to cluster in a short period and absent at other times. In this study, we investigate the weather (or extended weather) disturbances responsible for the extreme TCG clustering and how the weather disturbances are influenced by low-frequency climate oscillations including QBWO, ISO, and ENSO. We first calculate the maximum number of TCG per 10 days in each JJASO season from 1990 to 2019 to determine the most extreme TCG cases. Four cases with 5 TCGs in 10-days are selected. By analyzing the extended weather disturbances of the four periods, we find that 15 out of 20 TCs occurred in positive vorticity phase of QBWO and 4 TCs formed following a preexisting mature TC suggesting Rossby wave energy dispersion within positive vorticity region of QBWO. A further analysis of the climate conditions shows that, three of the four extreme TCG-clustering events occurred in late July-August during La Niña developing summer with inactive BSISO, while the other occurred in El Niño developing July with active BSISO in WNP.

The South Asian High (SAH) is the most important upper tropospheric summer system over Asia as its intensity and shape variation are closely linked with Indian and East Asian summer monsoon rainfall variability. However, SAH intraseasonal variability and its impacts on precipitation has not been fully untangled. In this study, we used the empirical orthogonal function (EOF) to analyze the 200-hPa daily geopotential height (GPH) during June-August in 1979-2020 over 20°-120°E, 10°-50°N. The first three leading modes are all significant at the intraseasonal time-scale. The regression patterns of 200hPa wind and GPH in the Northern Hemisphere show the first two modes are associated with different types of midlatitude circumglobal teleconnection patterns. The spatial pattern of EOF3 shows a negative sign over the Black Sea, a positive sign over the North China and another negative sign over the north of Indian and Indochina Peninsula, suggesting a possible north-south shift of the SAH center location. The regressed precipitation on PC3 shows negative correlation with the precipitation over the Yangtze River and Japan and positive correlation with the precipitation over the South China Sea (SCS) and Philippine. The PV tendency equation is used to diagnose how the SAH variations influence the precipitation anomalies over the SCS. The extreme cases in June and July 1998 are selected for detailed analysis. In June and July 1998, the SAH was extended eastward to the South China, and the Western Pacific subtropical high (WPSH) was extended westward to the SCS. Due to the anomalous descending motions associated with the WPSH, the rainfall over the SCS (110°–120°E, 10°–20°N) was the fewest during 1979-2020. The process of how the upper tropospheric circulation affects the lower tropospheric circulation and precipitation will be discussed.

During the first half month of April 2022 the Philippines experienced severe disasters associated with the weak but deadly tropical storm Megi that caused 214 deaths and the sinking of two ships. This prompted us to investigate the predictability of the springtime Philippine sub-seasonal scale rainfall extremes and to identify the forecast opportunity in the S2S prediction database. The results suggest that ENSO is the most influential climate driver for the sub-seasonal scale springtime (February-April) rainfall variability in the Philippines. MJO and equatorial Rossby (ER) waves can modulate the sub-seasonal rainfall extremes. The 15-day accumulated rainfall amount of the sub-seasonal peak rainfall events (SPRE) is higher during a La Niña spring, so is the occurrence probability of a extremely wet event. The wet extremes can be enhanced by strong MJO or equatorial Rossby (ER). For the April 2022 case, the S2S multi-model ensemble mean (MMEM) can successfully predict the extremity of the SPRE at least 10 days ahead, although both MJO and ER were weak during the occurrence period of the deadly SPRE. Thus the successful prediction was interpreted as a result of the predictability rooted in the rainfall variability driven by the La Niña event. Since the S2S prediction is built on the ENSO foundation, it is proposed to include the SPRE in the real-time extended-range forecast items to exploit the benefits of S2S prediction and applications.

This study examines the sensibility of WRF simulation to different options of the Noah land surface model (LSM) with multiple parameterization options (Noah-MP) and the feasibility of using LSM perturbation for the ensemble forecast of afternoon thunderstorms over the Taiwan area. Five land surface processes in Noah-MP were tested, and the results showed that the sensitivity of the processes in descending order are: surface layer exchange coefficient, canopy radiation geometry, canopy stomatal resistance, surface resistance to evaporation, and soil moisture factor for stomatal resistance. Based on the sensitivity analysis, we designed an LSM perturbation method for WRF based ensemble system by applying various Noah-MP configurations and land surface initial conditions. Two ensemble experiments, each with 24 members, were conducted to assess the impact of LSM perturbation on afternoon thunderstorm rainfall simulations. One applied perturbation on initial atmospheric conditions and the other with additional LSM perturbations. The preliminary results from the 24-hour forecast of five afternoon thunderstorm cases showed that the LSM perturbation scheme could effectively improve the spread-skill performance of atmospheric variables near the surface, and the probability match of precipitation is also improved. Further details of the probability forecast analysis will be presented.

The key to ensemble forecasting is the generation of initial ensemble perturbations which is aimed at sampling the probability distribution of analysis errors. Most studies assess the prediction skill of the ensemble forecasts, which, however, is simply an indirect reflection of the quality of initial ensemble conditions. This study will evaluate the performance of ensemble perturbations in sampling forecast errors at multiple spatial scales through the entire forecast lead times. The projection of ensemble perturbations onto forecast errors at multi-scales and different lead times is the major metric used in this study. The variations of the projection as a function of spatial scale, lead time, and vertical levels will be carefully investigated. How these perturbation-error relationships can influence the control and perturbed forecast skill will also be explored.

Quantification of the uncertainties in initial analyses against the real atmosphere (“reality”) provides a fundamental reference for the evaluation and development of operational data assimilation (DA) systems. Due to the unknown reality, most existing methods for analysis error estimation use reanalysis datasets or observations as a proxy for reality, which are empirical, nonobjective, and biased. Unlike these methods, our study adopted a modified Statistical Analysis and Forecast Error (SAFE) estimation method to objectively and directly quantify spatiotemporal errors in analyses compared to reality based on unbiased assumptions. In the present study, the SAFE method was first applied to estimate the annual variation and spatial distribution of analysis errors in the Global Forecast System of Global/Regional Assimilation and PrEdiction System (GRAPES_GFS) at the China Meteorological Administration since the beginning of its operational implementation (i.e., 2016–2021). Qualitative comparison to analysis error estimations in previous studies showed that SAFE can provide more reasonable spatial-mean analysis error profiles than can the estimation with the ERA-5 reanalysis as a reference (the approach hereafter called “ERAv”). Moreover, ERAv overestimates (underestimates) the spatial-mean analysis error below (above) ~500 hPa compared to SAFE because it neglects the uncertainties inherent in reanalysis. Overall, the SAFE estimation reveals that relative reductions of about 12.5%, 29%, and 24.5% were achieved for the spatial-mean analysis errors of wind, temperature, and geopotential height, respectively, in the GRAPES_GFS throughout the six-year study period. These results can largely be attributed to the DA scheme being upgraded from 3D-Var to 4D-Var. SAFE can also provide more reasonable and accurate point-wise analysis errors than ERAv can.

The ensemble forecast system based on the Korea Integrated Model (KIM), which is developed for Korea’s own numerical weather prediction (NWP) model, has been in operation at Korean Meteorological Administration (KMA) since October 2021. KMA ensemble forecast system consists of 50 perturbation members (25 members for long-range forecast) and 1 control simulation. Four-dimensional LETKF (Local Ensemble Transform Kalman Filter) with additive and RTPS inflation scheme is used to make initial perturbation.
Evaluation of forecast scores shows that our operational ensemble forecast system is generally more skillful compared to the deterministic simulation as forecast time is longer. We also tested ensemble forecast with increased ensemble size which is planned to be operated this summer. Increased ensemble size produces better representation of atmospheric fields especially in higher latitudes. Details of results from operational ensemble system and impacts of increased ensemble size will be discussed with introducing a brief overview of our ensemble forecast system and development plan in future.

The Table of Aerosol Optics (TAO) project is a community repository of optics computations (extinction, absorption, single-scatter albedo, lidar ratio, etc) that are useful for global models and remote sensing applications. TAO expands upon historical efforts (e.g., Hess et al., 1998) by building an open database that uses recent measurements and new computational techniques for non-spherical particles. The ‘open’ aspect of TAO is important, since the size distributions, hygroscopicities, refractive indices, and morphological recommendations of today will undoubtedly yield to different values in the future; the open framework of TAO allows scientists to keep adding new computations to the database as the science evolves.
TAO is meant to be a community repository where specialists can put their computations for other scientists to use. So for instance, some groups are advancing new techniques that can accommodate complex fractal aggregates of black carbon, other groups are working on realistic irregular shapes for mineral dust, and different groups are updating the hygroscopicity of various aerosol types using new techniques. The TAO database gives these scientists a place to distribute their products. As TAO grows, modelers and remote sensing specialists will look to TAO as a place to find a wide variety of choices for testing. Meanwhile, global modelers can also use TAO to lobby for new tables that accommodate their needs.
Eventually, TAO will provide mass extinction coefficients, mass absorption coefficients, lidar ratios, etc., at popular remote sensing and global modeling wavelengths (0.25-40 µm) for all pertinent species (sulfate, sea salt, BC, OC, BrC, dust, etc.). Multiple tables may be created for each species or type to account for the multiple valid size distributions, hygroscopicities, complex refractive indices, and shapes that can be found in the literature.
Hess, M., P. Koepke, and I. Schult (1998), Bull. Am. Meteorol. Soc.

Since the airborne particulate matter (PM) affects the respiratory health of human significantly, the monitoring of its quantity in a high spatial resolution is requested. But it is not an easy task because the installation of qualified surface monitoring station is expensive. The satellite monitoring can be an alternative but its spatial resolution is still sparse. To get over these limitations, we can estimate of PM_2.5 (PM having a diameter < 2.5 μm) based on the machine-learning system with the consideration of both surface in-situ and satellite measurements. In this study, we develop the PM_2.5 estimation platform based on the random forest (RF) algorithm using the Geostationary Ocean Color Imager (GOCI) 6km-AOD products and 500m-AOD products with surface PM_2.5 measured by KT company cheap sensor networks from June 2018 to December 2019. Our target region is Busan in South Korea. In addition to AOD, we use other 11 physical values such as the meteorological parameters (e.g., temperature, wind speed, etc.) from the ERA5 reanalysis dataset of the European Centre for Medium-Range Weather Forecasts. For the validation, we use two methods, which are the 10-fold cross validation and the comparison with the surface PM_2.5 observed by the Korean government monitoring system. In conclusion, the quality of our RF-based PM estimation is reliable, showing R² = 0.45 for the case to use AOD at 6 x 6 km², and R² = 0.71 for the case to use AOD at 0.5 x 0.5 km², showing that the usage of high resolution AOD enables us to have better PM_2.5 estimation. Our study clearly shows the advantage of high AOD resolution for the regional air quality monitoring, related to the machine-learning system. This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. 2020R1C1C1011624).

Reanalysis data prepared by the aerosol data assimilation was developed to compensate for the limitations of ground observation data, satellite data, and modeling. Since The second version of Modern-Era Retrospective analysis for Research and Applications (MERRA-2) reanalysis data including AOD assimilation provides the gridded PM_2.5 for the long-term period, it can be widely used for the analysis of regional aerosol pollution. To do this work, the validation of MERRA-2 PM_2.5 is strongly required. In South Korea, however, this evaluation research is still insufficient. In this study, from 2015 to 2021, the PM_2.5 concentration of MERRA-2 is evaluated by surface PM_2.5 observed by the Korean national intensive monitoring sites (NIMSs). We found some difference between the MERRA-2 and NIMSs PM_2.5. First, the monthly variation is different. While NIMSs PM_2.5 is highest in winter but lowest in summer, MERRA-2 PM_2.5 shows a maximum in April and minimum in winter. In other words, MERRA-2 PM_2.5 is largely overestimated in summer but underestimated in winter generally. Among the components used in the MERRA-2 PM_2.5, sulfate, black carbon, dust, and sea salt is overestimated but organic carbon is underestimated compared to the aerosol component of NIMSs. However, total MERRA-2 PM_2.5 is much lower than the PM_2.5 of NIMSs, especially in the urban area. This feature seems due to the absence of nitrate component in the MERRA-2 data. When we consider the amount of nitrate based on the empirical estimation suggested by some previous studies, we confirm that the quantity of MERRA-2 PM_2.5 becomes much closer to the NIMSs PM_2.5. Understanding of MERRA-2 PM_2.5 characteristics for each region is still required in future studies. This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. 2020R1C1C1011624).

Particulate matter with aerodynamic diameters smaller than 2.5 microns (PM_2.5) is a major contributor to air pollution worldwide and negatively impacts human health. For the past few decades, researchers have developed a variety of approaches to estimate surface PM_2.5 from satellite observations, due to a lack of ground station in situ PM_2.5 coverage in many regions globally. A well-documented technique is the use of column-integrated aerosol optical depth (AOD) retrievals from passive sensors to obtain surface PM_2.5 proxies. Active sensors, like lidars, provide aerosol vertical distribution, and thus can be used to scale column AOD to the surface to improve the proxies. In addition to this synergistic passive/active sensor approach, near-surface aerosol extinction retrievals from lidar measurements can also be used independently to derive PM_2.5 concentrations. Furthermore, researchers have used aerosol models to convert satellite AOD to surface PM_2.5, and data assimilation methods can be employed to improve forecast PM_2.5 model simulations. Benefits of space-based/model-assisted PM_2.5 estimates include helping to approximate PM_2.5 concentration levels for those areas lacking in situ ground station coverage, providing nighttime characterization of air quality, and assessing spatial and temporal variations of PM_2.5 pollution on both regional and global scales. For this Models, In situ, and Remote sensing of Aerosols (MIRA) Topic, we aim to provide inter-comparisons of various methods and techniques for retrieving surface PM_2.5 assisted by satellite remote sensors, global aerosol models, and in situ aerosol measurements. These comprehensive PM_2.5 estimates can be useful for current and future efforts in air quality research, modeling, forecasting, and applications. We seek international in situ datasets (e.g., mass scattering/absorption coefficients and aerosol hygroscopic properties) for various aerosol species to develop more robust PM_2.5 estimates. We also seek international ground-based in situ PM_2.5 datasets to validate the PM_2.5 concentration estimates.

Since 2019, the Korean government has been implementing a PM_2.5 Seasonal Management Policy during wintertime (December to March in next year) to protect public health by mitigating the frequency and concentration of high PM_2.5 events. To investigate the difference in the chemical composition of PM_2.5 at ground-level and mid-upper planetary boundary layer (PBL) during winter (January to March) in 2023, 24-h PM_2.5 filter samplings were conducted at both the ground-level and top of POSCO Tower (37.39°, 126.64°, 305 m a.s.l) along with co-located air quality monitoring stations (AQMS) which provide hourly concentrations of PM₁₀, PM_2.5, CO, NO₂, SO₂, and O₃. POSCO Tower is the sixth-highest building in Korea thus the Korean first AQMS at high altitude was established in 2023. Using low volume sampler (16.7 LPM), filter samples were collected on the 47 mm Teflon filter (PTFE Filter with support ring, 2.0 μm; MTL) for mass and trace metal concentration, and the quartz fiber filter (Tissuquartz™ 2500QAT-UP, Pall Pallflex) for water-soluble ions and carbonaceous concentration. The concentration of chemical components (ion, trace metal, and carbon) will be analyzed to utilize the input into a positive matrix factorization (PMF) analysis, which will allow us to identify different emission sources between two vertical sites by combining the backward trajectory analysis. We believe that this study will be contributed to not only identifying the major emission sources at two vertical sites, but also enhancing the accuracy of vertical PM_2.5 prediction in the chemical transport models and remote sensing techniques.

Aerosols are not only air pollutants that are linked to human health, but also an important factor in understanding earth radiation and climate change. As aerosol optical depth (AOD) is a quantitative estimate of aerosols present in the atmosphere, it can be used as a proxy of air quality. Satellite-based AOD is mainly obtained through a radiative transfer model, demanding heavy computation and having uncertainties. To overcome these challenges, this study proposed a machine learning-based model to estimate hourly AOD from the Geostationary Environment Monitoring Spectrometer (GEMS) over Asia-Pacific regions (5°S–45°N, 75°E–145°E). The model was constructed using light gradient boosting machine, which has top-of-atmosphere (TOA) reflectance and observation angle data obtained by the GEMS satellite sensor, meteorological variables from the numerical model, and other auxiliary variables as independent variables. Aerosol Robotic Network (AERONET)-measured AOD data were used as a dependent variable. The proposed model was evaluated using random, spatial, and temporal 10-fold cross-validation (CV) with in-situ data to examine the spatiotemporal generalization possibility. The results of the random and temporal CV showed good agreement with the reference data, resulting in the determination of coefficient (R²) of 0.87–0.90 and root-mean-square-error (RMSE) of 0.11–0.12. However, the spatial CV results yielded slightly lower accuracy, with an R² of 0.67 and an RMSE of 0.19. This explains the low spatial transferability in areas where ground stations are rare or in low-latitude regions where input data for modeling are often unavailable due to frequent precipitation. However, it is confirmed that the accuracy of AOD estimation is comparable to or even better than the operational GEMS AOD product. Our results indicate that the proposed approach using TOA reflectance data from geostationary satellite sensors has great potential for estimating AOD for operational purposes.

The extent of sea ice coverage over the Arctic Ocean has dramatically declined over the past few decades. The anomalously warm Arctic surface associated with the Arctic sea ice loss has been linked to the mid-latitude surface cooling in the subsequent boreal winter. Several studies have suggested that this linkage could involve the wintertime stratospheric circulation by enhancing the upward planetary wave (PW) activity and weakening the polar vortex. Temperature anomalies induced by the vortex weakening could subsequently descend into the troposphere on time scales of weeks to months, potentially leading to cold air outbreaks (CAOs) over the continents. Cases studies based on observations or model simulations even suggested that the enhanced wave flux could be strong enough to cause the demise of the polar vortex associated with the sudden stratospheric warming (SSW) phenomenon. The atmospheric response to Arctic sea ice loss remains however a subject of much debate. Most studies have focused on the sea ice retreat in the Barents-Kara Seas and its troposphere-stratosphere influence. Here [1], we investigate the impact of large sea ice loss over the Chukchi-Bering Seas on the sudden stratospheric warming (SSW) phenomenon during the easterly phase of the Quasi-Biennial Oscillation through idealized large-ensemble experiments based on a global atmospheric model with a well-resolved stratosphere. Although culminating in autumn, the prescribed sea ice loss induces a near-surface warming that persists into winter and deepens as the SSW develops. The resulting temperature contrasts foster a deep cyclonic circulation over the North Pacific, which elicits a strong upward wavenumber-2 activity into the stratosphere, reinforcing the climatological planetary wave pattern. While not affecting the SSW occurrence frequency, the amplified wave forcing in the stratosphere significantly increases the SSW duration and intensity, enhancing cold air outbreaks over the continents afterward. 

Contributions of the equatorial planetary waves (Rossby wave (RW), mixed Rossby-gravity wave (MRGW), Kelvin wave (KW), Inertio-gravity wave (IGW)) and small-scale gravity wave (SSGW) to the 2015/16 QBO disruption are examined using the ERA5 data, which have finer horizontal and vertical resolutions than some other global reanalysis data sets used in previous studies. Each wave mode is selected by frequency and horizontal wavenumber criteria, based on theoretical studies, in each month for 1 year from July 2015 to June 2016. The contributions of planetary wave modes are evaluated by calculating Eliassen-Palm flux divergence (EPFD), while those of SSGW are evaluated by considering both the EPFD of IGW with horizontal wavenumbers larger than 20 and parameterized gravity waves (PGWs) calculated from an offline convective GW parameterization. It shows that RWs contribute mostly to decelerating the westerly winds at 40 hPa during the disruption, with stronger contribution by meridional-propagating components than vertical-propagating ones. MRGWs weaken the center of the westerly jet, which leads to a favorable condition for RW breaking. The negative forcing by IGWs is gradually increased since November 2016, and it is larger than RW forcing after March 2016. The negative forcing by IGWs is mostly from small scale waves with horizontal wavenumber larger than 20, which is much larger than that from some coarse-resolution reanalysis data. The PGW forcing using ERA5 is smaller than that using coarse-resolution reanalysis data in some previous studies, as more IGWs are resolved from ERA5. Accordingly, forcing by SSGWs including both small-scale IGWs and PGWs in the present study is similar to that using coarse-resolution reanalysis data. The positive forcing by KWs is prominent nearly all month and altitude, except that SSGWs provide stronger positive forcing in March 2016 at 20 hPa where strong westerlies sustain by additional support from the vertical advection.

The Brewer-Dobson circulation (BDC) is driven by dynamics and balanced by diabatic heating, which is dominated by radiative heating above the tropopause. Clouds interact with radiation and cause radiative heating perturbations which in turn can influence the BDC. This study evaluates how the ERA5 and MERRA-2reanalyses and satellite data represent the effects of cloud-induced radiative heating perturbations on the BDC. The stratospheric residual circulation is diagnosed using all-sky and clear-sky radiative heating, and the difference between the two solutions is interpreted to be the cloud effect on the BDC. Above 70 hPa, clouds in reanalyses induce positive upward velocity anomalies near the summer hemispheric poles, suggesting that clouds tend to dampen the high-latitude subsiding branch of the BDC. Below 80 hPa, the three separate estimates show little agreement, with ERA5depicting weak cloud effects while the MERRA-2 and satellite-based results are larger in impact but opposing signs. Satellite-based estimates show that tropical cirrus strengthen equatorial upwelling as well as tropics-to-subtropics meridional transport, while the MERRA-2 solutions suggest the opposite. This discrepancy is attributed to significantly larger tropical tropopause-level cloud ice water content in MERRA-2 compared to satellite radar/lidar retrievals, suggesting that unrealistically thick high-altitude clouds result in unrealistic cloud effects on the BDC. Results suggest that the properties and frequency of tropopause-level clouds strongly influence representation of the cloud effects on the shallow branch of the BDC.

Atmospheric gravity waves (GWs) in the mesosphere-lower thermosphere (MLT) are crucial for the understanding of general circulation. However, their dynamical characteristics are hardly retrieved due to the difficulty in the high-resolution observation of wind. Therefore, this paper uses eight years (2013–2020) of meteor radar measurements in the MLT region at Mohe station (53.5°N, 122.3°E), China, to retrieve high temporal-resolution mesospheric wind data and further evaluate the temporal variation of GW kinetic energy. As the detected meteor trails exceed 6, the wind velocity is recalculated using the least square algorithm method, significantly increasing the temporal resolution of wind from 1 h up to 5 min. This resolution is sufficiently high for the investigation of GW kinetic energy, which exhibits a high spatial-temporal variability. For instance, it is enhanced in the winter season during the period of 0200 1400 UT and in the spring season during the period of 0800–1300 UT. The similarity between the climatological characteristics of GWs in MLT and the seasonal variation of GW total energy in the troposphere, determined from high-resolution radiosondes near to Mohe station, suggests that the meteorology in the lower atmosphere could be an important source of GWs in the MLT region.

The meridional overturning mass circulation in the middle atmosphere, i.e. the Brewer- Dobson circulation (BDC), was first discovered before decades based on the distribution of trace gases and a basic analytical concept of BDC has been derived using the transformed Eulerian mean equations. Since then, BDC is usually defined as consisting of a diffusive part, and an advective, residual mean circulation. Climate model simulations robustly show that the advective BDC part accelerates in connection to the greenhouse gas induced climate change and this acceleration dominates the middle atmospheric changes in climate model projections. A prominent quantity that is being studied as a proxy for advective BDC changes is the net tropical upwelling, which measures the amount of mass advected by residual circulation across a given level (the tropopause, the stratopause, 70 hPa, 100 hPa isobar) and upwards. Another robust aspect of the changes in greenhouse gas concentrations is the changing structure of the atmosphere across layers. Particularly, it was debated whether the increasing BDC is not driven by the vertical shift of the circulation. We developed an analytical method that allows us to attribute the changes in tropical upwelling to kinematic causative factors such as the accelerating residual mean circulation, changing density of air, vertical shift of the circulation and for the first time, changes in width and geometry of the upwelling region. Here we demonstrate that this is the complete set of kinematic factors influencing the net upwelling and that all these factors are important contributions to the net upwelling trend depending on the level, where the upwelling is analyzed. In our presentation we describe the method and show its utilization on ERA5 and CMIP6 data for the shallow BDC branch and on CCMI models for the deep BDC branch and the transport between the stratosphere and mesosphere.

As the NASA Earth Observing System (EOS) program ages, the new generation of advanced operational weather satellites are natural successors for extending important imager cloud climate data records begun by MODIS. In addition, many of these low-Earth orbit (LEO) observations are expected to be coupled with like observations from the new generation of advanced geostationary (GEO) imagers (e.g., ABI, AHI, etc.), allowing for the possibility of a consistent LEO/GEO cloud product Program of Record (PoR) that can enable enhanced climate and process studies by NASA investigators and the broader research community. Note that a cloud product PoR is desired to provide critical synergy with the NASA Atmosphere Observing System (AOS), which is currently in formulation and designed to address the Aerosols, Clouds, Convection, and Precipitation Designated Observables identified by the 2018 NASA Earth Science Decadal Survey. Here, we give an overview of a NASA research cloud optical property algorithm for GEO imagers that was developed to be consistent, to the extent possible, with NASA cloud continuity products from LEO imagers (Aqua MODIS MYD06, VIIRS CLDPROP for Suomi NPP and NOAA-20). Challenges and approaches the team has taken to-date to best allow for GEO/LEO continuity will be summarized (e.g., the need for common geophysical algorithms across the sensor records to account for instrument differences along with forward radiative models, ancillary data sources, and inversion approaches). We will also discuss the need for a flexible computational infrastructure that that can provide a robust and flexible science algorithm testing and intersensory evaluation tools (including colocation and combined inter-sensor match files for on-orbit evaluation).

For validation of the recently launched GEMS (Geostationary Environment Monitoring Spectrometer), the Korean government has supported expanding the Pandonia Global Network (PGN) over the GEMS field of view by donating over 20 instruments to Asian Countries. Currently, over 8 instruments were deployed in Thailand, Mongolia, Indonesia, and Cambodia and we are aiming to install the rest of the units by the end of 2023. As one of Asia's top air quality priorities is aerosols, the NIER (National Institute of Environmental Research) is developing an additional aerosol inversion algorithm that can be used for various types of Pandora measurements. This presentation aims to share recent updates of the PAN (Pandora Asia Network) including aerosol retrievals which will be publicly available on the NIER website. Merits of simultaneous retrievals of trace gases and aerosols will also be presented.

Particulate matter (PM) is one of the most harmful atmospheric aerosols which have influenced adverse human health such as respiratory and cardiovascular disease. Rapid industrial and economic growth has resulted in a significant increase in PM₁₀ and PM_2.5 in East Asia. The aerosol optical depth (AOD) from the geostationary satellite is actively used to estimate ground-level PM₁₀ and PM_2.5 concentrations due to their high temporal resolution. Since satellite-derived AOD, however, is provided under high-quality conditions, there are spatially missing values due to the bright surface, cloud contaminations, and shadows. In this study, I proposed a machine learning-based hybrid-random forest and regression kriging (Hybrid-RFK) approach for filling gaps in the geostationary satellite-based ground-level PM₁₀ and PM_2.5 concentrations over East Asia. The AOD from the Geostationary Ocean Color Imager (GOCI) was mainly used to estimate satellite-based PM concentrations by real-time learning (RTL) model, while the hybrid-RFK was developed using the meteorological and land-based products without GOCI AOD to represent full-coverage PM distributions. The estimated PM concentrations by hybrid-RFK and the RTL-based PM models have been combined with each other for better performance in the AOD available region through gaussian filtering (i.e., combined RFK-RTL). In the clear sky, in which the AOD product is available, the RTL models (R²=0.91, rRMSE=20.79% for PM₁₀ and R²=0.90, rRMSE=26.56% for PM_2.5) show more improved performance than hybrid-RFK (R²=0.56, rRMSE=45.89% for PM₁₀ and R²=0.83, rRMSE=32.22% for PM_2.5) during the cases study period (i.e., 23 May 2015). One of the main reasons is whether the AOD values were used or not. Although the hybrid-RFK is comparatively low performed than RTL-models, the hybrid-RFK also shows good agreement with observed PM concentration for all sky conditions (R²=0.77, rRMSE =44.52% for PM₁₀ and R²=0.81, rRMSE=36.60% for PM_2.5) during the same cases study period.

Clouds are a vital part of the hydrological cycle and play an important role in the weather and climate of the earth-atmosphere system and play a crucial role in the modulation of the Earth’s radiation balance, dynamics, and thermal balance of the atmosphere. The impact of clouds on the radiation budget of the atmosphere depends on the cloud properties, such as their frequency of occurrence, the number of cloud layers, cloud base height, cloud thickness, spatial distribution, and water/ice content. An excess surface radiative heating produces instability in the atmosphere, resulting in convection and transfer of energy from the surface and its subsequent cooling and it affects the atmospheric circulation pattern which initiates several processes and affects the meteorology of the regions. Understanding the vertical distribution of clouds, their properties, and temporal evolution is essential for understanding the origin and impact of clouds and their feedback on the above-mentioned processes and their parameterization in weather and climate models for improving forecasts. This is correspondingly essential for investigating the cloud-aerosol interaction, heterogeneous chemistry in the atmosphere, and other associated processes. Cloud properties can be investigated using in-situ observations, ground-based lidar and radar remote sensing, airborne detectors, and space-borne payloads (onboard satellites). These techniques and observing platforms are having their advantages and limitations. Lidar and Radar-based observations are having very good temporal and vertical resolution but have limited spatial coverage. Satellite observations have the advantage of providing global distribution of clouds. In this talk investigation of the atmospheric processes which are modulated due to clouds will be presented. Furthermore, the observation of clouds using various techniques and the implication of the advanced techniques in improving the understanding of atmospheric clouds will also be discussed and presented.

Aerosols are one of the major components that affects climate and air quality. Being able to measure global aerosol comprehensively has been a major goal for the last several decades. Now with the increasing number of sensors that are capable of retrieving aerosols on both geostationary orbit (GEO) and low-Earth orbit (LEO), getting a complete picture of global aerosol distribution is more achievable than ever. However, how to use these data with various temporal and spatial resolution synergistically is one of the urgent questions that needs to be answered before combining products. Using a consistent Dark Target algorithm on three GEO sensors (two Advanced Baseline Imagers (ABI) on GOES-E and GOES-W and Advanced Himawari Imager (AHI) on Himawari-8) and three LEO sensors (Moderate resolution Imaging Spectrometers (MODIS) on Terra and Aqua and the Visible Near-Infrared Imaging (VIIRS) on Suomi-NPP), we evaluate six level 2 DT aerosol optical depth (AOD) products against Aerosol Robotic Network (AERONET) as well as Marine Aerosol Network (MAN) and investigated the similarity and differences among the products with a special focus on the two GEO and LEO common regions, namingly North America and East Asia. The error statistics of these products are generated against observing conditions with major uncertainty sources identified. Our results provide baseline evaluation results before synergy of DT aerosol products can be pursued.

It is cost-effective to use X-band polarimetric radar to fill the large gap between C-band or S-band radar networks for severe weather monitoring and nowcasting; however, X-band radar suffers more attenuation than C- and S-band due to its short-wavelength characteristics, particularly in convective situations. In this work, a combined utilization of a radar network, which includes a C-band radar in Hangzhou located on the top of a western mountain (1512 m), an S-band radar in Ningno near the East Sea (454 m), as well as an X-band radar (816 m) located in their vertical gap, is used to demonstrate the capability of X-band radar in analyzing the developing convective rainstorms, after the attenuation correction for horizontal reflectivity (Z_H) of all three radar. The result shows that (i) after the significant attenuation on Z_H is corrected based on the self-consistency approach, X-band radar exhibits finer and even larger Z_H than S- and C-band radar when it captures the lowest part of the convective rainbands relative to the other two radars. (ii) The trigger and developing convective rainbands occurred on the lower atmospheric layers, which are insignificant in the view of S- and C-band radar because of longer distances, are all well captured by the X-band radar in their gap area. (iii) Benefiting from the nearest distance, S-band radar afterward captured the lowest developing part of another convective rainband with the most significant Z_H, which is different from but similar to that captured by X-band radar. In these senses, the X-band radar effectively favors diagnosing the trigger and developing of convective rainstorms in the gap area of this S- and C-band radar network.

Polarimetric radars provide new insights into the precipitation microphysical processes and characteristics in a tropical cyclone (TC, Kajiki 2019) inner rainband and a tropical squall line in the South China Sea. The precipitation of Kajiki is dominated by high concentrations and small (< 3 mm) raindrops, which contribute more than 98% to the total precipitation. The formation of precipitation in the inner rainband of Kajiki is dominated by warm-rain processes, and accretion and coalescence play a critical role in the formation of heavy rainfall. At the same time, the melting of different particles generated by the ice process has a great influence on the initial raindrop size distribution (DSD). The aloft DSD above heavy rain with the effect of graupel has a wider spectral width compared with the region affected by dry snow. A tropical squall line was also documented. The ZH is similar for both convective systems, while the ZDR (KDP) is higher (lower) at the front of the squall line, indicating a low concentration of large drops; the peak values of ZH, KDP, and ZDR in the inner rainband are all concentrated in the convention center. The trend of liquid water content toward the ground within the squall line is consistent with that of the inner rainband, while the mean value is lower, and is more affected by ice processes. This work may contribute to our understanding on cloud microphysics of tropical convective systems using polarimetric radars.

Mesoscale convective systems (MCSs) are frequent in Southern China (SCH) during warming season and can result in severe weathers. While the large-scale circulations and precipitation patterns of MCSs over SCH have been well studied, studies on finer-scale (meso- and micro-) structures of MCSs are still limited, especially changes in MCS structures associated with different large-scale circulations. In this study, we first identify and track nearly three warm seasons of MCSs over SCH using operational radar network data in China. Multi-source observations are then collocated into MCSs to examine the precipitation intensity, horizontal organization, vertical structure, and ensemble microphysics of MCSs. MCSs are stratified by six types of large-scale circulations, including front (7%), southwest vortex (SW-Vortex, 15%), trough (10%), tropical cyclone (TC, 18%), South China Sea vortex (SCS-Vortex, 10%), and other (40%). Our results reveal significant variations in the precipitation structures and convective characteristics of MCSs developed under various large-scale circulations. For example, MCSs of SW-Vortex and trough types exhibit the strongest and deepest convection as indicated by radar and IR observations, but the trough-type MCSs have a much smaller precipitation area and a shallower column of ice particles (microwave-based) likely due to a drier troposphere. Interestingly, MCSs associated with TCs and SCS-Vortex have the weakest convective intensity and the least lightning density, although they produce heavier precipitation at the surface. Mechanisms responsible for the variations in meso- and micro-scale structures among MCSs produced by various types of large-scale circulations will be further investigated.

Two X-band dual polarimetric radars were installed in the Seoul metropolitan area to supplement the limitation of low-level observation of the S-band radar network. High spatiotemporal resolution data of the X-band radar can be useful for flash flood monitoring and forecasting. However, there were unexpected beam blocking areas caused by artificial buildings, trees and terrain which cannot be detected using DEM (Digital Elevation Model) based on beam blocking simulation results and the real time beam blocking detection (consisting of the standard deviation of radar variables and co-polar cross-correlation coefficient). The frequency of occurrence of dual-polarization parameters (FOD) technique, an extended method of the frequency of occurrence of reflectivity technique (FOR, Chang et al. 2009) was used in this research to improve the beam blocking detection. This method is applied to the beam blocking detection and identifies ground clutter using feature parameters consisting of dual-polarization parameters (Z_DR, ρ_HV), proxy quantity dubbed depolarization ratio (DR) and their statistical variation (horizontal and vertical standard deviation, average, texture, etc.). The blockage areas are categorized using fuzzy logic based on FOD parameters. Different types of HSR masks (beam blocking simulation with DEM, real time beam blocking, FOR and FOD) were applied to Hybrid Surface Rainfall (HSR, Lyu et al. 2015) estimation with various methods (R(Z_H), R(Z_H, Z_DR), R(Z_H, K_DP), R(Z_H, K_{DP_P}) and R(Z_H, A_H)) to validate effect of FOD. QPE accuracy of the X and S-band radar was validated with rain gauges and compared with the QPE of S-band radar.

Acknowledgment
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646). This work was funded by the KMA Research and Development Program “Observing Severe Weather in Seoul Metropolitan Area and Developing Its Application Technology for Forecasts” under Grant (KMA2018-00125).

Raindrop Size Distribution (DSD) has been studied since the late 19th century by observing different disdrometers. Especially, the 2D-video disdrometer (2DVD) was widely used in many studies because the 2DVD has relatively high accuracy for the measurement of physical variables of drop by using two orthogonal optical cameras with high performance. However, the 2DVD data tends to underestimate the concentration of small raindrops, especially drop diameters below 0.7 mm. Against the underestimation issue, recent studies have tried to combine 2DVD data with Meteorological Particle Spectrometer (MPS) data showing a more realistic number concentration of small drops. Simply, the complete DSD is adding 2DVD DSD above 1.125 mm channel and MPS DSD less than 1.125 mm. Meanwhile, the complete DSD can be discontinuous when the number concentration difference between 2DVD and MPS nearby the binding point is too large. In this study, we propose an improved combination method based on the weighting factor. The new complete DSD was produced by applying the distance weighting factor between 2DVD and MPS about a specific drop channel range 1 mm to 1.25 mm. We compare the rainfall rate between the original complete DSD and the new complete DSD as a reference to Pluvio data. The comparison for summer rainfall events in 2022 in Incheon, South Korea shows the new complete DSD has relatively higher skill scores. We will also compare radar reflectivity between the two methods as a reference to Precipitation Occurrence Sensor System (POSS) data.

ACKNOWLEDGEMENT
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).

The dendritic growth layer (DGL) at an altitude of about -15 °C plays an important role in the production and growth of snowfall. In the DGL, the difference in saturated vapor pressure between ice and water is maximized, resulting in the rapid growth of dendritic ice particles and increasing ground precipitation. Despite the research efforts in previous studies, the microphysical characteristics of snowfall in the DGL and its role in the dynamical and thermodynamical processes are not yet fully understood. In particular, noticeable updrafts and bimodal particle size distribution in the vicinity of DGL are often observed, but their underlying mechanism remains unknown. Recently, Ye and Lee (2021) established a hypothesis that the generation of new small particles is favorable due to the supersaturation associated with the updrafts below.In this study, we will try to reveal the interaction between updrafts and snowfall microphysics in the DGL by using data from intensive observations during the ICE-POP 2018 (International Collaborative Experiments for Pyeongchang 2018 Olympic & Paralympic winter games). We analyzed the vertical profiles of polarimetric radar variables from surveillance radar, the Doppler spectrum of vertically pointing radar, and the 3-hr rawinsonde data. Furthermore, the vertical air velocity estimation technique was applied to the vertically pointing radar to quantify the updraft in the DGL. The three analyzed cases revealed the presence of the updraft around -15℃. There was an increase in reflectivity, a maximum value of differential reflectivity and specific differential phase, and a decrease in the cross-correlation coefficient in the region where the updraft was at its maximum intensity. We will suggest the possible mechanisms of the generation of small particles and updraft.

ACKNOWLEDGEMENT
This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).

Atmospheric motion vectors (AMVs) are of great significance for constructing initial fields of the numerical weather prediction and analyzing damaging weather processes such as typhoons and heavy precipitation. Based on the convolutional neural network (CNN) architecture, this study uses transfer learning in deep learning strategy to retrieve the AMVs from the infrared and water vapor images of the geostationary satellites Himawari and FY-4A respectively, and finally the output of the neural network model is stitched and merged to obtain the AMVs within the same observation area of the two satellites. The results show that CNN can effectively acquire the information from the satellite images and convert it into vector wind fields at various height layers. Compared with the traditional algorithms, AMVs retrievals with machine learning algorithms have the features of no requirements for the height assignment, the ability to display the complete atmospheric circulations at different height layers and the uniform AMVs distribution in the vertical direction.

Tropical cyclones (TCs) always bring torrential rains triggering loss of property, even lives. Nowcasting heavy rainfall from TCs within 2 hours is thus an important topic in meteorology and hydrology. Traditional numerical weather forecasting fails to provide accurate results due to spin up in such short lead time while deep learning-based extrapolation has proven its potential. In this work, we proposed a multi-tasking learning framework to forecast the rotating features of TCs. Spatial and temporal difference (STD) are coupled with structural similarity (SSIM) to better predict periphery rainfall bands and rotation. The framework adopted two different backbones namely convolutional long short-term memory (ConvLSTM) and U-Net neural network with fed of radar reflectivity composite. The results showed that: (1) the coupling of STD and SSIM could significantly increase performance matrix compared to traditional deep learning with SSIM loss function only. The average improvement of critical success index (CSI) of 30 dbZ are 84% and 8.9% in ConvLSTM and U-Net, respectively, and the decreasing of mean absolute error (MAE) are 7.1% and 4.6%. (2) The movement of TCs, including transformation and rotation are observed in our proposed framework, indicating the effectiveness of spatial and temporal difference. However, the blur issue exists, especially with lead time over 1 hour, which calls for physical driving coupled in deep learning.

In recent years, extreme precipitation has brought serious losses, accurate and timely forecasting of extreme precipitation has become more and more important. This study analyzes the radar echo data of some extreme precipitation in Shenzhen using Convolutional Long Short-Term Memory (ConvLSTM) network. Based on the radar echo data, our study calculates the optical flow field by variational echo tracking (VET) method, which is added to model input and prediction respectively. This study also explores why optical flow field can improve the performance of the model. The outcome shows that the total loss is reduced by approximately 7% compared with pure ConvLSTM when the optical flow field is added to model input, and 12% when it is also added to model prediction. It is also shown that the ConvLSTM model with optical flow field as additional input is equal to a more complex ConvLSTM model since optical flow field calculated by radar echo data can’t provide more real-world information. Meanwhile, when optical flow field is also added to model prediction, the model performs better because additional information is considered, which also provides the basis for combining with numerical weather forecast results.

Climate change will induce more extreme rainfall events, leading to less reliable flood susceptibility analysis based on the existing hydrological handbook. To better distinguish flood-prone areas in a more variable environment, this study aims to propose a hybrid assessment framework that uses a hydrodynamic model referred to as the Two-dimensional Runoff Inundation Toolkit for Operational Needs (TRITON) and three machine learning models, namely Convolutional Neural Networks (CNN), Random Forest (RF), and Support Vector Machine (SVM), for flood susceptibility analysis in Changhua county, Taiwan. TRITON, which makes use of graphics processing units to achieve high-performance computing, simulates Probable Maximum Precipitation (PMP) and Probable Maximum Flood (PMF) in various scenarios to identify flood susceptible sites that are adopted as the training and validation points for the machine learning models to develop flood susceptibility mapping. Preliminary results show that the “very high” flood susceptibility category accounts for 13% to 20% under the SSP5-8.5 scenario, suggesting a 6% increase compared to the baseline analysis. It is expected that the hybrid modeling framework can not only deal with data scarcity regarding historical flood susceptible sites but also produce flood susceptibility maps for disaster preparedness.

With the requirement of high accuracy and short responding time, rainfall nowcasting has become a significant challenge for weather authorities to provide timely and accurate alerts of floods to the public. Hence, three state-of-the-art deep learning models, the Convolutional Long Short-Term Memory (ConvLSTM), the Trajectory Gated Recurrent Unit (TrajGRU), and the Predictive Recurrent Neural Network (PredRNN), are trained and tested for rainfall nowcasting with a newly collected radar echo maps dataset provided by the Meteorological Bureau of Shenzhen Municipality in China. All three deep learning models outperform the traditional optical flow-based extrapolation method in all performance metrics, improving the performance by up to 63%. Performances of those models are further compared and discussed with three typical cases of different rainfall intensities in South China, showing the prediction accuracy for 0-2 hours nowcasting using deep learning. Data fusion between different cities in the same area is also tested for the building of high-quality rainfall dataset for nowcasting tasks. The difference is less than 5% when the model has multiple data sources which proves the robustness of the deep learning models. Analysis are carried out regarding the effectiveness of deep learning techniques on rainfall nowcasting, suggesting ways of improvement in the future.

Making sure of the safety of embankments, which are powerful measures to prevent river from flooding thus safeguarding people’s lives and property, is crucial, especially when human society is under rapid development and the impacts of climate change are becoming bigger and bigger. However, engineering measures previously constructed may not meet the risk resistance requirements at the time of embankments construction. This study evaluates the resilience of embankments against future risks of Maozhou River as a case study, which is a cross-border river connecting to the Pearl River Estuary. The modelling of water dynamics in coastal basins is very complex and challenging, for flood events in estuary areas are often compounded by extreme rainfall and tidal patterns. To overcome this challenge, this study uses the water level data of Maozhou River in Shenzhen, China and employs wavelet transform method to decompose the water level data into two time series parts: direct runoff related part and base flow related water level part, which correspond to the part influenced by precipitation and the part by tidal events, respectively. These two parts time series data are modelled separately using a hybrid machine learning model framework, and then combined together for the final composite water level analysis. This method preprocesses water level based on the rainfall-runoff process and can significantly improves model’s prediction performance when compared with models that simply uses rainfall and tidal events as covariates. Finally, with our method, using the rainfall data predicted from IPCC6, this study forecasts the future changes in water level with better accuracy, providing a scientific evaluation method for embankment safety.

From Oct-2019 to Sep-2020, a high-volume sampler was used to monitor atmospheric dry deposition over Mazu Island which is located at the southern end of the East China Sea. The collected samples were divided into coarse particles and fine particles using a six-stage particle size separator, and the nitrogen (N) and phosphorus (P) species were categorized as either water-soluble (WS) or water-insoluble (WI). In this study, WS-TN and WI-TN accounted for 74% and 26% of TN, while WS-TP and WI-TP accounted for 47% and 53% of TP, respectively. According to the historical literature, the proportions of WI-TN and WI-TP in TN and TP are 39-51% and 95% in inland areas, which are significantly higher than 8-26% and 53-63% observed in coastal atmospheres. Correlation analysis results showed that there was a high correlation between WS-TN and WI-TN (r = 0.79, p < 0.01) and a moderate correlation was noted among nss-K⁺, nss-Ca²⁺, and nss-SO₄^2-. The results show that WS-TP and WI-TP may be derived from fossil fuel combustion and biomass combustion, respectively. In terms of particle size distribution, the ratios of the concentrations of WS-TN, WI-TN and WS-TP within fine particles are 72%, 93% and 72%, respectively. However, WI-TP displays the opposite pattern since the proportion of fine particles is only 37%. Flux calculation results show that the annual fluxes of TN and TP are 48.4 and 0.29 mmol m^-2 yr^-1, respectively. Both TN and TP had the highest average daily flux during winter; of these, up to 88% of the annual N flux consisted of WS species. On the other hand, approximately 69% of P was dominated by WI species. The above results show that N has a higher bioavailability than P in the areas, which may cause nutrient imbalances in adjacent ecology. 

Nitrogenous organics (CHONs) are one of the most important components of secondary organic aerosol (SOA). However, the formation mechanism of CHONs in clouds has rarely been studied. Here, we observed the evolution of CHONs during cloud processes (CPs) and found that CPs significantly enrich the diversity of CHONs, specifically generating CHONs with O numbers of 1–10 and double bond equivalent (DBE) values of 2–10, which can be described by the formula C_nH_{(2n-16)–(2n)}O_1–10N_1–4. We proposed that the CHONs formed during CPs are likely formed through aqueous phase reactions with CHO compound precursors via nucleophilic attacks by NH₃. Nearly 77% of the CHONs in cloud water can be accounted for by this scheme, and 66% of all CHONs are formed through reactions between NH₃ and carbonyl-containing biogenic volatile organic compound (BVOC) ozonolysis intermediates. This study provides the first insights into the evolution of CHONs during CPs and reveals the significant roles of CPs in the formation of CHONs.

The clean heating renovation has been executed for improving particulate matter (PM) pollution in northern China since 2017. This study determined particle size distributions of nitrated phenols (NPs) in personal exposure (PE) samples and its associations with biomarkers in saliva and urine from homemakers in rural households of the Fenwei Plain, China. Remarkable reductions of 28.6–66.3% and 52.2–82.4% on PMs and total quantified NPs were found with the substitutions of raw coal chunk and biomass by advanced clean coal. 4-Nitroguaiacol (4NG) showed the largest reductions of 81.2% among individual NP. In addition, the clean coal efficiently reduced interleukin-6 (IL-6) and 8-hydrox-2'-deoxyguanosine (8-OHdG) in the urine and saliva by 12–72%. Furthermore, significant positive correlations between urinary 8-OHdG with most of NPs in all particle sizes, urinary IL-6 with 4NG for particles with D_p > 2.5 µm and D_p = 0.25–1.0 µm, salivary IL-6 with 4-nitrocatechol (4NC) and 4-methyl-5-nitrocatechol (4M5NC) for particles with D_p > 2.5 µm, D_p = 0.5–1.0 µm, and D_p < 0.25 µm were observed, but not for salivary 8-OHdG or PMs. The results provide scientific support for the clean energy reformation and demonstrate the strong particle size dependence between NPs and biomarkers.

The characteristics of the regional nitrogen cycle have not been fully understood and its response to the future emissions control strategy is warranted. Here, we used the WRF-CMAQ model to reveal the regional nitrogen cycle (emissions, concentrations, and depositions) in the atmosphere in January (winter) and July (summer) 2015 and anticipate its changes under emissions control by 2030 in YRD. The model has reasonable simulations for the meteorological factors, chemical concentrations, and depositions in 2015. The results show that the nitrogen concentration (deposition) is generally higher (lower) in January than in July due to higher precipitation and dry deposition velocity in summer. Local emissions of YRD are the major contributors to the concentration in this region. For the deposition, HNO3, NH3, NO3-, and NH4+ are the major species. The depositions of oxidated nitrogen and reduced nitrogen account for 62% and 38% of the total nitrogen deposition in January, while their proportions are 51% and 49% in July. To achieve the carbon peak emission in 2030, the NOx and NH3 emission is projected to reduce by ~51% and ~22%, respectively. The concentration and deposition of oxidated nitrogen will decrease by ~50%, generally in consist with the reduction of NOx. However, the NO3- concentration is less responsive to the emission reduction in January because the contribution of NOx reduction will be offset by the increase of atmospheric oxidation capacity during nighttime. The lower reduction of NOx than NH3 results in a higher proportion of reduced nitrogen than oxidated nitrogen in deposition. In addition, the smaller reduction of reduced nitrogen wet deposition than sulfur wet deposition and oxidated nitrogen wet deposition can raise the pH of precipitation and help alleviate the acid rain problem, especially in July.

The Integrated Multi-satellitE Retrievals for GPM (IMERG) product from the U.S. Science Team of the Global Precipitation Measurement (GPM) mission provides estimates of surface precipitation rate at 0.1° every half-hour globally, with three different Runs to cater to various latency requirements. The latest version (V07) of IMERG, which involves a uniform reprocessing of the full 20+ years of record spanning the TRMM and GPM projects, includes numerous changes to the algorithm that are expected to improve the precipitation estimates in different ways. This poster will describe the major changes in the algorithm from the previous V06 to latest V07 and show examples of the associated improvements. One prominent issue with IMERG V06 is an overestimation of the satellite-only precipitation estimates, especially during the TRMM era; this poster will summarize the cause and the solution adopted in V07. Another issue that was identified in the development of V07 is a 0.1°-eastward spatial offset in the gridded microwave estimates and hence most of the inputs; this poster will demonstrate the improvement resulting from the correction. As well, changes to the Kalman filter in IMERG are expected to improve the distribution of estimated precipitation rates and mitigate the differences arising from the use of diverse types of sensors. Other changes in V07 include a modification of the motion vector source to address deficiencies in the precipitation propagation near orography, a modernized infrared precipitation algorithm, an update to the precipitation phase parametrization for improved consistency, and a renewed effort to implement climatological gauge adjustment to the near-real-time Runs. These enhancements in IMERG V07 are focused on the goal of providing the community with a long record of reliable high-resolution precipitation observations.

The Global Precipitation Climatology Project (GPCP) products address the need for long-term precipitation products that emphasize homogeneity, following Climate Data Record (CDR) principles.  The new-generation Version 3.2 provides key improvements over the operational Version 2.3 such as: finer spatial resolution of 0.5°x0.5°; wider geosynchronous infrared estimation (58°N-S) upgraded with the PERSIANN-CDR algorithm; upgraded retrievals from selected passive microwave sensors (GPROF algorithm) that calibrate the IR input; revised intercalibrations of TOVS and AIRS data (used at high latitudes); climatologies based on CloudSat, TRMM, and GPM to provide overall calibration by modern satellite estimates; the latest Global Precipitation Climatology Centre (GPCC) precipitation gauge analyses over land areas; regional modifications to the gauge undercatch correction; and IMERG half-hourly data input to the Daily V3.2 product. We will show sample analyses that demonstrate aspects of the Version 3.2 precipitation record, such as the global climatology, the time series for global land and ocean total precipitation and snowfall, and the time series of tropical land and ocean daily precipitation rate histograms.  For selected analyses we will show improvements in both the Monthly and Daily products in Version 3.2 compared to the operational Version 2.3.  In particular, the climatological zonal profile of precipitation in the Southern Ocean, extending south of 40°S, improves a suspected artifact in V2.3.  Similarly, the Daily histograms over ocean in Version 3.2 lack the jump in the predecessor Version 1.3 Daily over ocean at the start of 2009, although a smaller jump is introduced in June 2014. The presentation will conclude with a prospectus for the future satellites/sensors and community datasets necessary to continue computation of a consistent CDR product on the one hand, while also potentially contributing to improvements in the historical record.

A new balloon-borne precipitation particle imaging radiosonde named "Rainscope" has been developed. The decisive difference between the Rainscope from the ground-based disdrometer is that it can know the vertical distribution of microphysical parameters. In other words, from the liquid phase to the ice phase such as graupel and snowflakes that exist above the freezing level, their types, shapes, sizes, spatial concentrations, etc. in clouds are provided. It is a powerful tool for the ground validation of remote sensing techniques. For example, it can provide observational data useful for the validation of GPM DPR products such a "binHeavyIcePrecipTop", which was newly released in 2021 and shows the upper end elevation of strong solid precipitation in clouds below -10ºC. The Rainscope was first deployed in the intensive observation campaign, which was conducted in Kyushu region during the Baiu rainy season in 2022. It was launched into convective clouds with active lightning and gust on 25 June 2022. It transmitted images of ice particle in the process of melting just below 0ºC, and frozen particles with semi-transparent and smooth outlines around 0ºC. And white and irregularly shaped graupel were observed across all layers up to -30ºC level. The clear particle images captured by the Rainscope enable us to get more detailed information of particle shapes, surface conditions, and contours, making it easier to evaluate their shapes quantitatively. The circularity of graupel was smaller in upper layer. It has a longer circumference and more irregular shapes, suggesting an active riming process originated from ice crystals. On the other hand, graupel in lower layer with larger circularity was suggested to be originated from a frozen particle. The different graupel formation processes was considered to exist in convective clouds.

The Global Precipitation Measurement (GPM) project leads the global distribution of precipitation by combining observations from multiple microwave radiometers (PMWs) onboard low Earth orbit satellite. Since temporal resolutions shorter than 3 hours cause sampling gaps only by PMW observations, a combined PMW-infrared (IR) combined algorithm has been developed to interpolate precipitation areas based on cloud areas and their moving vectors using IR images from geostationary meteorological satellites. It is difficult to accurately capture cloud areas at high latitudes due to radiometric effect at large zenith angle from the geostationary Earth orbits. Therefore, the coverage of major PMW-IR combined satellite precipitation map products has been limited to the equatorial region below 60 degrees latitude. The Global Satellite Mapping of Precipitation (GSMaP) product version 05 (algorithm version 8), released in December 2021, provides precipitation estimation up to the poles only in the PMW observation area. However, a polar extension of the PMW-IR combined algorithm has been expected to fill the remaining PMWs sampling gaps. This study examined the derivation of moving vectors using Nonhydrostatic Icosahedral Atmospheric Model-Local Ensemble Transform Kalman Filter (NICAM-LETKF) Japan Aerospace Exploration Agency (JAXA) Research Analysis (NEXRA), which is in regular operation together with GSMaP to extend the PMW-IR combined algorithm to the polar regions. The moving vectors of column water vapor and precipitation were calculated by modifying the current version of GSMaP moving vector calculation program to extend poleward and to allow the use of geophysical quantities other than IR. In this presentation, we will compare the results with other numerical weather forecast data used in previous studies and discuss issues for their use.

Akisame season is the autumn rainy season in Japan, with heavy rainfalls particularly recorded in eastern Japan. However, research on heavy rainfalls in the Akisame season is not as sufficient as those for the Baiu season. The objective of this study is to clarify the environment and precipitation characteristics of widespread heavy rainfalls during the Akisame season. We utilized gauge data collected by the Automated Meteorological Data Acquisition System (AMeDAS) stations for 1979-2020 to define the widespread heavy rainfalls. Japanese 55-year Reanalysis (JRA-55) and rainfall data from the Global Satellite Mapping of Precipitation (GSMaP) are utilized to analyze the environments and rainfall distributions. GPM/DPR data are utilized to examine the characteristics of rainfall events embedded in the widespread heavy rainfall events in Akisame season. Widespread extreme precipitation event cases in western Japan in Baiu season and in eastern Japan in Akisame season are compared. A northwest-southeast dipole structure in the geopotential height anomalies is found from the lower to upper level for the Baiu case, while a northeast-southwest dipole structure is found in the Akisame case. These different structures of anomalous field tend to enhance the water vapor flux associated with the different mean large-scale circulations in the two rainy seasons, preparing large precipitable water field. Next, three-dimensional rainfall characteristics are examined with the GPM/DPR overpasses for four representative cases of widespread extreme precipitation events in Akisame/Eastern Japan cases. Two of the four representative cases are associated with precipitation systems with strong convective precipitation. Other two cases are characterized as organized precipitation systems with very intense precipitation, but with moderate heights. The latter cases are similar to the precipitation system observed in the record-breaking extreme rainfall event in July 2018.

Satellite observations have brought opportunities to quantify the amount and distribution of precipitation over the globe, critical to understand how the Earth system works. The amount and spatial distribution of oceanic precipitation from the latest versions (V7 and the previous version) of the Global Precipitation Measurement (GPM) core-instruments and the constellation of passive microwave sensors are quantified and compared with other products such as the Global Precipitation Climatology Project (GPCP V3.2), the Merged CloudSat, TRMM and GPM (MCTG), and ERA5. Results show that GPM V7 products have higher precipitation rate than the previous version, except for the radar-only product. For the ~ 65^oS-N region, covered by all of the instruments, this increase ranges from about 9% for the combined radar-radiometer to about 16% for radiometer-only product. While GPM precipitation products still show lower values than MCTG (except over the tropics and Arctic Ocean), the V7 products (except radar-only) are generally more consistent with MCTG and GPCP V3.2 than V5. Precipitation products are found to be least consistent over the Southern Oceans, displaying the largest spread in mean precipitation rate and location of latitudinal peak precipitation. Over the tropics, (25°S-25°N) passive microwave sounders show the highest precipitation rate, and the highest increase (~19%) compared to their previous version, among all of the studied precipitation products. Furthermore, analysis showed that precipitation products are generally less consistent (larger spread) over lower and higher values of sea surface temperature and total precipitable water. Analysis of mean monthly precipitation variation (MPV) showed that MPV is relatively small when averaged over 90°S-90°N and 65°S-65°N (e.g., ~4-6% of the annual mean precipitation rate), but is large over the tropics (e.g., 9-11%) and high latitudes (~26-55%). 

Over the Arabian Sea, the tropical cyclone (TC) formed during the post-monsoon season is investigated through the Global Precipitation Measurement (GPM) satellite level 2 observation. The precipitation characteristics of the TC developed between 2014 and 2021 are analyzed to understand the cyclonic rainfall's precipitation characteristics. 2-D distribution between the liquid water content (LWC) and non-liquid water content (IWC) shows considerable variations in the different precipitating cloud. In the convective cloud, the LWC is 0-800 g/m^2, and IWC is about 0-300 g/m². The amount of LWC is low, and IWC is higher in the stratiform cloud than in the convective cloud. In the stratiform (convective) cloud of the TC, the mean of normalized intercept parameter (N_w) and mass-weighted mean diameter (D_m) is 34.0 (35.54) mm^-3m^-1 and 1.27 (1.31) mm, respectively. Moving from the smaller to the bigger rain droplets, the value of N_w decreases, as observed in the 2-D distribution between D_m and N_w. The concentration of the bigger rain droplets is high in the convective cloud, and stratiform cloud is dominated by the smaller rain droplets. The contoured frequency altitude diagram (CFAD) of the D_m and reflectivity (Z_e) for the stratiform and convective cloud of the TC indicate significant variation in the droplet's state above and below the melting layer. Ice particles and droplets in the supercooled liquid state dominate above the melting layer, and below it, a high concentration of rain droplets in the liquid form observed. Below the melting layer, the contribution of the various microphysical processes (size-sorting, evaporation, breaking-up and collision-coalescence process) in the rain droplets formation is different for the stratiform and convective cloud. The breaking-up process acts as the primary microphysical process in the stratiform cloud, and the collision-coalescence process is predominating in the convective of the TC.

The Korean Integrated Model (KIM) is a global NWP model, which was developed in 2020 for the operational medium-range forecast at Korean Meteorological Administration. The Korea Institute Atmospheric Prediction Systems has a mission to improve KIM predictability for high-impact weather especially over the Korean Peninsula with an approach to increase resolution and advance physics parameterization. In this study, we evaluate heavy rainfall events simulated by KIM during summer 2022 by using both in-situ and remote sensing measurements. It is found that KIM simulates overall synoptic conditions for summer heavy rainfall events, which are characterized with upper level trough, low level jet, and migratory cyclones with North Pacific High located in the east of the Korean Peninsula. However, it also shows that precipitation prediction performance can be significantly affected by the simulated thermodynamic conditions by the model. Meanwhile, the model shows strong sensitivity to the cumulus parameterization scheme especially in events with an east-west rain belt passing from the West sea to the central part of Korea. The characteristics of the vertical structures of simulated rainfall events are further evaluated using remote sensing measurements to give comprehensive understanding to improve model predictability.

Various terrain-following vertical coordinates have been developed to mitigate the influence of the terrain structure in atmospheric numerical models. However, it still remains the problems in computing horizontal pressure gradients and advection along steep terrain surfaces. These small-scale structures can further promote artificial circulations as the model resolution increases. Choi and Klemp (2021) proposed the smoothed hybrid sigma-pressure (SMH) coordinate to reduce small-scale components more rapidly with height than basic sigma or hybrid sigma coordinates. The SHM coordinate is designed to flexibly control smaller scale of terrain influences with height by separating a reference surface pressure to smoothed and deviated parts. The performance of SMH coordinate was demonstrated in two-dimensional slice version of KIM through several idealized tests. Compared to the hybrid sigma coordinate, the SMH coordinate showed significant improvement in reduction of topographic features. This study examines the impact of the SMH coordinate in a full version of KIM. The SMH coordinate has potential to well control smaller scale of terrain features so that the model can be allowed to have less smoothing of the terrain in high resolution. With testing idealized experiments and real forecasts, the SMH performance will be discussed in terms of model stability and accuracy. Reference: Choi, S.-J., and Klemp, J. B., MWR, Vol. 149 (2021), 4077-4089. Acknowledgement: One of the authors, Suk-Jin Choi, wishes to acknowledge this study was supported by 2023 Academic Research Support Program in Gangneung-Wonju National University.

The uneven spatiotemporal distribution of rain gauge data in Taiwan demands a need for a comprehensive gridded product that can be obtained from satellite observations, such as the Integrated Multi-satellitE Retrievals for Global satellite precipitation measurement (IMERG) product. However, IMERG still poses known biases and can sometimes be erroneous. In order to enhance the accuracy of satellite rainfall estimates, this study aims at applying two deep learning (DL) models, namely the Gated Recurrent Unit (GRU) model and the Trajectory Gated Recurrent Unit (TrajGRU) model, to correct IMERG Early and Final Run data in Taiwan at different sub-daily scales. Compared to traditional recurrent neural networks, GRU exhibits better model efficiency with a low memory usage, while TrajGRU incorporates a deformable convolution kernel able to effectively capture spatiotemporal features and improve predictive capabilities. Gauge data from Taiwan’s Central Weather Bureau are adopted as ground truth in the training and validation of DL models. Relative performance in the corrected IMERG products will determine the capabilities of the DL models as well as the limitation of time scales for such correction scheme.

The intense winds in a Typhoon environment over mountainous regions can trigger strong convections and produce enhanced precipitation, yet few previous studies have been done in understanding how the orographic precipitation in a TC environment might respond to global warming. Here, we focus on this problem and use large-eddy simulation (LES) to estimate the global warming-induced response of precipitation over and near an idealized mountain with 10km half width and 1km height. Pseudo-global warming experiments are further conducted by adding mean thermodynamic structure changes predicted by a CMIP6 climate simulation. Two regions exhibit locally enhanced precipitation. One is over the mountain with rainfall maximum on the lee slope; the other is in the downstream region 30 to 50 km away from the mountain. Regression analysis suggests the enhanced precipitation in both regions are related to the seeder-feeder mechanism. Though the enhancement in the downstream regions differs from the conventional definition. Mountain waves generate an intense cloud formation center in the upper troposphere above the mountain, resulting snow and graupel drift downstream which intensify downstream convection when they fall into a proper height. This precipitation enhancement mechanism in the downstream region is called as Pseudo Seeder-Feeder Mechanism (PSF). Under global warming, the downstream enhanced region exhibits mean precipitation sensitivity ~18%/K which is significantly higher than the Clausius–Clapeyron (CC) scaling of ~7%/K of surface warming. This super-CC scaling in the downstream region is induced by both the precipitation efficiency and thermodynamic enhancement, of which the precipitation efficiency is the dominant factor. Additionally, the larger sensitivity in the downstream region is partially due to the upwind shift of the enhanced precipitation, which in turn is attributed to changes in the mountain wave patterns.

Accurate quantification of future extreme tropical precipitation is challenging because global climate models have insufficient resolutions to resolve convection. Yet reliable estimation is needed to prepare our society for the increasing risks of flooding and landslide as extreme rainfall is expected to intensify generally as the climate warms. Here, we combine deep learning (DL) and convection-permitting simulations to provide reliable estimations of the future change in extreme precipitation in South China. The DL model was trained to statistically downscale low-resolution climate model data and proved more effective in identifying extreme precipitation events than simply using global climate model rainfall. High-resolution (1km) downscaling simulations are conducted for DL-selected cases so as to verify DL prediction and provide a more accurate, physical modeling-based prediction of precipitation. We found that the sensitivity of extreme rainfall at the grid-point scale is close to the Clausius-Clapeyron (CC) scaling, around 7%/K. However, for large spatial scales and temporal accumulation, the sensitivity can be as high as 17%/K. Such accelerated response of extreme precipitation in the urban-relevant scales is found to result from more compact organizations of convective cores in tropical cyclones. Such super-CC scaling has important implications for future urban flooding and slope safety.

Convolutional neural network (CNN) is applied to identify the extreme precipitation in southern China in summer and the physical contributions are quantified for the interdecadal changes in the extreme precipitation. The CNN correctly identifies 93% of the observed extreme precipitation events based on the given large-scale atmospheric circulation through hyper-parameter optimization and model training. The discrimination made by the neural network is revealed by using the layer-wise relevance propagation method. The CNN is inclined to classify the circulation pattern as an extreme precipitation circulation pattern (EPCP) when the low pressure anomalies of sea level pressure and 500-hPa geopotential height appear in southern China. Both the extreme precipitation amount and the frequency of EPCP show significant interdecadal changes around the early-1990s. Result of the partition analyze indicates that the interdecadal increase in the extreme precipitation in the last three decades is mainly attributed to the dynamic influence of the increasing frequency of the EPCP. The thermodynamic effect of the increasing watervapor also plays an important role in the increase of both the extreme and non-extreme precipitation after the early-1990s.

Aiming at the characteristics of high noise, nonlinearity and non-smoothness of ionospheric Total Electron Content (TEC) time series, this study constructs a combined short-term ionospheric TEC prediction model based on long short-term memory (LSTM) neural network model, incorporating singular spectrum analysis (SSA). The data of geomagnetic storm periods in high solar activity year and geomagnetic calm periods in low solar activity year are selected, and the trend component, period component and noise residual component of TEC time series are firstly extracted using SSA, and then the LSTM model is used to implement TEC prediction at different geomagnetic conditions and geographic latitudes. The results show that the relative accuracy of SSA-LSTM model is 91.17% and 95.46% respectively, which is 4.92% and 3.17% higher than that of single LSTM model, during the period of geomagnetic storm periods and geomagnetic calm periods.

It is well known that there is a strong correlation between the intensity of nighttime OI 135.6nm emission and the critical frequency of ionospheric F2 region (foF2). However, the relationship depends on the different longitudes and latitudes. In this study, an improved algorithm was developed to estimate foF2 from nighttime OI 135.6nm emission. The proposed method is suitable for different longitude, latitude, local time, season and solar activity. This study estimated foF2 from OI 135.6nm emission observed by the Special Sensor Ultraviolet Spectrographic Imager (SSUSI) instrument on board the Defense Meteorological Satellite Program (DMSP), and then the estimated results were compared with the results of ground-based ionosonde. Results show that the data with relative error less than or equal to 20% accounted for 92.99%, and the average relative error was about 7.08% during the high-solar activity year (2013). During the low-solar activity years (2017), the data with relative error less than or equal to 20% accounted for 80.76%, and the average relative error was about 12.64%. Furthermore, we discussed the possible reasons about the difference of retrieval accuracy of the algorithm during the high and low solar activity years.

The ionospheric sporadic E layer, the ionospheric irregularities of enhanced electron density, appears in the Earth’s ionosphere at altitudes between 90 and 120 km, which supports the real-world radio communication needs of many sectors reliant on ionosphere-dependent decision-making. The prediction of the occurrence of sporadic E layers has been extremely difficult due to the highly complex behavior. Conventional numerical methods are limited because of the inability to extract high-level information from data. Deep learning is a powerful tool for mining latent features from data, which can theoretically avoid assumptions constraining physical methods. Inspired by feature extraction, we applied deep learning to explore latent relationships between mapping observable lower atmospheric data and ionospheric data from limited observations. The proposed model was trained with high-resolution ERA5 data during Jan. 1, 2007-Aug. 30, 2018 as input and corresponding ionospheric sporadic E data collected from COSMIC RO measurements as output. The results show that the model can learn complex relevance as bridges connecting the input and the desired output and obtain excellent performance and generalization capability by applying multiple evaluation criteria. Additionally, we established several model architecture training methods to explore the performance of the model with different input data. The statistic results show that model inference performance is proportional to the abundance of input information and is impacted by intraseasonal variability. The inference capability of the model achieves the best performance in the Jun.-Aug. (JJA) and Dec.-Feb. (DJF) seasons, which is the exact period of sporadic E layer significant occurrence, although different models are evaluated.

The increase in vegetated areas in China led to increased biogenic volatile organic compounds (BVOC) emissions. As the important precursor for ozone (O₃) and secondary organic aerosol (SOA), an increase in BVOC emissions will impact the formation of O₃ and SOA. Thus, accurate estimation of BVOC emissions is critical to understanding their impacts on air quality. In this study, Model of Emissions of Gases and Aerosols from Nature (MEGAN) v2.1 was used to investigate the impact of different leaf area index (LAI) and land cover (LC) datasets on BVOC emissions in China in 2016 and the effects on O₃ and SOA were evaluated based on the Community Multiscale Air Quality Modelling System (CMAQ). Three LAI satellite datasets of the Global LAnd Surface Satellite (GLASS), the Moderate Resolution Imaging Spectroradiometer (MODIS) MOD15A2H version 6 (MOD15), and the Copernicus Global Land Service (CGLS), as well as three LC satellite datasets of the MODIS MCD12Q1 LC products, the Copernicus Climate Change Service (C3S) LC products, and the CGLS LC products were used in five parallel experiments (cases: C1-C5). Results show that changing LAI and LC datasets of the model input has an impact on BVOC estimations. BVOC emissions in China range from 25.42 to 37.39 Tg in 2016. Changing the LC inputs for the MEGAN model has a more significant difference in BVOC estimates than using different LAI datasets. Besides the impact on BVOC emissions, changing the MEGAN inputs further impacts the concentrations of O₃ and SOA. The highest O₃ and biogenic SOA (BSOA) concentrations appear in the C1 (using GLASS and MCD12Q1 LC) simulation, which can reach 12 ppb and 9.8 μg m^-3, respectively. Due to the effect of BVOC emissions, the relative differences between C1 and C4 are over 52% and 140% in O₃ and BSOA in central China.

Ozone (O₃) pollution is influenced by a combination of NOx and VOCs emissions. The biogenic volatile organic compounds (BVOC) emitted from plants could be an important source and the production is significantly impacted by temperature changes. This study will investigate the effects of temperature on Biogenic VOCs emissions and its contributions to ozone formation in Pearl River Delta in summer, 2022 using Model of Emissions of Gases and Aerosols from Nature (MEGAN) and the Community Multi-scale Air Quality (CMAQ) model. The meteorology will be generated using Weather Research and Forecasting (WRF) model version 3.9.1 and biogenic emissions will be generated based on MEGAN version 2.4 with two scenarios: base case and temperature increased 2K case. The differences in spatial variations of biogenic emissions, ozone concentrations and SOA concentrations between two scenarios will be analyzed. The results show a significant increase in BVOC emission (24%) and ozone concentrations (~2 ppb) with 2K increase of temperature in Pearl River Delta in summer, 2022. BVOC emissions are highly sensitive to the temperature changes, and we need to adjust our air pollution control strategy based on temperature effects especially in Pearl River Delta.

Biogenic volatile organic compounds (BVOC) can contribute a significant fraction to secondary organic aerosols (SOA) through atmospheric oxidation, which plays a critical role in climate change and human health. The ozonolysis of α-pinene, one of the most important BVOCs, is a canonical SOA system. At present, the effects of relative humidity (RH) on SOA composition from α-pinene+O₃ reaction are still unclear. In this study, we report the SOA composition on molecular level formed in α-pinene+O₃ reaction under various RH. The SOA components were measured by an Extractive ElectroSpray Ionization inlet coupled with a long Time-of-Flight Mass Spectrometer (EESI-TOF-MS). We observed RH-dependent SOA chemical composition, including larger contribution of monomer products with increasing RH, although the total O:C remained largely unchanged. The effect of RH may be attributed to the particle-phase reactions of SOA components. This study highlights the necessity of characterizing SOA composition on molecular level and of considering RH dependence of SOA chemical composition and it physicochemical properties in atmospheric models.

Secondary organic aerosol (SOA), formed by oxidation of volatile organic compounds, significantly influence air quality and climate. Highly oxygenated organic molecules (HOM), particularly those formed from biogenic monoterpenes, play a key role in their formation and growth. As the most important daytime oxidant, hydroxyl radical (OH) initiated HOM from monoterpenes is believed to be mainly formed via the OH addition channel. However, for α-pinene, the most abundant atmospheric monoterpene, we found that the minor hydrogen abstraction channel is underappreciated in the HOM formation. We will present our observations and theoretical calculations, showing the role of hydrogen-abstraction and alkoxy radicals for fast autoxidation leading to HOM. We also provide a mechanism and yield, suggesting the non-negligible contribution of the hydrogen abstraction channel to ambient SOA, particularly in OH rich areas.

Dust aerosols exert an enormous influence on human health, global weather, and climate due to their large mass loading and wide spatial distribution. A large uncertainty still exists in quantitatively assessing the impact of dust aerosols on radiative transfer and climate simulation, which can partly be attributed to the complex refractive indices of dust. The complex refractive indices of dust are related to mineral composition. However, the mineral composition of dust is manifold and varies with dust origins, which results in sampling limitations in the experimental measurement of the complex refractive indices of dust and the contents of mineral components. In response to this problem, this study has obtained the complex refractive indices and the contents of the mineral composition of dust aerosols from satellite data by utilizing the mineral composition characteristics in the infrared band. The results showed that the retrieved complex refractive indices were higher than the experimental reference values, while the retrieved contents of mineral components were in good agreement with the experimental measurements.

In order to represent aerosols for radiative transfer simulations in the atmospheric model, we developed an internally-mixed aerosol scheme in which both particle nonsphericity and inhomogeneity effect were considered. In this scheme, black carbon was assumed to be fractal and soil dust was modeled as super-spheroid. These two insoluble aerosols (black carbon and soil dust) were partially coated with hygroscopic aerosols such as sulfate, nitrate, and aerosol water. The aerosol hygroscopic growth and composition variations were parameterized. Then, we built a database of the optical properties for the aforementioned models using the Invariant Imbedding T-Matrix Method(IITM). In addition, deep neural networks (DNN) were trained based on the database. We found that the DNN can predict the optical properties of aerosols efficiently and accurately. With these efforts, the aerosol-optics scheme was implemented into the Global/Regional Assimilation and Prediction System with Chinese Unified Atmospheric Chemistry Environment (GRAPES/CUACE). Two case studies were performed in East Asia on the summer and winter of 2020. Compared to the uniform and spherical aerosol scheme, the inhomogeneity and nonsphericity effect caused negative radiative forcing on the surface and positive radiative forcing on the top of atmosphere (TOA), enhancing the solar heating rate. The surface was cooler and the air beyond the boundary layer was warmer. Thus, the boundary layer was found to be more stable, leading to less precipitation and lower boundary layer heights. Finally, the meteorological and optical profiles computed from the GRAPES/CUACE were implemented in the RTTOV (Radiative Transfer for TOVS). The simulated solar reflectance on the TOA was compared with observations from the FY-4A satellite. We found that the comparisons were improved when the inhomogeneity and nonsphericity effect of aerosols were incorporated.

It is a challenge for improving the simulation effect of thick clouds over the Tibetan Plateau due to the rapidly and deeply cloud convection occurs frequently in summer. A cloud optical thickness assimilation system is developed by introducing Four-Dimensional Ensemble Kalman Filter theory (4D-LETKF) which establishing the covariance relationship between 2D cloud optical thickness and 3D cloud water content and taking advantage of the high spatial and temporal resolution of the next generation of geostationary satellites. A typical simulation case of convective clouds in summer is carried out over the Tibetan Plateau using the new assimilation system. The cloud optical thickness is assimilated hourly from the next generation data of Himawari-8 satellite to correct the cloud water mixing ratio in the WRF model. The assimilation significantly improves the simulation accuracy of cloud optical depth on thick clouds which solves the problem that the assimilation effect is weak on thick clouds in the existing cloud assimilation system. Using Himawari-8 satellite data as the true value for effect evaluation, the new assimilation scheme can improve the accuracy of cloud optical thickness correlation coefficient by 34% after the assimilation.

Cloud top pressure (CTP) is a critical cloud property that significantly affects cloud radiation. Multi-angle polarization-based sensors can employ polarized bands (490 nm) or O₂ A-bands (763 and 765 nm) to retrieve the CTP. However, the CTP retrieved by the two methods shows inconsistent results in certain cases, and large uncertainties in low and thin cloud retrievals may negatively affect subsequent applications. This study proposes a synergistic algorithm that appends both O2 A-bands and polarized bands using a random forest (RF) model. LiDAR CTP data are used as the sample true values, and polarized and non-polarized measurements are concatenated to train the RF model of CTP. Additionally, through analysis, we discovered that the polarized signal is saturated with increasing cloud optical thickness (COT), thus necessitating a special treatment for COT intervals <10 to improve the stability of the algorithm. We then applied the synergistic method to directional polarized camera (DPC) and Polarized and Directionality of the Earth’s Reflectance (POLDER) measurements for evaluation, and the validation accuracy and retrieval accuracy of the POLDER-based measurements (RMSE=121.431 hPa, R²=0.827 and RMSE=30.611 hPa, R=0.470, respectively) were higher than that of the MODIS and POLDER Rayleigh pressure measurements. Moreover, POLDER also showed good performance relative to DPC. This algorithm is expected to provide data support for atmosphere-related fields as an atmospheric remote sensing algorithm in the Cloud Application for Remote Sensing, Atmospheric Radiation and Updating Energy (CARE) platform.

The morphology of black carbon (BC) plays an important role in the determination of its aging scale and constraining its radiation effect. Therefore, it is urgent to characterize the distribution and evolution of observational-based morphology of BC-containing particles during atmospheric aging. In this study, a tandem differential mobility analyzer–single-particle soot photometer (DMA–SP2) system was used to estimate the size-resolved effective density (ρ_eff) of atmospheric BC-containing particles. The results show that there are differences in the morphology evolution with different BC core diameters (D_c) during the emitted-to-age process. With less coating materials, the small D_c particles (100-120 nm) tend to be spherical (ρ_eff = 1.31-1.33 g cm^-3) while the large D_c particles (280-300 nm) are fractal (ρ_eff = 0.95-1.08 g cm^-3). During atmospheric aging, the non-BC materials dominate the ideal growth of particle diameter for small D_c particles, while used to fill the void for large D_c particles and cause the collapse of the fractal branches. Furthermore, this study classified the BC-containing particles into two groups, i.e., near-spherical and non-spherical particles. For the first time, we calculated and parameterized the coating-to-BC mass ratio (M_R) of the shape transition points as a function of D_c and the mean aging state of BC-containing particles. After considering this parameterization, compared with the core-shell prediction, there are ~ 18% reductions for calculated absorption cross section over the entire population. This study highlights that the strong dependence of the ρ_eff and shape transition M_R point of atmospheric BC-containing particles on D_c can improve the representation of its morphology in the global climate model and reduce uncertainty in the evaluation of BC climate effects.

Real time precipitation is an important parameter in flood forecast. Geostationary satellites which have real time observations at visible and infrared bands with high spatial and temporal resolution provide a good way to estimate precipitation. A lots of precipitation estimation algorithms have been developed for geostationary satellites. Cloud top information which includes cloud top reflectivity, cloud top brightness temperature, gradient of cloud top brightness temperature and motion vectors of cloud top was usually used to retrieve precipitation in these precipitation estimation algorithms. However, it is hard to accurately select precipitation pixels from non-precipitation clouds only based on cloud top information. In this study, a new real-time precipitation estimation method for geostationary meteorological satellite Himawari-8 considering cloud microphysical parameters was proposed. Besides the cloud top information, the cloud microphysical parameters including cloud effective radius and cloud optical depth in both day and night were used in the new precipitation recognition method to improve the accuracy of precipitation identification. Based on the identified precipitation area, the cloud top brightness temperature and cloud liquid water path which was calculated from cloud effective radius and cloud optical depth were used to build a dynamic calibrating model to estimate the rainfall rate by taking the GPM-IMERG early run precipitation product as reference. The estimated rainfall rate was further calibrated using rain gauge observations on hourly scale. According to the validation using independent hourly rain gauges observations over Chinese mainland in summer, the probability of detection (POD) of the real-time estimated precipitation can reach up to 0.7, and the false alarm ratio (FAR) is lower than 0.6. From the aspect of precipitation identification, the accuracy of the newly developed precipitation algorithm has a obvious improvement compared with that of the GSMaP_NOW over Chinese mainland.

The intracloud (IC) to cloud-to-ground (CG) lightning ratio (IC:CG ratio) provides not only essential information for extrapolating IC flashes from the continuously monitored CG activities but also important implications for the kinematic and microphysical structures of thunderstorms. This study examines the IC:CG ratio climatologies in China and seeks the underlying reasons responsible for regional variations on IC:CG ratios. Analyses are based on both grid-scale (2 degree) and thunderstorm-scale (convective feature), using coincident observations from TRMM LIS (total flash) and the China Lightning Detection Network (CG lightning) during 2009-2014. National mean IC:CG ratios in China are about 3-5, consistent to those reported in other regions around the world. However, Guangxi Province (GXP) and Sichuan Basin (SCB) have exceptionally high (> 10) and lower-than-average (< 2) IC:CG values, respectively. While the IC:CG in GXP significantly maximizes in the spring season and diurnally peaks in the morning, SCB exhibits virtually no seasonal nor diurnal variation on IC:CG. The contrasting IC:CG ratios between GXP (~7.5) and SCB (~2.5) are also found on the storm scale. Thunderstorms in GXP have a much lower CG flash rates than those in SCB, even though they share a similar total flash rate, resulting in a far higher IC:CG in GXP. All convective proxies indicate that SCB has a higher likelihood in producing CG flash than GXP under the same convective intensity. In addition, the CG flash rate in SCB increases much more rapidly with the increase of total flash rate or convective intensity. The underlying mechanisms for regional variations on IC:CG may be associated with storm morphology, microphysical property, and/or convective environments, which will be further investigated.

In this study, based on the ERA5 dataset, the orographic precipitation characteristics during an extraordinary heavy rainfall event in Henan Province on July 19, 2016 are analyzed. Besides, by using WRF, a high-resolution numerical simulation is carried out, including a series of sensitivity experiments on the Taihang Mountains, the Yunmeng Mountain and low-altitude areas. The distribution of observed precipitation shows that the precipitation over the Taihang Mountains in northern Henan during the daytime of July 19 is significantly enhanced due to the topographic effect compared with that over the plain on the east. The simulation results demonstrate that the distribution of simulated precipitation over the steep terrains on the east side of the Taihang Mountains is closer to the observational precipitation when SRTM-90m is introduced to the WRF model. When the altitude of the Taihang Mountains is lowered by 50%, the accumulated precipitation along the Taihang Mountains in the daytime of the 19th decreases by 26%, accompanied with the northward movement of the heavy precipitation center over the northern Taihang Mountains. Besides, the Yunmeng Mountain plays an important role in blocking or diverting the easterly and southeasterly low-level jet. When the terrain elevation of the Yunmeng Mountain is reduced, the convergence zone of the wind field in the low-level moves to the areas along the Taihang Mountains, which leads to the expansion of the area with heavy rainfall, with the average accumulated precipitation in mountain areas increasing by 7%. Besides, When the low-altitude terrain in the study region is reduced by 50%, the center of heavy precipitation moves westward obviously. Combined with the diagnostic analysis of the topographical forced vertical velocity, the results show that the uplift of low-level jet caused by the low-altitude terrain is one of the reasons that affect the location of heavy rainfall center.

In order to enhance the understanding of the environmental field characteristics of the convective regeneration triggered by thunderstorm gust fronts. Based on Beijing sounding observation, automatic stations, S-band Doppler radar and new detection data to analysis environmental field characteristics of two typical cases in which convection initiation triggered by thunderstorm gust fronts. The results show that: (1) One type is convection triggered by collision of multiple gust fronts or the encounter of gust fronts with other boundary layer convergence lines in the region with high Convective Available Potential Energy (CAPE) and sufficient water vapor （collision triggering）. In this type, the vertical velocity of two collision gust fronts is much larger than that of a single gust front. Strong upward movement is easy to provide good lifting conditions for the triggering of convection, and convection is easy to trigger. (2)The other type is the gust front triggers convection in the unstable region (non-collision trigger). In this type, the low-level wind direction is perpendicular to the gust front, which is conducive to the strong convergence between the ambient wind near the ground and the gust front. Meanwhile, the updraft of the convection triggered by the gust front tends to be vertical, which is beneficial to convective initiation. (3)By comparing the environmental field of gust fronts trigger and non-trigger Convective Initiation (CI) cases, the results show that there are deep wet convection potential with high CAPE and low Convective Inhibition (CIN) of the environment field in these cases which CI are triggered by gust fronts. There are cold dry advection at the upper level and warm wet advection at the lower level in most cases CI triggered by the gust fronts, which also happened at the wind speed or wind direction convergence areas with high CAPE. 

The South China Sea frequently experiences boundary layer jet (BLJ) events, which are closely related to the coastal rainfall of South China during the pre-summer rainy season. The gust front associated with the cold pool can interact with the low-level vertical wind shear, potentially affecting convection. The BLJ with a unique vertical wind shear may also interact with the cold pool, but the nature of this interaction is yet to be fully understood. The present study uses data from 356-m high Shenzhen Met-Tower to analyze the statistical characteristics of cold pools and investigate how these characteristics vary between BLJ- and nonBLJ- related cold pools. A total of 54 cold pools were observed by Shenzhen Met-Tower from April to June during 2018-2020, which included 26 BLJ-related cold pools and 28 nonBLJ-related cold pools. Results showed that the BLJ-related cold pools had a weaker temperature drop and less pressure increase compared to the nonBLJ-related cold pools, but their structure were deeper, greater specific humidity decrease and recovery was faster. The BLJ-related cold pools caused the decreases of relative humidity in the upper levels but increases in the lower levels. In contrast, both the upper and lower layers of the nonBLJ-related cold pools showed an increase in relative humidity. It is found that BLJ-related cold pools moved faster and were affected by the type of convective systems. The BLJ-related cold pools associated with mesoscale convective systems are deeper, longer-lasting, and had less pressure increase and more specific humidity reduction than those associated with convective cells.

This work compares the characteristics of the first echoes of thunderstorms and nonthunderstorms retrieved from S-band polarimetric radar observations. Observations of 57 (39) isolated thunderstorm (nonthunderstorm) cells with roughly equivalent aerosol and water vapor conditions but different convective available potential energy were obtained with a S-band polarimetric radar and three independent lightning location systems during 2016/2017 in southern China. Storms with the first echoes were divided into three types based on echo top heights, namely, type 1 (below 0°C layer), type 2 (0°C to −10°C), and type 3 (above −10°C layer). Our observations show median values of radar reflectivity (ZH) and differential reflectivity (ZDR) of type 1 and type 2 in warm phase layer (below 0°C layer) are obviously greater in nonthunderstorms than in thunderstorms, but this feature is not significant in type 3 storms. In the mixed 1 phase layer (0°C to −10°C), median ZH in type 2 is greater in nonthunderstorms while median ZDR in type 3 is slightly smaller. In the mixed 2 phase layer (−10°C to −38°C), median ZH is greater in thunderstorms while median ZDR is smaller, and ZDR values in nonthunderstorms are closer to zero. Although results of ZDR comparisons in the mixed phase are likely affected by random errors and/or residual bias errors, these different signatures suggest different characteristics of liquid or ice particles between thunderstorms and nonthunderstorms. This study is expected to advance our understanding of physical processes responsible for the generation of the first flash.

Cold surge events are divided into four types depending on their correlation with North Pacific extratropical cyclones. Climatologic relationships between the two phenomena reveal that 92% (39%) of all cold surges are accompanied by extratropical cyclones (explosive extratropical cyclones), while 31% of extratropical cyclones are accompanied by cold surges. The occurrence and development of extratropical cyclones favor eruptions of cold air from higher latitudes, which in turn produce conditions conducive to cold surges. In the North Pacific, extratropical cyclones travel in a northeastward direction, ultimately contributing to the Aleutian low. Meanwhile, the westerly jet is observed to strengthen following cold surge events. Both actions drive further extratropical cyclone activity, which in turn facilitates subsequent cold surges. Therefore, when extratropical cyclones occur before and after cold surges, the cold surge event itself tends to be relatively strong and long-lived. The transmission of energy by extratropical cyclones is a primary link between the high and middle latitudes, and contributes to the impact of cold surges on low latitudes. This study also explores the respective influences of the Siberia High and Aleutian Low on cold surges.

Lightning affects significant hazards to society directly or indirectly, such as by striking individuals, causing damage to electrical infrastructure, and disrupting transportation systems. Accurate prediction of lightning is crucial for forecasters to respond effectively to its imminent occurrence. Lightning usually occurs in convective storms, which develop rapidly and occurs in a narrow area. This causes difficulties in predicting their location using numerical weather prediction (NWP). Lightning occurrence is often predicted by utilizing thermodynamic parameters (K index, CAPE, Lifted index, etc.). A high-resolution NWP model provides a prediction of thermodynamic variables at high spatiotemporal resolution with high accuracy for a few hours. However, a complicated algorithm is required to handle all the useful high-resolution variables from the NWP model. Recently emerging machine learning (ML) technique can solve this issue because it properly handles these “big data” without any model distributional assumption.
In this study, we have examined the capability of the use of “random forest” (RF) in the prediction of lightning probability for 3 years (2020–2022) of the warm season (June to August). The RF was trained by a lightning occurrence as a predictand and characteristic parameters from the NWP as predictors. The precision of RF was higher than 0.9 for lightning and higher than 0.85 for non-lightning, respectively. The trained RF was applied to the independent analysis and forecast fields and showed high agreement with the observation, suggesting that the trained model had the potential to assist forecasters in improving their lightning forecasting skills using real-time probabilistic forecasts from the trained model.

ACKNOWLEDGEMENT
This work was funded by the KMA Research and Development Program “Observing Severe Weather in Seoul Metropolitan Area and Developing Its Application Technology for Forecasts” under Grant (KMA2018-00125). This work was supported by the NRF of Korea grant funded by the Korea government (MSIT) (No. 2021R1A4A1032646).

It is known that the eastward moving Tropical Intraseasonal Oscillation (TISO) has impact on the climate over the Asian-Australian monsoon region. This study summarizes that the TISO can have the impact on the Taiwan climate in the onset of the Meiyu season and the summer and winter monsoon rainfall. For the onset of the Meiyu in Taiwan, it is found that for almost half of the cases, the first transition of the Asian summer monsoon can be classified as a sharp onset, which is occurred with an eastward-propagating TISO from eastern Africa and the western Indian Ocean to the Maritime Continent. This efficient and persistent transport of moisture presumably provides a favorable condition for the maintenance of the Meiyu front and marks the onset of the Taiwan Meiyu season. Following the Meiyu, during the summer, the TISO can be represented by the boreal summer intraseasonal oscillation (BSISO) better. Our results found that the BSISO and the associated horizontal patterns of the western North Pacific typhoon frequency are closely related. Taiwan has larger rainfall when the major BSISO convection moves northwestward from the Philippine Sea to the Taiwan area. The anomalous low-level cyclonic flow and the increased typhoon frequency directly result in the larger rainfall in Taiwan. The enhanced low-level southwesterly flow which transports the moisture to Taiwan is responsible for more summer rainfall in Taiwan. On the other hand, during the winter season, the rainfall in Taiwan is also related to the TISO. The results show that Taiwan has larger rainfall when TISO reaching the Indian Ocean and the western part of the Maritime Continent, and less rainfall when it moves to the western Pacific warm pool area. The mechanisms suggested by this work are: (1) Tropics to mid-latitude wave train and (2) Increase of moisture supply from South China Sea.

Droughts can cause enormous damage to agriculture, ecosystems and socio-economics because of their longevity and widespread spatial extent. As a densely populated region in the world and with a large area of arid and semi-arid regions, Asia is largely affected by droughts, especially in summer. The drought variations are determined by precipitation that affects water supply and temperature that affects water depletion through surface evaporation. In this talk, we will present an analysis of contributions of precipitation and temperature to the spatial patterns and temporal variations of summer drought over Asia measured by 3-month Standard Precipitation Evaporation Index (SPEI03) in July and the factors for year-to-year variations of summer drought. We will perform numerical experiments with the Community Earth System Model (CESM) to explore the influences of different regional SST anomalies on summer drought in Asia. Our results show that the July SPEI03 displays a north-south dipole distribution across approximately 30°N, which is dominated by interannual component. Precipitation and temperature both contribute to the summer drought variations over Asia. Asian drought in summer is associated with both the tropical Pacific and North Atlantic SST anomalies. Preceding tropical Pacific SST anomalies influence the summer Asian drought via the large-scale divergent fows and soil moisture-evaporation feedback. The North Atlantic SST anomalies influence the Asian drought through atmospheric wave trains over the North Atlantic-Eurasia. 

The changes in the spatial distribution of Indian summer monsoon (ISM) rainfall pose significant risks of extreme weather events in the Indian subcontinent and Arabian Peninsula. Monsoon intraseasonal oscillation (MISO) plays a key role in the intraseasonal distribution of the ISM rainfall, and changes in MISO determine the ISM rainfall patterns in the warming climate as well. However, the trend in MISO remains unclear in recent decades. Here, we find that the variance of MISO has an increasing trend from 1982 to 2017 over the northeastern Arabian Sea (AS), accompanied by increasing intraseasonal rainfall. The enhancement in rainfall is mainly nourished by the increasing moisture supply. Accordingly, the intraseasonal moisture budget diagnosis suggests that the changes in vertical moisture advection dominate the increasing amount of moisture over the northeastern AS. In contrast, the horizontal moisture advection and moisture processes associated with condensational heating exhibit a decreasing trend. By further decomposing the timescales, the upward advection of background moisture by intraseasonal vertical wind velocity is the main contributor. Given sufficient moisture availability, the changes in the intraseasonal vertical velocity are more important than the background moisture, which is a response to the strengthening of intraseasonal convergence in the PBL. Moreover, the faster warming of the AS and the increasing cyclonic vorticity associated with the poleward movement of the summer mean low-level jet are responsible for the increasing trend in PBL convergence. Thus, the MISO accompanying the intraseasonal rainfall displays an increasing trend over the northeastern AS. This recent intensified intraseasonal rainfall has contributed to the trend in the mean and extremes over the central-western parts of the Indian subcontinent. The results could assist the understanding and better projections of spatial pattern of rainfall during the ISM.

El Niño–Southern Oscillation (ENSO) and the Tibetan Plateau snow cover are important factors in interannual climate variability. The relationship between ENSO and the Tibetan Plateau snow variation is still undetermined. While some studies suggested that ENSO is a key factor of changes in snow cover over the Tibetan Plateau, other studies noted independence between the two. The present study revealed a prominent interdecadal change in the relationship between ENSO and the spring Tibetan Plateau snow-cover variation in the early 2000s. There is a significant positive correlation between ENSO and the spring Tibetan Plateau snow-cover variation in the period 1988–2003, but an obvious negative relationship is detected in the period 2004–19. The interdecadal change in the ENSO–snow relationship is related to the distinct pathway of ENSO influence on the spring Tibetan Plateau snow-cover variation during the two periods. In the period 1988–2003, ENSO induces anomalous convection over the tropical western North Pacific that in turn causes atmospheric circulation and moisture anomalies over the Tibetan Plateau. The resultant winter snow anomalies over the central-eastern Tibetan Plateau persist to the following spring. In the period 2004–19, ENSO induces North Atlantic sea surface temperature (SST) anomalies in winter that are maintained to the following spring. The North Atlantic SST anomalies then stimulate the atmospheric circulation anomalies extending to the Tibetan Plateau that induce snow-cover anomalies there in spring. The different processes of ENSO influence lead to opposite anomalies of spring snow cover over the Tibetan Plateau in the two periods.

The topographic effect of lands on monsoonal climate variability is one of the key topics in monsoon dynamics studies. Here, we re-visit the topographic effect of varying-shaped lands on Asian monsoon formation using a fast AGCM with a large number of experiments. It is found that the C-shaped continent located over the tropics and subtropics favors a distinct Asian summer monsoon pattern most. The preliminary results show that the C-shape land creates more evident monsoonal responses than other shape lands through three key processes, including intensified cross-equatorial winds, water vapor convergence along with subtropical monsoonal troughs, and one significant low-pressure departure near the southeast corner of the NH land. We would like to show more detailed results derived from our numerical experiments, which help to improve our understanding of monsoon formation and variability.

Water vapor flux in the lower troposphere controls the occurrence of heavy rain in the Bai-u season. The strong wind as well as enough humidity is essential to enhance the flux. We analyzed the remarkable south-westerly strong wind zones whose maxima are present below 900 hPa over the central East China Sea during the Bai-u season. Our definition of a “strong wind zone” is that the maximum wind speed within the layer below 900 hPa is greater than 15 ms^-1 and the wind speed decrease is larger than 3 ms^-1 from the 900-hPa level up to the 700-hPa level. The zones have a maximum occurrence rate from the end of June to the beginning of July over the central Eastern China Sea. We focused on the events in which the geostrophic component has remarkable shear above the wind speed peak below 900-hPa. Most of them are detected far from low-pressure systems and Bai-u front and are not directly included in them. We showed that warm air advected by the rather strong westerly inherent to this season over the heated Asian continent and cold air above the ocean made the steep zonal temperature gradient, which strengthens vertical shear through the thermal wind relation. The mechanism is similar to the coastal jet resulting from the land-sea thermal contrast (e.g. Tu et al. 2018) except for the large distance (several hundred kilometers) from the continent. The occurrence rate of the strong wind zone has a maximum (minimum) at 02LT (20LT); this diurnal cycle was synchronized with the strength of the southerly ageostrophic wind near the surface reported in some previous studies. The combination of the geostrophic component produced through thermal wind relation and ageostrophic one inherent to the East China Sea produces a very strong jet structure.

This paper examines the impact of Ural blocking (UB) on winter sub-seasonal Siberian cold anomalies under the different background conditions of the winter East Asian trough (EAT). It is found that the meridional displacement and the strength of the winter EAT possibly modulate the persistence or lifetime of UB and associated sub-seasonal cold anomalies over Siberia through changing background circulation conditions. Statistical analysis reveals that UB events tend to occur more frequently in strong or southward EAT winters than in weak or northward EAT winters. A composite analysis of UB events further shows that UB has larger amplitude and longer lifetime in strong or southward EAT winters (especially the latter) than in weak or northward EAT winters, thus more likely leading to stronger sub-seasonal cold anomalies over Siberia. Moreover, the possible physical cause of why a strong or southward EAT favors UB is further investigated. It is found that winter zonal wind and meridional gradient of potential vorticity (PV_y) tend to be weakened on the upstream side of Siberia near the Ural Mountains when the EAT is stronger or located more southward, favoring the maintenance of UB due to the strengthening (weakening) of the nonlinearity (energy dispersion) of UB. In contrast, when the EAT is weak or northward-displaced, a strengthening of winter zonal wind and PV_y is more likely seen over the Ural Mountains and adjacent regions, which may suppress UB through enhancing its energy dispersion and reducing its nonlinearity.

Global warming has contributed to the recent increase in heat waves, which have become more intense, frequent, and longer lasting. Furthermore, heat waves that last for a longer period of duration are more likely to cause socioeconomic damage. Therefore, to understand how heat waves are sustained, it is imperative to identify and predict their mechanisms. In this study, we categorized the summer heat waves in South Korea into two categories based on their duration: short-duration events (5-7 days) and long-duration events (≥16 days). Following that, we examined the associated atmospheric circulation patterns pertaining to the persistence of summer heat waves (June through August from 1973 to 2022) using local and synoptic atmospheric variables obtained from the JRA-55 dataset. During both events, South Korea’s upper and lower atmospheres are dominated by high-pressure anomalies, with weakened jet streams, sinking motion, and suppressed convection. In short-duration events, there is a predominant meridional wave propagation with low-pressure anomalies in the Russian region of Siberia. Compared to short-duration events, long-duration events are related to strong and zonally extended high-pressure anomalies in the southern region of the Kamchatka Peninsula, which have suppressed the movement of anticyclones, thereby prolonging the duration of heat waves. Long-duration of heat waves in South Korea may be linked with persistent anomalies in mid-latitude induced by changes in stationary waves with global warming. As a result of this study, we will be able to improve our understanding of how summer heat waves persist, as well as be able to diagnose and predict heat waves in the future.

In the winter of 2020/21, an extreme and long-lasting drought affected Southeast China (SEC), exerting significant social and economic impacts. In this study, the synergistic effects of La Niña and Tibetan Plateau (TP) warming on the drought are emphasized. Moisture budget analysis reveals that the drought was mainly attributed to the negative anomalies in horizontal and vertical moisture advections, namely the weakened atmospheric dynamic processes. The negative anomalies in horizontal moisture advection were previously claimed to be caused by La Niña events, as noticed by previous studies; however, the La Niña in 2020/21 presented as a cold tongue mode that induced more westward-extended Walker circulation and significant convection suppression over the western equatorial Pacific. Compared to the canonical La Niña events, warmer sea surface and stronger cyclonic anomaly occurred in the western North Pacific, inducing stronger northeasterlies and negative horizontal moisture advection anomalies over the SEC region. On the other hand, significant warming appeared over the TP in winter 2020/21, and it consequently enhanced the tropospheric meridional temperature gradient and further accelerated the westerly jet stream. The accelerated westerlies over the TP led to air accumulation and sinking motions in the downstream region, weakening the vertical moisture advection and being unfavorable for heavy precipitation in the SEC region. We highlight that, besides the La Niña event, the abnormal TP warming played a synergistic role in the extreme SEC drought in winter 2020/21. This multifactorial mechanism deepens our understanding of the drought formation and potentially improves the prediction of extremes in SEC.

The Western North Pacific (WNP) area is the main precipitation region with energetic typhoon activity in the boreal summer. A decreased precipitation and anomalous anticyclonic circulation in the WNP region are shown in the high-resolution atmospheric model (HiRAM) global warming simulation (RCP8.5_Ens). This study conducts a series of sensitivity experiments with various patterns of regional sea surface temperature warming (SST’spa) to explore the causes of the projected changes in the WNP in HiRAM. A slightly decreased precipitation and enhanced anticyclonic circulation in the WNP can be reproduced in experiments with 2-degree uniform warming, the spatial pattern of SST warming, warming over the Tropics, and the warming from greenhouse gases emission. More identifiable changes are shown in cases with Tropical Atlantic warming, the Tropical Indian Ocean, and the Pacific meridional mode-like pattern. These experiments, to some extent, reproduce the WNP response in HiRAM RCP8.5_Ens, indicating that most SST’spa can lead to decreased precipitation and enhanced WNP subtropical high. That might be because SST’spa reduces the spatial contrast of SST climatological distribution. On the contrary, the SST warming in WNP shows the opposite response to RCP8.5_Ens, revealing more precipitation and anomalous cyclonic circulation in the WNP. Contrasted responses are also detected in typhoon activity changes. The warming over the tropical eastern Pacific also generates an anomalous cyclonic circulation. These findings indicate that the forcing from remote SST warming has a more decisive influence in the WNP region than in situ warming, implying a high probability of decreased precipitation and enhanced subtropical high in WNP in the warmer future.

The El Niño–Southern Oscillation (ENSO) plays a critical role in predicting the winter surface temperature over East Asia. Numerous studies have attempted to improve the seasonal forecasting skill in view of the combined effects of ENSO and oceanic–tropospheric factors. However, high uncertainty and notable challenges still exist in using the ENSO to predict the surface temperature. Here, we showed that tropical stratospheric forcings (Quasi-Biennial Oscillation, QBO) could disrupt the response of the surface temperature to ENSO. The response of the East Asian surface temperature to El Niño/La Niña events evidently weakened in winter during the westerly/easterly phase of the QBO. This disruption has shown an increasing trend in recent decades, limiting the usefulness of ENSO alone as a seasonal predictor of the surface temperature. The modulation of the QBO on the East Asian surface temperature is achieved mainly by affecting subtropical zonal winds and North Pacific wave activity. Our analyses suggest that the QBO is a nonnegligible predictor in improving seasonal forecasts of the East Asian surface temperature.

This study analyzes the relationship between the Scandinavian (SCA) teleconnection pattern and summer precipitation over the eastern side of the Tibetan Plateau (ESTP) between 1960 and 2020 and further studies the underlying physical mechanism. Statistical analysis showed that there is a significant and high negative correlation coefficient (around −0.59) between ESTP summer precipitation and the SCA teleconnection pattern over the period 1960–2020. When the SCA teleconnection pattern is in its negative phase, there is a significant increase in precipitation over most of the ESTP, and vice versa. Moisture budget analysis showed that vertical moisture advection makes a larger contribution to precipitation changes related to the SCA teleconnection pattern than evaporation and horizontal moisture advection. Specifically, positive precipitation anomalies related to the SCA teleconnection pattern are dominated by the enhanced dynamic component of vertical moisture advection, which is induced mainly by the significant ascending motion over almost the whole of the ESTP region. In summer, under the effect of SCA teleconnection pattern, the abnormal westerly wind transports the mean warm air from TP to ESTP and forms an abnormal warm advection over the ESTP, resulting in a significant vertical upward motion and more summer precipitation over the ESTP. This study reveals the possible physical mechanism of the impacts of SCA teleconnection pattern on summer precipitation over the ESTP region and provides a scientific basis for ecological protection and disaster mitigation measures associated with major projects such as the Sichuan–Tibet railway and the Chengdu–Chongqing Economic Circle.

On August 8th, 2022, Seoul experienced unprecedented rainfall with a record-breaking total of 381.5 mm/day, the highest amount in 115 years. Heavy rain is a prevalent and dangerous weather phenomenon in Korea, leading to many investigations and studies on its mechanism of occurrence. However, In the Summer of 2022, the pressure system surrounding the Korean Peninsula exhibited an atypical summer pressure pattern. Therefore, this study analyzed the specifics of the pressure system in the Korean Peninsula and the Eurasian continent to uncover the factors and mechanisms behind the heavy rain in August 2022. At the surface level, high pressure increased significantly near the Ural Mountains and Lake Baikal in Siberia during the beginning of August. In terms of temperature, the analysis focused on the advection of cold air with a temperature anomaly of below -5°C from high latitudes to the northwest of the Korean Peninsula via Siberia along the flow of the high-pressure system in the lower atmosphere. And at 500 hPa, blocking was evident in the Ural. The blocking over the Eurasian continent hindered the zonal flow and amplified the meridional flow, resulting in the transport of cold air from high latitudes to East Asia through the strengthened meridional flow. The movement of cold air towards East Asia encountered the rim of the North Pacific high pressure and established a stationary front, leading to heavy rainfall. This study defined the development of a summertime cold continental high pressure affecting the mid-latitude region as a "Summer Cold Wave" (SCW) and the formation of the front caused by the SCW as a "Summer Cold Front" (SCF). Furthermore, by examining previous instances of SCW, it was discovered that the formation of Ural blocking in summer played a role in the emergence of SCW.

This study evaluates the performance of the high-resolution atmosphere general circulation model (AGCM), namely the Community Atmospheric Model (CAM) in its representation of seasonal mean, interannual variability and global teleconnections over the south Asian summer monsoon region (SAM), Western North Pacific (WNP), Equatorial Indian Ocean (EIO), Maritime Continent (MC), and West & East Pacific Ocean basin (W-PAC & E-PAC). Based on the AGCM experiment conducted for CAM model with observed SST (similar to AMIP type experiment), the model’s skill in simulating summer monsoon rainfall over different domains (SAM, WNP, EIO, MC, W-PAC & E-PAC) is low for the climate period (1981–2008). The decreased skill of the model over the south Asian monsoon region is evident, when the model is compared with the tropical mid-Pacific regions (W-PAC & E-PAC) and MC, which indicates that the impact of atmospheric dynamics is more important than SST forcing alone over the Asian monsoon regions (SAM, WNP & EIO). Our result shows that the AGCM simulation over south Asia exhibit large uncertainty in capturing the anomalies associated with the large global forcing like El Niño and La Niña. The analysis clearly brings out the presence of large systematic biases in the model simulation. The lack of air-sea interaction processes limits the skill of the model forced with SST alone over the Asian summer monsoon regions (SAM, WNP & EIO). The improved skill over the tropical mid-Pacific regions (W-PAC & E-PAC) and MC may be attributed to the strong SST forcing over the Pacific and the better representation of upward and downward limb of walker cells. While reduced skills over the Asian summer monsoon regions (SAM, WNP, EIO) may be attributed to the lack of Air-Sea Interaction and lack of dynamical feed backs from the local effects.

During the winter of 2022/23, East Asia experienced extreme temperature fluctuations on both inter- and intra-monthly timescales. November 2022 was warmer than normal, while December was colder than normal, with the second-highest difference in monthly temperature anomaly between the two months over the past 44 years. This regime shift in monthly time-scale can be explained by the alteration of month-long dominance of warmer air from south and cold air from north-west. The long-lasting La Niña condition over three years contributed to near record-high sea surface temperatures in the western Pacific region, affecting the warm East Asian climate until November. In December, the WACE pattern emerged over Eurasia due to the amplified warming especially in the Barents-Kara Sea causing the East Asian region to be cold climate. In January 2023, East Asia experienced abrupt regime shift from warm to cold climate within in just a few days. This event was unprecedentedly largest temperature fluctuations, dropping 12.34℃ in 12 days, observed on an intra-monthly timescale during the winter. In early to mid-January, warm and moist air intruded into East Asia from the south driven by expansion of the North Pacific High leading to above normal warm climate in East Asia. Later in mid-January, blocking occurred in the Ural region, and the meandering of the jet stream caused cold air to flow into East Asia from the north, resulting in a cold wave at the end of January. The extreme temperature fluctuations in East Asia this winter on both inter- and intra-monthly timescales are attributed to the tug-of-war between tropical and polar air masses manifested by such as WACE patterns due to arctic amplification and equatorial warm and moist air intrusion partly explained by the long-lasting La Niña condition which contributes to the intermittent expansion of North Pacific warm High.

The frequent incidence of extreme precipitation in India during the Indian Summer Monsoon (ISM) season has resulted in various devastating consequences such as flash flooding, crop destruction, loss of life, and damage to infrastructure. To mitigate the potential effects of extreme rainfall, it is essential to understand its impacts across the country. In this study a Maxstable process-based model has been used to model the Spatio-temporal variability of ISM rainfall extremes and the risk associated with various return levels. In addition to the spatial factors, the impact of large-scale climatic drivers, such as the El Niño Southern Oscillation (ENSO), has also been incorporated to demonstrate its influence on the extreme precipitation field. Due to the complexity of terrain across the country, spatial clustering has been implemented. The computational efficiency was substantially improved by the implementation of parallel computing on clusters. Twenty models with different combination of spatial and temporal covariates for location, scale and shape parameters have been selected and the best fit model was identified based on Takeuchi Information Criterion (TIC) values. Results indicates that the regions have noticeable spatial variabilities extreme precipitation and around 75% of the land area in India is under the risk of heavy rainfall for more than 50-year return period. Compared to the spatial extreme models, inclusion of climatic oscillations information (ENSO) in a model framework greatly improves the suitability of the model. The results of this investigation hold significant scientific and practical implications for monitoring and managing the risks associated with extreme rainfall. They can aid in the development of appropriate mitigation strategies and inform planning efforts for potential adverse effects caused by such extreme weather events.

The scientific explanation of speleothem δ¹⁸O in the East Asian summer monsoon (EASM) domain is a crucial issue restricting stalagmite-inferred paleoclimate research. Long-term cave monitoring is an effective solution to deal with this issue. Here, we investigate the transfer mechanism of oxygen isotopic signals from rainfall to dripwater and modern calcite, using 11-year-long cave monitoring data from Yuhua Cave, southeast China, located in the frontal zone affected by the EASM. The δ¹⁸O in the precipitation (δ¹⁸O_p) shows a high response to EASM activities associated with the intertropical convergence zone (ITCZ) migrations and the activities of West Pacific Subtropical High (WPSH) affected by El Niño Southern Oscillation (ENSO) activities. Upstream convection and rainout process play a crucial role during the transmission of the ENSO signal to the δ¹⁸O_p in southeastern China. Affected by the mixing effect of the epikarst reservoir, the amplitude of δ¹⁸O in dripwater (δ¹⁸O_d) is much smaller than δ¹⁸O_p, and there is no obvious seasonal variation. The δ¹⁸O_d reflected ENSO-related δ¹⁸O_p during monitoring periods, with greater amplitude. The modern calcite δ¹⁸O (δ¹⁸O_c) series show significant seasonal variations controlled by the fractionation coefficient varied with cave temperature. At the interannual scale, the δ¹⁸O_c series can inherit the δ¹⁸O_d ENSO-related variations. Our results suggest that high‐resolution stalagmite δ¹⁸O reconstructions from Yuhua Cave could characterize past ENSO‐related variability in the EASM on the annual-decade scale.

The difficulty in constraining the large-scale Asian summer monsoon (ASM) variability in the Chinese monsoon region (CMR) during Termination II lies in our poor knowledge of the western part of the CMR largely due to the sparsity of the paleoclimate records. To get a better picture of the ASM during Termination II, we examined a precisely dated stalagmite δ18O record between 133.1 and 127.0 kyr B.P. from Wanxiang Cave in western China. In combination with published δ18O data from this cave, we have identified the ‘Weak Monsoon Interval’ (WMI) in the Wanxiang δ18O record and confirmed that the severe cold event in the North Atlantic caused the weakened ASM over the CMR via reorganization of the large-scale ocean-atmospheric circulation. However, the post-WMI change in δ18O is gradual, in contrast with the abrupt shift shown in the other cave records. The rapid northward migration of the westerly jet relative to the Qinghai-Tibet Plateau is probably responsible for such discrepancy. This northward migration prevents the more 18O-depleted moisture from reaching the study site and also causes the earlier northward movement of the East Asian summer monsoon (EASM) rainband that carries positive precipitation δ18O to obscure the possible abrupt decrease in our δ18O record. After the last interglacial onset, no obvious Younger Dryas (YD)-like event was recorded in Wanxiang Cave, which suggests a minimal impact of the YD-like event on the ASM variabilities. The relatively large amplitude of δ18O variation observed in Wanxiang Cave between the late penultimate glacial and the last interglacial corresponds to a dominant control of the Indian summer monsoon (ISM), whereas smaller δ18O amplitudes were recorded in those cave sites mainly under the influence of both ISM and EASM. Therefore, the hydroclimate change over the CMR during Termination II has resulted from a combination of multiple processes.

The hydro-meteorological climate of the Himalayas is the lifeline for the most densely populated region in South Asia. Annual floods in the Himalayan rivers often occur during the monsoon season, affecting millions of people in the Himalayas and downstream regions. The Himalayan region is extremely vulnerable to the ramifications of extreme precipitation events (EPEs), such as flash floods, landslides, and agricultural and infrastructural damages during the Indian summer monsoon (ISM). Complex terrain, high meteorological diversity and uncertainty in observations over this region make it challenging to comprehend the precipitation disparities and predict the EPEs across the Western Himalayas (WH). Therefore, a better representation of EPEs characteristics over the WH using high-resolution data is crucial for precisely understanding the precipitation variability and mechanisms of climate-triggered localised natural disasters. This study investigates EPEs using Asia's first highest resolution (10 km), a regional atmospheric reanalysis, High Asia Refined analysis version 2 (HAR v2), during ISM. HAR v2 precipitation dataset has been generated by dynamically downscaling global ERA5 reanalysis data using Weather Research and Forecasting model (WRF). Majorly, heavy rainfall over the region during the ISM occurs by convection followed by orographically locked system. The intricate interaction of regional topography with moist airflow and temperature gradient accentuates the EPEs. The present study will investigate the dynamic and thermodynamic processes associated with EPEs over the study region. Overall, this study aims to provide scientific insights to investigate the potential impacts of climate change on extreme events, which in turn could help mitigate disasters in the Himalayan region. Detailed results of precipitation variability over the Himalayas, and mechanisms altering the atmospheric conditions attributed to the EPEs will be discussed.

Proxy and model-based studies suggest multi-scale temporal variability in the Indian summer monsoon (ISM). In this study, using the CESM1 atmospheric general circulation model, we carried out multiple ensemble AGCM simulations for the Mid-Holocene (MH; ≈ 6 kyr BP), Medieval Warm Period (MWP; ≈ 1 kyr BP), Little Ice Age (LIA; ≈ 0.35 kyr BP), and Historical (HS; ≈ CE 2000) periods. We used the PMIP3/CMIP5 boundary conditions for this purpose. Our simulations indicate that the ISM during the MH was stronger compared to HS and the rainfall higher, in agreement with several proxy studies. The experiments also suggest that the ISM rainfall (ISMR) was higher during MWP relative to the LIA in agreement with our earlier results from the PMIP3 models. A relatively northward migration of the ITCZ over the Indian region and strengthening of the neighboring subtropical high over the northwestern Pacific, both associated with stronger insolation associated with the obliquity and precision during the MH, seem to be important reason Indian summer monsoon during the MH.

Most future projections conducted with coupled general circulation models simulate a non-uniform Indian Ocean warming, with warming hotspots occurring in the Arabian Sea (AS) and the southeastern Indian Ocean (SEIO). Although the corresponding spatial temperature gradients are associated with large-scale atmospheric circulation anomalies and rainfall trends with far-reaching societal implications, little is known about the underlying physical drivers. Here, we analyze a suite of large ensemble simulations conducted with the Community Earth System Model 2 to elucidate the causes of non-uniform Indian Ocean warming. Strong negative climatological air-sea interactions in the Eastern Indian Ocean are responsible for a future weakening of the zonal equatorial sea surface temperature gradient, resulting in a slowdown of the Indian Ocean Walker circulation and the generation of southeasterly wind anomalies over the AS. These contribute to anomalous northward ocean heat transport, reduced evaporative cooling, a weakening in upper ocean vertical mixing and an enhanced AS future warming. In contrast, the projected warming in the SEIO is related to a reduction of low-cloud cover and an associated increase in shortwave radiation. Therefore, the regional character of air-sea interactions plays a key role in promoting future large-scale tropical atmospheric circulation anomalies with implications for society and ecosystems far outside the Indian Ocean realm.

In recent years, extreme weather events have increased and in this work, we explore whether volcanic eruptions- subaerial, submarine or mixed, may have a role in or trigger, any of these extreme events. Volcanic materials (sulphur dioxide, water vapour and aerosols) released by volcanic eruptions can be carried by the wind and transported to a long distance and can affect the regions far away from the eruption sites. Based on location, the strength of eruptions and seasonal timings, volcanos have the potential to influence and alter atmospheric and oceanic circulation too. Hence, the impact of each volcanic eruption is different. Apart from circulation fields, related mechanisms that can modulate rainfall are the reduction of solar radiation, acting as condensation nuclei and complex interaction involving clouds among others. Strong volcanic eruptions can even alter ozone distribution in the stratosphere and hence can influence stratosphere-troposphere coupling. This study using observation from satellite data and rainfall records from Hong Kong Observatory revealed that volcanic eruptions have contributed to several extreme rainfall records in Hong Kong since 1963. The 1963 Agung eruption and the 1991 Pinatubo eruption can be linked to the driest year and 11th driest year of Hong Kong respectively. On the other hand, the 1982 El Chichón eruption and the 2008 Chaitén eruption could be linked to much excess precipitation in Hong Kong, the second and sixth wettest years on record respectively. Analyses of the recent strong submarine volcanic eruption Hunga Tonga in 2021 (December) that reached its peak in 2022 January is also explored in connection to the recent extreme monsoon rainfall in some east Asian countries. Such analyses and in-depth understandings of the impact of volcanic eruptions on climate will not only provide a useful direction for volcanic risk assessment but also lead to improved regional climate prediction.

An extremely strong and long-lasting (more than 8 months) oceanic warm core eddy existed in the South China Sea (SCS) from February–October 2010. From July–August 2010, three tropical cyclones (TCs; TC Conson, Chanthu, and Mindulle) consecutively passed over this eddy and sustained at least 21 h. The intensity change of all three TCs reached 20 kt within 24 h when they encountered this eddy. In mid-late July, tropical cyclone heat potential (TCHP) is overall stronger in the eddy region than in its surrounding region, thus TCHP plays an important role in the intensification of TC Conson and Chanthu. It is also found that the intraseasonal oscillation (ISO) and the quasi-biweekly oscillation (QBWO) can be important in favor of the further enhancement of TCs. The TCHP is too low to favor the intensity increase of TC Mindulle in late August, 2010, but weak vertical wind shear, ISO and QBWO act as key roles in the intensification of TC Mindulle.

Most strong typhoons undergo rapid intensification (RI), and it is difficult to predict them. In that sense, the analysis was conducted to understand better the trends and characteristics of the rapid intensification of tropical cyclones (RITCs) that can cause significant damage. Although it is difficult to explicitly define the rapid intensification of typhoons, generally defined as a period during which the intensification rate is greater than 15.4 m/s within 24 hours. This study used JTWC best-track data from 1982 to 2020 for the northwest Pacific. As a first result, the time series of RITCs ratio and LMI intensity shows a clear trend of increase in the autumn TC period than in the summer. In addition, the longitude of occurrence and strongest point of tropical cyclones also showed a tendency to move westward, which was evident in the autumn, especially in September. Secondly, an analysis was conducted to find the reason for the tendency of occurrence and the LMI location to be biased toward the west. When the correlations between RI genesis and LMI longitude and the various oscillation indices in the autumn period were calculated, the possibility that PDO was most likely to have an effect was found. In addition, values of sea surface temperature and tropical cyclone heat potential increased in the recent RI increase zone during the autumn TC period. This may have influenced the recent increase in the RI ratio, and further analysis of several synoptic variables will be conducted. 

Fine particles pollution associated with high sulfate and secondary organic aerosol (SOA) still often occur in winter, despite efficient pollutants mitigation has been achieved in China. The fast SO₂ oxidation mechanisms revealed recently are controversial but consistently recognized associating with aerosol water. Here we report that both multiphase oxidation of SO₂ by N(Ⅲ) involving aged soot-containing/iron-rich particles, and aqueous hydroxymethanesulfonate formation dominated high sulfate yield in wet particles with pH of 4.0~4.5 under relative humidity >85% in wintertime of northern China. We also provide a new evidence in terms of single particles for enhanced SOA generated from combustion-derived primary organic aerosol (POA) under high humidity. High humidity benefits the water uptake of particles, and aerosol high water content buffers aerosol acidity, favorites sulfate formation and the aging of soot-containing particles, particularly facilitates the conversion of combustion-derived POA to SOA. This study highlights a co-reduction on reactive gases and particles those directly emitted from solid fuels combustion for mitigating sulfate and SOA under high humidity.

Organic vapors from biomass burning are a large source of secondary organic aerosol (SOA). Previous smog chamber studies found that the main SOA contributors in biomass burning emissions are volatile organic compounds (VOCs). Intermediate volatility organic compounds (IVOCs), thought to be efficient SOA precursors, are a considerable fraction of biomass emissions, but their contribution to SOA formation has not been directly observed. Here, by deploying a newly-developed oxidation flow reactor to study SOA formation from wood burning, we find that IVOCs can contribute ~70% of the formed SOA, i.e., >2 times more than VOCs. This previously missing SOA fraction is interpreted to be due to the high wall losses of semi-volatile oxidation products of IVOCs in smog chambers. The finding in this study reveals that SOA production from biomass burning is more than 3 times higher compared to previous studies, and highlights the urgent need for more research on the IVOCs from biomass burning and potentially other emission sources. In addition, by applying source apportionment and clustering methods, we are able to track the chemical evolution of SOA molecules. The results provide insights into the multi-generation chemistry when biomass burning emissions are transported in the atmosphere.

Cooking emissions represent a major source of air pollution in the indoor environment and exhibit adverse health effects caused by particulate matter together with volatile organic compounds (VOCs). A multitude of unknown compounds are released during cooking, some of which play important roles as precursors of more hazardous secondary organic aerosols in indoor air. Here, we applied secondary electrospray ionization high-resolution mass spectrometry for real-time measurements of VOCs and particles from cooking peanut oil in the presence of 300 ppbv nitrogen oxides (NOx) generated by a gas stove in an indoor environment. More than 600 compounds have been found during and after cooking, including N-heterocyclic compounds, Oheterocyclic compounds, aldehydes, fatty acids, and oxidation products. Approximately 200 compounds appeared after cooking and were hence secondarily formed products. The most abundant compound was 9-oxononanoic acid (C9H16O3), which is likely the product formed during the heterogeneous hydroxyl (OH) radical oxidation of oleic acid (C18H34O2) or linoleic acid (C18H32O2). Real-time detection of an important number of organic compounds in indoor air poses a challenge to indoor air quality and models, which do not account for this extremely large range of compounds.

It is necessary to convert fire radiant energy (FRE) to biomass burning consumption by the combustion conversion coefficient, when adopting the satellite FRE to estimate the open biomass burning emission. To date, the combustion conversion coefficients are considered to be scalar constants, which ignores the influence of multiple factors during the combustion process and regional differences. This project aims to conduct a systematic analysis of combustion conversion coefficients, by the view of ground experiment measurement, satellite data verification and emission inventory improvement. Firstly, the combustion conversion coefficients of typical biomass sources in China are obtained and the effects of fuels type, combustion state, combustion scale and combustion temperature on the combustion conversion coefficient are revealed, based on the ground experiment. Then, the satellite data are used to estimate the regional combustion conversion coefficient, so as to verify the reliability of the ground experimental measurement results on the regional scale. Finally, the verified combustion conversion coefficients are applied to improve the accuracy of the emission inventory. The uncertainty of the emission inventory caused by the combustion conversion coefficient has been estimated. The research results provide a new parametric scheme for correcting the uncertainty caused by the combustion conversion coefficient and improving the accuracy of the open biomass burning emission inventory in China. 

Household stoves are widely used in developing regions for cooking and heating purposes, but they also represent a significant primary emission source of gaseous pollutants. To investigate their impact on air pollution, we collected samples of ammonia (NH₃) and nitrous acid (HONO), two important reactive nitrogenous gases, from various sources including the flue gas of typical household stoves. We measured their stable nitrogen and oxygen isotopic signatures (δ¹⁵N and δ¹⁸O) to identify their sources. In a year-long case study, we found that during autumn and winter, the stable isotopic compositions of atmospheric NH₃ and HONO closely matched those of stove flue gas. This suggests that household stoves play an important role in the local budgets of NH₃ and HONO. These findings underscore the need for targeted interventions to address the use of household stoves in developing regions as part of broader efforts to reduce air pollution and its associated health impacts.

We investigate the effect of forest fires on aerosol and gaze composition of the stratosphere, using the interferometric measurements of MIPAS instrument, that monitored the atmosphere from to 2002 to 2012 from ENVISAT platform. Two events are analyzed: 2009 Australian Black Saturday fires and 2003 Siberian Taiga fires. The impact on the composition of CO, aerosols, H2O, ozone and temperature are analyzed and the 3-dimensional plume trajectory is reconstructed, using the data of MIPAS scientific Processor run at IMK-ASF/IAA.

As is well known, water vapor is an important greenhouse gas in the atmosphere that plays a critical role in atmospheric dynamics process. It is closely related to weather and climate phenomena that affects our daily life. Therefore, precisely determining, mapping and monitoring water vapor contents has been a vital research concentration in deepening our understanding of atmosphere and our earth’s environment. Precipitable water vapor (PWV), which measures the contents of the water vapor in the atmosphere, is an essential parameter in climate and meteorological research. Among the PWV sensing methods, Global Navigation Satellite Systems (GNSS) is characterized by its remarkable accuracy, super capacity to operate under diverse weather conditions, and high ability to deliver real-time products. This contribution focuses on the algorithm, software development, and accuracy evaluation of the real-time PWV retrievals based on GNSS precise point positioning (PPP) algorithm. First, some advanced models were developed to determine high-accuracy zenith hydrostatic delay (ZHD) and weighted mean temperature (Tm). Then, an enhanced multi-GNSS real-time PPP software was developed to calculate the zenith tropospheric delay (ZTD) over the GNSS station. Finally, an exhaustive analysis was undertaken to assess the accuracy of real-time PWVs derived from the PPP-based ZTDs. The high-accuracy real-time PWV obtained is promising for time-critical meteorological applications. Additionally, our 20-years efforts towards GNSS meteorology were also summarized in this contribution.

Using multisource data including 5-yr dual-polarization radar observations, convective and microphysical characteristics of extreme precipitation features (EPFs) over a monsoon coastal region in South China are investigated including dependence on rainfall extremity and subseasonal variations. The EPFs are sorted into three groups according to the extreme rainfall intensity: 84 – 126 mm hr^-1 (ER1), 126 – 186 mm hr^-1 (ER2), ≥186 mm hr^-1 (ER3). The more extreme rainfall shows a notable increase and decrease in the fractions of “intense” convection (7.6%, 20.6%, 31.6%) and “weak” convection (41.3%, 22.9%, 18.9%), respectively, while that of the “moderate” convection remain about 50%. The higher rainfall extremity is accompanied by statistically significant increases in ice and liquid water contents and a slight decrease in the fraction of coalescence in liquid-phase processes. While the raindrop size distributions (RSDs) of ER1 to ER3 similarly feature a mean size larger than “maritime-like” droplets and a concentration much higher than “continental-like” raindrops, the mean size and concentration of raindrops tend to increase with the increasing rainfall extremity. During the pre-monsoon period, precipitation systems are the largest in area but their EPFs are the least frequent and have the lowest raindrop concentration, likely due to the colder, drier environment with large vertical wind shear (VWS). Onset of the summer monsoon increases the frequency and convective intensity of EPFs, leading to an increase in raindrop size, consistent with the substantial increases of CAPE and moisture during the active-monsoon period. EPFs share similar convective intensity and RSD between the post-monsoon and active-monsoon periods, although the post-monsoon EPFs are slightly less frequent and have a smaller horizontal scale related to the reduced 0–6-km VWS. EPFs associated with tropical cyclones have the weakest convective intensity but the most active warm-rain processes with the RSD being closer to the maritime regime.

This study reveals that the significant increase of winter precipitation over the southeastern United States (SEUS) is associated with El Niño and negative phase of North Atlantic Oscillation (NAO^-). Diagnosis of large-scale dynamics shows that El Niño and NAO^- have a synergistic effect on the enhancement of transient eddies and stationary waves in the eastern Pacific, southern United States, and North Atlantic. These enhanced transient eddies are associated with subtropical jet stream acceleration and maintenance of subtropical low from the eastern Pacific to the Atlantic, influencing stronger stationary waves propagation from tropical Pacific to the SEUS and North Atlantic, and from the northwestern Atlantic to the SEUS and Pacific. This favors a positive phase of meridional dipole in geopotential height anomalies between the tropics and subtropics in the Western Hemisphere (WTSD) during the co-occurrence of El Niño and NAO^-. The strong positive WTSD-like pattern, accompanied by a zonally extended and southerly shifted subtropical westerly jet, induces a northward tilted secondary circulation with ascending motion over the SEUS and Gulf of Mexico. Simultaneously, intensified trough over the SEUS promotes moisture and warm advection from the subtropical and tropical Pacific converging with cold advection from the northwestern North Atlantic in the adjacent Gulf of Mexico, which is beneficial for front and cyclone generation, and induces heavy precipitation in the SEUS. This study suggests that the synergistic effects of El Niño and NAO^- help to understand the variability of winter SEUS precipitation.

The southeastern extension of the Tibetan Plateau (SETP) is distributed by the typical longitudinal mountains and has unique climate characteristics and significant regional differences. The diurnal cycles of rainfall amount, frequency, and intensity over the SETP in the warm seasons (May to September) were investigated using high-density hourly station rainfall data and the possible mechanisms were discussed by analyzing the ERA5 reanalysis and satellite data. The largest amounts of rainfall appeared on the southern and western margins of the SETP. Located at the southern margins of the SETP, the adjacent western (Reg_W) and eastern (Reg_E) regions presented a similar rainfall amount, frequency, and intensity, but the diurnal features of the two regions were quite different. The rainfall amount, frequency, and intensity had dominant peaks in the late afternoon in Reg_W, while in the early morning in Reg_E. A secondary peak in the early morning can be found in Reg_W. The diurnal features of the two regions were closely related with the low-level atmospheric conditions and the distribution of clouds. The enhanced convergence of the anomalous southerly winds and sufficient water vapor after midnight contributed to the nocturnal rainfall at the southern edge of the SETP. In the afternoon, the lower surface air temperature and more stable stratification in Reg_E, as compared with Reg_W, were relatively unfavorable for the occurrence of afternoon convection.

The Weather Research and Forecasting (WRF) Double-Moment 6-class (WDM6) microphysics scheme only predicts the number concentrations for liquid-phase hydrometeors. Even though Park and Lim (2023) recently revised the WDM6 scheme by implementing the prognostic number concentration of cloud ice, the precipitating solid-phase hydrometeors such as snow and graupel are still treated as the single-moment approach, in which only mixing ratio is prognosed. In this study, the new version of WDM6 microphysics scheme is introduced by adding the prognostic number concentrations of snow and graupel to the WDM6 (Park and Lim, 2023). Therefore, the new WDM6 scheme predicts the number concentrations of all hydrometeors. By introducing the double-moment approach for snow and graupel, several microphysical processes related to them are modified and added. The newly added processes include the processes of accretion of rain by snow, self-collection of snow, sublimation of snow and graupel, and evaporation of snow and graupel. The new scheme has been tested for four summer-precipitating (Kim et al., 2019) and seven winter-precipitating convection cases (Ko et al., 2022). The mixing ratio of snow tends to increase and that of graupel decreases in the new scheme. Both mixing ratio and number concentration of rain reduces. The new scheme improves the equitable threat score (ETS) and false alarm ratio (FAR) for eight cases and probability of detection (POD) for seven cases among 11 total cases. *This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government. (MSIT) (RS-2023-00208394)

The Weather Research and Forecasting (WRF)-Double-Moment 6-class (WDM6) microphysics scheme treats the solid-phase hydrometeors as the single-moment approach. The scheme only predicts the mixing ratio of graupel, snow and cloud ice. In addition, each category of solid-phase hydrometeors is characterized by static parameters defining pre-defining density, diameter-mass relationship, and diameter-fall speed relationship. Previous studies have reported a considerable sensitivity of simulated precipitation systems according to these static parameters. This study revises the ice microphysics in the WDM6 scheme through the implementation of prognostic graupel density and cloud ice number concentration. The new version of WDM6 has been tested in the idealized squall line and winter-time snowfall cases. the simulations with the new WDM6 show better statistical skill scores such as Root Mean Square Error (RMSE) and Equitable Treat Score (ETS), relative to the ones with the original WDM6. It is also confirmed that the new WDM6 simulates a similar fall velocity-density relationship of graupel to the observation. More detailed analysis will be presented in the conference.* This work has been supported by the South Korean Ministry of Science and ICT (MSIT) and the National Research Foundation of Korea (NRF) (grant no.2021R1A4A1032646).

In this study, the convection-permitting model (CPM) with 3-km horizontal resolution has been used to investigate the simulated ability on diurnal cycles of pre-summer rainfall over southern China with various subregions and rainfall systems. Compared with the observations, the model captures the peak time of nocturnal rainfall in the western inland related to the eastward-propagating rainfall system over the eastern slope of the Yungui Plateau. The simulated eastward-propagating rainfall system moves a short distance and has a short lifetime due to the weaker westerly-steering flow and southerly winds in the late night bringing less warm and moist air with the weaker upward flow of the mountain–plain solenoid circulation that is unfavorable for convection instability. Over the eastern inland, the model shows good performance on simulating the quasi-stationary rainfall system in the morning because of the high resolution of topography. However, the model cannot simulate the climatology of the secondary morning rainfall peak due to overestimation of the rainfall from noon to evening. On the other hand, the weaker simulated southwesterly winds especially on the north side of the Nanling Mountains with decreased upward movement and moisture convergence and the more stable atmospheric condition lead to the disadvantage of convective rainfall. The CPM reproduces the afternoon rainfall peak but with a lead of 2–3 hr and still cannot simulate the morning peak well over the coastal region. The main source of mode bias is the over estimation of the onshore winds in the boundary layer over the land with the reduced temperature difference between land and sea resulting in insufficient convergence in the coastal region.

High-resolution observations show that the intraseasonal precipitation variance accounts for 20-50% and 40-70% of the total daily precipitation variance in the wintertime western and eastern North Pacific, respectively, implying the indispensable ingredient of the intraseasonal precipitation for the hydrological cycle over the East Asia-North Pacific-North American region. In this study, using daily observational data, characteristics and origins of the intraseasonal precipitation in the wintertime western and eastern North Pacific are investigated. The changes of the intraseasonal precipitation in the western and eastern North Pacific are featured by the anomalous precipitation in the Kuroshio-Oyashio Extension (KOE) and eastern North Pacific along 30°N, respectively. The intraseasonal precipitation in the western North Pacific is associated with the migratory weather systems with anomalous anticyclone originating from Japan and the cyclone originating from Northwest Asia, accompanied with the upward draft in the KOE region. The instability conditions are different in the north and south of the KOE. The unstable static and baroclinic conditions are both responsible for the enhanced precipitation in the south of the KOE. The anomalous precipitation is mainly determined by the atmospheric baroclinic instability in the north of the KOE. The anomalous atmospheric circulations associated with the intraseasonal precipitation in the eastern North Pacific are characterized by a dipolar structure, in which the anomalous quasi-stationary cyclone in the eastern North Pacific plays a major role. The upward draft associated with the anomalous cyclone and atmospheric static instability induced by the anomalous southeasterly wind provide dynamic and instability conditions for the anomalous precipitation. Furthermore, the moisture sources of the intraseasonal precipitation in western and eastern North Pacific are both dominated by the anomalous water vapor transport by the atmospheric circulation and the change of the precipitable water.

Scale-aware cumulus parameterizations and their precipitation forecast skill over South Korea are compared at different horizontal resolutions. A series of single-domain experiments are conducted at 3-, 9-, and 27-km spatial resolution with three different scale-aware cumulus parameterization schemes available in the Weather Research and Forecasting (WRF) Model, that is, the multiscale Kain-Fritsch scheme, the Grell-Freitas scheme, and a revised version of the simplified Arakawa-Schubert scheme developed by the KIAPS (KSAS). The results from the experiments are compared with each other and with those obtained with the cumulus parameterization scheme switched off and using conventional cumulus parameterization schemes with no scale-aware capability, particularly focusing on the contribution of parameterized convection, which is determined by the convective updraft fraction or scale-dependent parameter as a function of horizontal resolution in the scale-aware cumulus parameterization, and the short-range precipitation forecast skill over South Korea against automatic weather station rain gauge observations. Based on the results, the method of defining convective updraft fraction in the KSAS is revised to use a more physically-based method and its effect on the precipitation forecast skill at each spatial resolution is examined.

Soil liquid water (SLW) has an important impact on near-surface soil hydro-thermodynamics and surface energy budget. A new term about vertical movement of SLW (VMSLW) was added to the soil heat conservation equation in Noah-MP coupled with WRF, and the effects of the local and upstream VMSLW on a heavy summer rainfall event during June 15-25, 2010 in South China were investigated. Results show that when precipitation intensity is ≥ 2 mm/day, the magnitude of the VMSLW term in the soil heat conservation equation is comparable to that of the original term due to soil thermal conduction. Local VMSLW (contribution rate: 5.88%) leads to high pressure anomalies in South China, which weakens precipitation and reduces the positive bias of precipitation. The upstream VMSLW (contribution rate: 9.24%) decreases water vapor transport in Indo-China Peninsula, causing reduced precipitation in South China. The modified soil heat conservation equation with the new VMSLW term improves the precipitation simulation in the coupled WRF-Noah-MP model and deepens the understanding of interactions between SLW and precipitation. 

The surface heating in complex urban areas affects concentration of secondary inorganic aerosol (SIA) via flow and chemical environment changes. Previous studies on heating effects on pollutants have focused on effect of dispersion change. However, the heating of buildings also affects SIA concentration via chemical environmental changes, such as reaction rate and thermodynamic equilibrium changes. Accordingly, we investigated the effect of heating, by not only flow changes but also chemical environment changes, on SIA using a coupled chemistry-computational fluid dynamics (CFD) model. We investigated the change of temperature and wind in each scenario to separately identify the effects caused by the chemistry and dynamics of thermal heating on reactive aerosols. We found that both the chemical and dynamical effects reduce SIA concentration. The effect of heating caused by chemical environment changes on SIA was 49.4% larger than that of flow changes, suggesting the importance of chemical reactions in aerosol calculations in urban microscale aerosol simulations, which are not considered in most CFD-based models. We conclude that chemical speciation and detailed aerosol chemistry calculations, including an inorganic aerosol thermodynamic equilibrium model, should be considered for accurately accounting for the effect of heating on urban areas. This work was funded by the Korea Meteorological Administration Research and Development Program under Grant KMI2021-03312.

In order to obtain VOCs emission characteristics of different products in corresponding processes of pharmaceutical industry, and evaluate the impact of different treatment technologies on VOCs emission, this study took 11 typical pharmaceutical enterprises in Zibo City, Shandong Province as the research object, collected and analyzed VOCs samples of different processes and treatment technologies. The VOCs source spectra of different emission stages were compared using the coefficient of divergences (CD) method. The total non-methane hydrocarbon (NMHC) was used to characterize the overall emission of VOCs. The emission factors of pharmaceutical industry were calculated using the measurement method. The effects of different waste gas treatment technologies on VOCs emissions were evaluated according to the purification efficiency. The results show that the average VOCs emission factor of typical enterprises A and B in Zibo pharmaceutical industry is 7.97 ± 8.21 g/kg-product, which was 1/54 of the recommended value (430 g∙kg^-1) provided by literature (Ministry of Ecology and Environment of the People's Republic of China, 2014). There are significant differences in VOCs emission characteristics among different processes. After RTO treatment, the proportion of aromatic hydrocarbon species increased from 49% to 58%, while the proportion of other species decreased. However, after water spraying +UV photo-oxygenation treatment, the proportion of alkane species decreased from 75% to 69%, and the proportion of other species increased. The VOCs discharged by chemical synthesis enterprises are mainly alkanes, and the proportion of each species has almost no change after the treatment of condensation, spraying and activated carbon adsorption (CSA). Different treatment technologies have different purification efficiency for different kinds of PAMs. UV photo-oxygenation + water spray is more suitable for the removal of VOCs and odor in sewage stations, and RTO is more suitable for the removal of VOCs with higher concentration.

Secondary organic aerosol (SOA) in the atmosphere can be formed from the gas phase precursors and evolve in the particle phase. Compared to numerous studies on gas phase oxidation of volatile organic compounds (VOCs), the evolution of OA and corresponding mechanisms in the particle phase are still unclear. In this study, a thermodenuder (TD) was coupled with both a long time-of-flight aerosol mass spectrometer (L-ToF-AMS) and a soot particle aerosol mass spectrometer (SP-AMS) at an urban site in southern China during wintertime of 2018, to explore whether there is significant difference for OA evolution on black carbon (BC) aerosols. The average mass concentration of PM₁ during the sampling period was 19.5±9.5 μg m^-3, with OA as the most abundant chemical components. The OA measured by L-ToF-AMS and SP-AMS both were resolved into different primary and secondary factors, with the OA factors on BC particles showed different volatilities to that on the bulk aerosol. The BC-containing organics was more aged with a higher fraction of SOA contribution (83.3%) and higher O/C (0.63) compared with bulk OA. During the aging process of OA on BC particles, the fraction of BC-containing OA in total mass increased with the decreasing of volatility and correlation with BC. The more-oxidized oxygenated OA (MO-OOA, an agent of aged SOA) exhibited a unique preference on BC particles compared to other OA components. The abundant transition metals detected on BC particles was possibly the catalyser for the transformation from less oxidized OOA (LO-OOA, an agent of fresh SOA) to MO-OOA in aerosol aqueous-phase on BC particles and thus facilitates SOA formation. More studies are needed to further explore the potential role of BC and transition metals as catalysers for OA aging and the co-benefit of BC and transition metals pollution control on SOA reduction in the atmosphere.

Arsenic (As) exposure is a worldwide public health issue. Inhalation of atmospheric particulate matter enriched with As is an important exposure route. Little research has been conducted for atmospheric As, especially in Indian context. This study has been conducted over Shyamnagar (22°50’N, 88°23’E, 8masl) a semi-urban station situated in the eastern part of IGP near Kolkata metropolis for an entire year (2019). 24hr samples of fine mode aerosols (PM_2.5) were collected using a 5-channel aerosol sampler on every alternate day covering all the seasons. The Energy Dispersive X-ray Fluorescence technique was used to determine As concentrations along with other trace elements in PM_2.5. The major elements observed were Chlorine, Sulphur, Potassium, Silicon, Aluminum, Zinc, Iron, Sodium, Calcium, Barium, and Lead along with Arsenic. The annual mean concentration of As was found to be 381.8 ngm^-3 which is more than 60 times the Indian Standard (NAAQS; 6 ngm^-3). The concentrations were highest during winter (~800 ngm^-3) followed by postmonsoon (~ 630 ngm^-3), premonsoon (~ 160 ngm^-3), and monsoon (~ 80 ngm^-3). This implies the constant emission of atmospheric As throughout the year. As showed very strong correlations with Lead (Pb) throughout the year (R² = 0.99) suggesting some continuous stationary emission sources, e.g. battery industry. Low-grade coal containing arsenates (AsO₄^3-) can also be a source of Arsenic as they are used in local eateries and domestic cooking purposes as easily affordable fuel. Irrespective of the source, this kind of high concentration of As in the atmosphere can be of huge adverse impact from the perspective of human health for the local residents. This study draws the serious attention of the scientific community as well as policymakers for regular surveillance and action plans for such overlooked semi-urban stations.

China is infronting severe ozone (O₃) pollution although particulate matter reduced significantly, causing damages to public health and ecological systems. Here we utilized the comprehensive methods of ground-level observations, satellite data, and source-oriented chemical transport model to interpret O₃ variations throughout China from 2016 to 2019. A remarkably worsened trend of O₃ levels has been found both by observation and simulation in these years. Our results showed that the remarkable O₃ elevation was found in the NCP and YRD (maximum daily 8h average O₃ ~60 ppb) with an annual increasing rate of 10%. In addition, O₃ formation regimes also changes obviously in the NCP, which shifted from VOC-limited to transition regimes (4.9%) and transition to NO_x-limited regimes (9.6%). The elevation of MDA8 O₃ was mainly attributed to the enhanced atmospheric oxidation capacity (AOC) in above regions. Particularly, the increasing ratios of OH and HO₂ radicals (major oxidants) in NCP and YRD reached to ~15% and ~5%, respectively. The comprehensive study of long-term O₃ changes, formation regimes, and AOC based on a multimethod approach should be considered when designing O₃ control policies.

The contribution of organic aerosol (OA) to particle formation, mass, and number concentration is one of the major uncertainties in current climate and air quality models. In this study, a global aerosol model considering the volatility distribution of organic components was developed to simulate detailed microphysical processes of organic species and other components. The model calculates the kinetic condensation of low-volatility organic compounds and equilibrium partitioning of semi-volatile organic compounds. Using this model, the response of particle number concentration to emission change was investigated. The model results strongly suggest the important role of OA volatility distribution in particle formation and the impacts of emission change on particle number concentration over the whole globe, especially over the areas influenced by anthropogenic sources.

The relationship between air quality and the pollutants generated by vehicles is of great concern in many cities around the world. In this research, we aim to understand the interaction between these two factors and the role of air filters in mitigating air pollution. To do so, we have used a combination of the Weather Research and Forecast model with a chemistry version 3.7.1 to study the impact of anthropogenic emissions from vehicles on air quality. In our study, we have used the CAPSS, Clean Air Policy Support System dataset as the source of anthropogenic emissions. The WRF-Chem model is used to the dispersion of these pollutants in the atmosphere and then predict the impact of these emissions on air quality. The simulation was executed from April 28 00 UTC, 2020 to May 30 00 UTC, 2020. Analysis data was made by the results 7 days after simulation began. This study prescribes that the mitigation of air pollutant emissions is valid only for the road emissions. i.e., line sources. Our results show that the use of air filters in vehicles can significantly reduce the concentration of air pollutants in the atmosphere by 5%. In conclusion, our study highlights the importance of considering the interaction between air quality and air pollutants generated by vehicles when developing air quality management strategies. The results of our research demonstrate the potential benefits of using air filters in vehicles to mitigate air pollution, and provide a basis for further investigation into the impact of air pollution on public health and the environment.

China began implementing the Air Pollution Prevention and Control Action Plan (APPCAP) in 2013 to improve air quality. Although the concentration of particulate matter in most areas has decreased significantly, some regions have yet to reach the second-level standard of China Environmental Air Quality Standards (CAAQS), such as Beijing-Tianjin-Hebei (BTH) and Yangtze River Delta (YRD). In this study, we used a source-oriented chemical transport model to quantitatively estimate the effects of inter-city transport on fine particulate matter (PM_2.5) among the 41 cities in the YRD. And use the Michaelis-Menten equation to quantify the relationship between the cumulative contribution rate and the distance between cities in the YRD region. The results show that it is more appropriate to use the Michaelis-Menten equation to indicate the relationship between the cumulative contribution rate and distance, 71% of R² values are greater than 0.9 in six regions and four seasons. The maximum contribution of regional transport (K1) and the distance where the regional transport contribution is half of the maximum contribution (K2) are different in different regions. The average value of K1 is 73.6, which is smaller in the northern part of the YRD and higher in central Jiangsu. K2 is larger in northern Jiangsu and central and southern Zhejiang. Primary PM_2.5 and secondary organic aerosols (SOA) contribute more within the YRD, with 82.9% and 88.6%. The local contribution in autumn and winter in the northern part of the YRD is lower than that in spring and summer, especially in northern Jiangsu, where it is 90.4% in summer and 53.0% in autumn and winter. K2 is larger on polluted days, which means that it will be affected by a larger range on polluted days. The results can provide a scientific basis for regional joint prevention and control in the YRD region.

In this study, firstly, three first-order local closure vertical diffusion schemes used in multiple mesoscale chemical transport models (CTM) were embedded in the CHIMERE CTM model and tested over a one-year simulation covering the whole of France. Three model configurations present a fair reproduction of pollutant concentrations both in urban and rural areas, indicating it is an effective way to reproduce the dispersion of pollutants in chemistry transport modeling. However, it cannot be expected to significantly improve the vertical mixing under the first-order closure scheme. Secondly, a 1.5-order turbulence kinetic energy-based eddy diffusivity closure scheme called the new eddy diffusion (NED) is implemented in CHIMERE to describe more realistic diffusion processes near the surface. A fifteen-day simulation encompassing a winter pollution episode was performed for three major cities with a horizontal resolution of 1.67 km and the first layer height at 12 m, respectively. The NED scheme improved NO2 simulations at most urban sites compared to the initial Kz diffusion scheme (IKD). Taking the root mean square error as evaluation criteria, the average improvements are 18.8%, 24.5% and 9.5% for NO2 simulation in Paris, Lyon and Bordeaux respectively. For the model performance of PM2.5 and PM10 simulations in the urban areas of Paris, the improvements are 13.5% and 19.1%, respectively. Overall, preliminary outcomes of this study are encouraging. The simulation with more sophisticated and realistic eddy viscosities are better than for IKD that is widely used in CTMs, but we need to realize that this is only a fifteen-day simulation for three cities, for further research, longer periods are needed with a greater variety of meteorological situations to prove the universality of the NED scheme.

In order to distinguishe effects of chemical mechanism and meteorological factors in the megacity Beijing, the sensitively experiments with the third-Generation Air Quality Modelling System is designed. The modelling system consists of the meteorological model (WRF), the emissions process model (SMOKE) and the air quality model CMAQ, and the experiments focus on the megacity Beijing for one year simulation in 2018. The baseline group experiment, named CB05_AER05 experiment, choses the gas-phase chemistry module CB05 and the aerosol mechanism AERO5, while the NOOP group experiment deactivates gas-phase chemistry module and the aerosol mechanism, and standing for the only meteorological factor without the chemical mechanism in the simulation. The results indicates that: 1) the modelling system plays well performance in the megacity Beijing in 2018. The correlation coefficient(R) of the simulated PM_2.5 concentration and the observation value reaches to 0.7, which the normal mean bias (NMB) is -14%~32% in the 12 National Standard Air Quality (NSAQ) observation stations in Beijing. 2) The chemical mechanism has an effect on the concentration of PM_2.5 up to 135% compared to the NOOP group results, while the PM_2.5 concentration in NOOP group is 43.3 ug/m³, which stands for the influencing of the meteorological factor. The influencing factors of the pollutant concentration are complex, which also proves that the atmosphere is a complex system. 3) For the photochemical species, the peak frequency of NO₂ concentration in the baseline group (CB05_AER05) and the NOOP group experiments are 26.9 ppbV and 2.2 ppbV, respectively. The numerical model is a good tool to quantify the effects of chemical mechanism and meteorological factors.

To ameliorate the substandard air quality, a comprehensive national emission database is crucial for determining the aggregate emission reductions that are necessary to conform to air quality standards. Thus, it is imperative to assess the uncertainties associated with these emission inventories. This study endeavors to evaluate the accuracy of emission inventories in replicating the satellite-observed AOD for the Indian region, and to identify the most suitable emission inventory among the available options for the common baseline year of 2015. The study contrasts four databases (three global and one regional) that furnish emission estimates for air pollutants in India, which include EDGARv5, REASv3.2, and SMoGv1. Simulations were conducted for different seasons using WRF-Chem V3.8.1. The key finding of the study is that EDGAR emerged as the best performing emission inventory among all the databases across all seasons, with an average Root Mean Square Error (RMSE) of 0.38 for the entire year.

Chlorine chemistry play essential roles in tropospheric physicochemical processes, such as affecting the ozone (O₃) level and secondary aerosol formation. Tropospheric chlorine chemistry was initially known to be important in the marine and polar atmosphere; however, more and more recent studies have indicated that chlorine chemistry was also active in inland areas. Various of gaseous chlorine species, such as nitryl chloride (ClNO₂), molecular chlorine (Cl₂), hydrochloric acid (HCl) and others, have been observed to be ubiquitous in the urban atmosphere, and were linked to the influence from anthropogenic sources like coal combustion and biomass burning activities. Upon photolysis, these chlorine species can release chlorine atom, which can react with O₃ rapidly. The chlorine atom can also oxidize volatile organic compounds (VOC) with a reaction rate up to two order of magnitude faster than the reaction of hydroxyl radical (OH) with VOC. Despite the importance, a full understanding of the chlorine cycling and its impacts in the atmosphere remains unclear, particularly in urban environment. Our recent findings have shown that the presence of reactive halogen compounds (such as bromine and iodine) can change our understanding on the chlorine cycling in the atmosphere. Here, we will present the vital chlorine species observations and current understanding of reactive chlorine cycling in the urban atmosphere of China, and further discuss on the future challenges of reactive chlorine chemistry.

The global emission inventories of Emission Data Base for Global Atmospheric research (EDGAR) and Evaluating the Climate and Air Quality Impacts of Short-Lived Pollutants (ECLIPSE) uses top-down approach considers the total fuel consumed or the number of vehicles registered, to estimate vehicular emissions in India. However, this might not represent the actual traffic emissions in a city. Therefore, the following study aims to estimate the vehicular emissions for the year 2019 using bottom-up approach for seven cities in Punjab, India viz. Jalandhar, Patiala, Mandi Gobindgarh, Khanna, Dera Baba Nanak, Naya Nangal and Dera Bassi. The traffic volume data for two wheelers, three wheelers, four wheelers, light commercial vehicles, heavy commercial vehicles, buses, and others (tractors etc.) were calculated at twenty-five pre-identified locations in all the cities. Jalandhar which had the maximum traffic volume data, contributes to 8.02 kt/day PM₁₀ and 3.54 kt/day PM_2.5 vehicular emissions (tail pipe and non-tail pipe emissions combined) respectively. The silt loading value observed for the seven cities lie in the range of 0.89 to 9.86 g/m². The study also compares the total estimated vehicular emissions with EDGAR and ECLIPSE global emission inventories. This study signifies the importance of bottom-up approach for the combined vehicular emissions. It is expected to provide an insight in the scientific community on the need for developing local emission inventories to devise city specific action plans for combating the air pollution.

Air pollution, especially PM2.5, is a significant environmental health concern globally, particularly in the Indian subcontinent. A minor proportion of PM2.5 is comprised of heavy metals, that can be toxic even at low concentrations, such as Cr, Cd, As, Pb, and Ni. Scientific evidence has illustrated that these metals can accumulate in adipose tissue and the circulatory system of the human body, leading to potential adverse effects on the digestive, cardiovascular, and central nervous systems, and may also act as a co-factor in the development of other diseases. This study focuses on the seasonal variation and health risk assessment of PM_2.5 bound heavy metals in two north Indian cities, Alwar and Amritsar in the states of Rajasthan and Punjab respectively. The study was conducted through a combination of field measurements and comparative analysis. PM_2.5 samples were collected during summer and winter seasons in both cities and analyzed for the presence of heavy metals using ICP-MS. The health risk assessment was carried out to understand the risk associated from the exposure to PM_2.5 and heavy metals to the population, particularly in terms of respiratory and cardiovascular diseases.

This study evaluated the Atmospheric Assimilation Capacity (AAC) of Vijayawada city in Andhra Pradesh, India. Emission inventory of 1 × 1 km² resolution was used to estimate the emission load of PM₁₀, PM_2.5, SO₂ and NOx under the Business as Usual (BAU) and Alternative (ALT) scenarios for the year 2030 compared to the reference year 2021. Subsequently, the emission load was fed into the AERMOD view v9.7.0 dispersion model to simulate the concentration of the pollutants in various scenarios for the city sources. The study reported the city's overall emission load of PM₁₀~5648 tonnes/day, PM_2.5~2131 tonnes/day, SO₂~1378 tonnes/day and highest for NOx~8912 tonnes/day. PM₁₀, PM_2.5 and SO₂ emissions were projected to increase by 36%, 22% and 11%, respectively, compared to a 14% decline in NOx emission load between the reference year and BAU 2030. For ALT 2030, the simulation of control and abetment measures resulted in a 16% decline in PM₁₀ and a 23% decline in PM_2.5. Implementing progressive measures in the transportation and industrial sectors also projected a significant decrease in gaseous emission load. Annual city average projected PM₁₀ and PM_2.5 concentrations reported the exceedance of Indian National Ambient Air Quality Standards (NAAQS) that resulted in reduced AAC due to an increase of 24% and 15% concentrations under the BAU 2030 scenario. However, the stringent implementation of ALT scenarios reported a high AAC with 16%, 24% and 38% decline in PM₁₀, PM_2.5 and NOx concentration, respectively. Among urban sources, construction and road dust were the prominent contributors to the city's PM₁₀ emission load, whereas transportation and coal-burning industries contributed significantly to PM_2.5, NOx and SO_2, respectively. Such measures will help improve the air quality and livability of the city.

Air quality in Beijing-Tianjin-Hebei and its surrounding regions has been significantly improved during the Three-Year Action Plan for Winning the Battle Against Air Pollution from 2018 to 2020. The effectiveness of anthropogenic emissions reduction on PM_2.5 over the North China Plain remains unclear. In this study, the Nested Air Quality Model System and observations were used to quantitatively examine the contribution of meteorological and emission variations to autumn-winter (November to February) PM_2.5 concentrations in Beijing and its surrounding regions from 2018 to 2020. The model reproduced the temporal evolution of atmospheric pollutants and components well. The results showed that the observed mean PM_2.5 concentrations during the winter were reduced by 15% and 29% in 2019 and 2020, respectively, compared with that in 2018. In 2019, the meteorological conditions over the Beijing-Tianjin-Hebei (BTH) region were poor, causing an increase of PM_2.5 by 7.6% in BTH and 11% in Beijing; therefore, the control of air pollutant emission compensated for the unfavorable influence of meteorology and emission reduction was the decisive factor for PM_2.5 reduction; In 2020, the meteorological conditions were relatively favorable, and the changes in meteorological factors and emission reduction led to a decrease in PM_2.5 concentration by 18.6% and 10.5%, respectively, compared to 2018. Emission reduction also contributed to decrease in nitrate, ammonium, organic matter, element carbon and precursors. This work confirmed the obvious environmental benefit of pollution control measures from 2018 to 2020.

Recently, because the fine particulate matter (PM_2.5) concentrations monitored in Taiwan have substantially decreased due to less long-range transport of air pollutants in East Asia and the COVID-19 pandemic outbreak, many environmental groups have suggested that Taiwan’s Environmental Protection Administration should adjust the current regulatory standard for PM_2.5 to an annual average of 12 μg/m³, which would accord with that of the United States. Thus, to determine whether such a regulatory adjustment is needed, assessing air quality on the basis of spatiotemporal PM_2.5 distributions is critical for establishing risk maps and maintaining human health. This study spatiotemporally characterized the air quality in Taiwan by using multivariate indicator kriging (MVIK) according to current Taiwanese and US regulatory standards for annual average PM_2.5 concentrations (i.e., 15 and 12 μg/m³, respectively). First, long-term PM_2.5 concentrations were statistically analyzed to determine recent stable temporal PM_2.5 data. Multivariate integration was implemented using the 2019-2021 and 2020-2022 PM_2.5 data. MVIK was then adopted for modeling probabilities according to the two standards. Finally, compared with PM_2.5 observations, quantile estimates based on the probabilities were employed to determine the optimal classifications for establishing risk maps of different PM_2.5 standards. Suitable regulatory standards for PM_2.5 are discussed according to risk classification maps. The results indicated that the multivariate integration of temporal PM_2.5 data used in MVIK can effectively streamline the analytic process. The multivariate integration of 3-year PM_2.5 data was suitable for the assessment of adjusting a regulatory standard for annual average PM_2.5. The air quality in the Central, Yunchianan, and Kaoping air quality regions was found to pose high risks to human health due to long durations of high levels of PM_2.5 concentrations. Adjusting the regulatory standard of annual average PM_2.5 in Taiwan to 12 μg/m³ would be inappropriate at this time.

Previous lab studies have suggested that other non-BC pollutants can promote the aging of black carbon (BC). However, most of the experiments were carried out with a single component mixed with BC, and the results obtained in previous studies under simple laboratory conditions may not be applicable to atmospheric relevant conditions. Here we will use a robust experimental system to perform a laboratory simulation of the internal mixing of flame-generated BC aggregates with atmospheric-relevant secondary pollutants in a smoke chamber. The impact on BC optical properties and CCN activities from the concentration of pollutants, coating sequence, as well as relative humidity, will be investigated. Variations in particle size, mass, coating thickness, effective density, dynamic-shape-factor, chemical composition, optical properties and CCN activities were determined online by a suite of instruments. Accordingly, this study will help to attain a comprehensive and clear understanding of the effects of coatings on the morphology, optical properties and CCN activities of BC and provide a foundation for accurately predicting the radiative forcing of BC as well as its role in air pollution control and climate change.

Dust emission acts as a crucial part in the dust cycle that determines dust related processes at both regional and global scale, such as long-range transport, dry/wet deposition, and radiation forcing. However, most dust emission simulations utilizing dust emission scheme merely investigate natural dust, neglecting the contributions of anthropogenic dust induced by direct or indirect anthropogenic activities, resulting in great uncertainties in estimating dust emissions by previous numerical modelling. To comprehensively reproduce the anthropogenic dust emissions process, both “indirect” and “direct” anthropogenic dust emission schemes were constructed to simulate anthropogenic dust emissions originated from diverse kinds of source regions in the study. Results showed that using both indirect and direct anthropogenic dust emission schemes show good performance on reproducing the spatio-temporal distributions of anthropogenic dust at the global scale during 2007–2010. The natural dust sources contributed 81.0%(6.34 ± 0.31 μg m⁻² s⁻¹) of the global dust emissions and the anthropogenic contributed 19.0%(1.01 ± 0.07 μg m⁻² s⁻¹) of the residual. The anthropogenic dust emissions concentrate in semi-arid, semi-humid and humid regions and generally fluctuated between 0.1 and 10 μg m⁻² s⁻¹. Anthropogenic dust accounted for about 42.99% of the total dust emission in the semi-arid areas. In semi-humid and humid areas, represented by cities, direct anthropogenic dust emissions are much higher in developing regions than in developed regions. This study provides a basis for accurately estimating dust emissions, further studying the radiation forcing and climate effects of dust.

Sudden mega natural gas leaks of two Nord Stream pipelines in the Baltic Sea (Denmark) occurred from late September to early October 2022, releasing large amounts of methane into the atmosphere. We inferred the methane emissions of this event based on surface in situ observations using two inversion methods and two meteorological reanalysis datasets, supplemented with satellite-based observations. We concluded that approximately 220 ± 30 Gg of methane was released from September 26 to October 1, 2022. This figure is about 0.08% and 85% of global and Danish annual anthropogenic methane emissions, respectively, and comparable to the annual anthropogenic methane emission in Austria. It surpasses the Aliso Canyon gas leak (100 Gg) that occurred in California in 2015, making the Nord Stream leak the largest gas leak ever reported.

Air quality especially in urban regions is known to be highly influenced by anthropogenic activities (e.g. vehicle engine combustion, industrial processes, solvent use, etc.). The Chinese New Year (CNY) is a well-celebrated festivity in Taiwan and anthropogenic behavior in the region changes during this time - people are on holiday vacation, there is a probable decrease in manufacturing operations, and some people celebrate it with fireworks and firecrackers. In this study, we investigated the effect of CNY on the air quality in the region particularly in terms of VOCs, PM, NO_x, CO, and O₃. The hourly ambient measurements were done at Urban Air Pollution Research Station, Taichung City, Central Taiwan from January 9 to February 8, 2023. The sampling period was subdivided into three periods, the pre-CNY (Jan 9-19), the CNY (Jan 20-29), and the post-CNY (Jan 30- Feb 8). Results showed a significant decrease in the concentrations of the measured CO and VOCs particularly of benzene, toluene, xylene/ethylbenzene, and trimethylbenzene during the CNY period. In particular, the sum of BTX average concentrations was 11.83 ppbv, 2.60 ppbv, and 10.97 ppbv for pre-CNY, CNY, and post-CNY, respectively. PM 2.5 concentrations also showed a huge decrease during the CNY period (20.46±3.48 µg/m³) as compared to pre-CNY (27.68±14.18 µg/m³) and post-CNY (25.70±5.58 µg/m³). NO_x is also found to be lowest during the CNY period (6.28±2.56 ppbv), as compared to pre-CNY (15.26 ±7.32 ppbv) and post-CNY (15.38±9.12 ppbv). CO, BTX, and NO_x are known to be tracers of vehicle emissions hence the decrease can be mainly attributed to the low traffic in the area during the CNY holiday. Meanwhile, ozone concentration did not vary that much which is subject to further analysis. Diurnal variations during each period will also be investigated to further understand the observed measurements.

The present study investigated the seasonal trend and source attribution of n-alkanes in dry deposition during 2021 at an urban site in Indo-Gangetic Plain i.e., Delhi. Dry deposition samples (n=30) were collected weekly on 24 hourly basis (except COVID lockdown). The samples were extracted in methanol and analyzed for n-alkanes (C21-C40) using Gas Chromatograph-Mass Spectrometer. However, only C21-C31 n-alkane homologues could be quantified in the samples. The annual average dry deposition flux of total n-alkanes was 51.5±4.5 µg/m2/day following the seasonal trend as winter (75.6, n=9) > post-monsoon (61.8, n=7) > summer (34.3, n=6) > monsoon (28.4, n=8). The seasonal variability in deposition fluxes of total n-alkanes was more pronounced during winter and post-monsoon due to higher emissions arising from burning of woods for domestic heating and biomass burning respectively. The micrometeorology of the region also aids to the formation of shallow planetary boundary layer resulting in accumulation of the pollutants during this period. The drastic decline in the average dry deposition flux of total n-alkanes during summer could be due to volatilization of the n-alkanes at high temperature during summer season. The minimum deposition flux of total n-alkanes during monsoon provides the insights on the wet scavenging of the atmospheric pollutants. The molecular distribution of ≤C25 n-alkane homologues with maxima at C23 and C25 indicates the signature for vehicular emissions while the odd-even predominance of the >C27 n-alkanes homologues with maxima at C29 suggests the prevalence of biogenic sources. The annual CPI value (1.9±0.1) confirms the presence of mixed emission sources coming from fossil fuel burning and biogenic emissions at the sampling site. Concentration Weighted Trajectory (CWT) showed the prevalence of long-range transport from Arabian, Gulf, Myanmar and Indo-Gangetic Plains during summer and monsoon while prominent local and regional sources during winter and post-monsoon seasons.

The SuperDARN network of meteor radars near 60^o N have been used in a number of studies to characterize the climatology of planetary waves (PW) in the mesosphere and lower thermosphere (MLT). Classically, planetary waves are considered to propagate westward relative to the background flow due to planetary vorticity. However, instabilities in the zonal-mean flow produce vorticity gradients that provide a rich spectrum of PWs generated in-situ in stratosphere and mesosphere, including eastward propagating PW. Such instabilities are generally associated with sudden stratospheric warming (SSW) events, and eastward propagating planetary waves (EPW) have typically been associated with SSW. Here we present both super-posed epoch analyses of EPW during SSW as well as an extensive climatology of EPW during winter seasons from 2007 to 2019 using the SuperDARN meteor network. The details of the analysis to extract the EPW will be presented as well as the results showing that while EPW do occur during SSW events, their presence in the MLT is ubiquitous throughout the winter season.

The secondary ozone layer is a global peak in ozone abundance in the upper mesosphere-lower thermosphere (UMLT) around 90-95 km. The effect of energetic particle precipitation (EPP) from geomagnetic processes on this UMLT ozone has not been well studied. In this research we investigated how the secondary ozone response to EPP from the Microwave Limb Sounder (MLS) and the Sounding of the Atmosphere using Broadband Emission Radiometry (SABER) instrument on the Aura and TIMED satellites, respectively. In addition, the Whole Atmosphere Community Climate Model with thermosphere and ionosphere extension and specified dynamics (SD-WACCM-X) was used to characterize the residual circulation during EPP events. By comparing ozone and circulation changes under High- and low-Ap conditions, we report regions of secondary ozone enhancement or deficit across low, mid and high latitudes as a result of circulation and transport changes induced by EPP.

Keystone is a mission concept for a limb sounding satellite to explore the upper mesosphere and lower thermosphere, as well as the ionosphere. Keystone was a candidate missions for the ESA Earth Explorer 11 Call (EE-11). It’s a maturation of the EE-10 and EE-9 mission concept LOCUS. In its present form Keystone comprises of three instruments: A Supra-THz radiometer, an infrared radiometer, and a UVIS spectrometer.The infrared radiometer has four channels at 15um (CO₂), 9.6um (O₃), 5.3um (NO), and 4.3um (CO₂). The UVIS spectrometer ranges from 230nm-780nm (O, O₂, O₃, O⁺, M/Mg⁺, Fe/Fe⁺, and Temperature). These two instruments are a continuation of upper atmospheric measurements such as MIPAS, HiRDLS or SABER, and OSIRIS, SCIAMACHY or ICON/GOLD, respectively. Keystone addresses the looming gap in upper atmospheric measurements at these wavelengths. The novelty of the Keystone mission is the heterodyne Supra-THz sounder with frequencies of 770GHz (O₂), 1.1THz (NO,CO), 2THz (O), 3.5THz (OH), and 4.7THz (O). The THz instrument - made possible at these very high frequencies by novel quantum cascade laser (QCL) local oscillators - will provide vertical profiles of its target species. Prime among these is the elusive atomic oxygen (O), which is the kingpin of upper atmospheric chemistry and energy balance. Atomic oxygen is the missing keystone in the interpretation of both infrared and UVIS measurements. The unique combination of the three Keystone instruments presents a strong and novel synergy. By resolving age old questions - e.g., the conundrum of the quenching rate of CO₂ with O - Keystone will significantly improve our understanding of the upper mesosphere and thermosphere, the least well-known region of the atmosphere, maybe the entire planet.

It has been believed that a strong pre-existing tropical cyclone (TC) can contribute to the genesis of another TC to the east or southeast in the western North Pacific through the Rossby wave dispersion (Ritchie and Holland, 1999; Li et al. 2006). However, these TCs are usually observed in the easterly wind where the stationary Rossby wave cannot exist, and a well-coordinated numerical experiment has not been conducted to elucidate the impact of a pre-existing TC. Therefore, we conducted a set of numerical simulations in which a pre-existing TC is removed sufficiently prior to the genesis of a subsequent TC. It turned out that a subsequent TC was generated even without a pre-existing TC in all 10 simulations for a pre-existing TC case according to the genesis environment database of Fudeyasu and Yoshida (2018). Also, the removal of a pre-existing TC hardly affects the intensity of a subsequent TC. This type of a TC is typically located south of a subtropical high, where the horizontal cyclonic shear prevails from west to east widely. Further analysis showed that these TCs were frequently observed in MJO phase 7, El Nino, and positive SST anomaly to the east of the western North Pacific. These are favorable for the genesis of a TC in the broad area of the western North Pacific. In other words, the characteristics are simply explained by the favorable condition in a broader area for the genesis of two TCs in which an eastern TC is generated later because the SST is lower in the east of the western North Pacific, not necessarily relying on the Rossby wave dispersion from a pre-existing TC.

Analysing the intensity of typhoons that have affected Korea over the past ten years revealed that half of them were very strong typhoons (with a maximum wind speed of 44 m or more per second). In particular, one of the four typhoons had a maximum wind speed of more than 55 m/s, which is the superpower level newly established in 2021. As such, the typhoons affecting Korea are gradually getting stronger. The World Meteorological Organization (2021) has highlighted that the world will be exposed to risks from complex weather disasters rather than single weather disasters. In particular, weather disasters caused by typhoons are not a single phenomenon, such as strong winds, heavy rains, and tsunamis, but natural disasters that can have simultaneous adverse effects on air quality; therefore, preparation and evaluation are required. Accordingly, the National Weather Service (NWS) declared a paradigm shift for Weather-Ready Nation (WRN), an influence-based weather information service to support decision-making. This is a national strategic plan to minimise damage by acting in advance against the effects of extreme weather events (such as heavy snow, tornadoes, hurricanes, floods, and droughts). Researchers have established partnerships with the NWS WRN and AMBASSADOR and are conducting research to predict the impact of typhoons in Korea. In line with this, this study intends to establish and introduce the Typhoon-Ready System (TRS), which produces pre-disaster information (risk) on complex weather disasters associated with typhoons across Korea and in each detailed area. We believe that this system can be used as a countermeasure to the typhoon climate crisis on the Korean Peninsula. This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT) (No. 2020R1F1A1068738)

In general, tropical cyclone (TC) rainfall accumulation usually decreases with faster TC translation speed but increases with heavier rain rate. However, how the TC rain rate changes with translation speed is unclear. Here we show that, in all TC basins, the average TC rain rate significantly increases with translation speed. On average, the rain rate in a fast-moving TC is 24% higher than in a slow one. This difference increases with TC intensity, with category 3–5 TCs having a 42% increase while tropical depressions exhibit only a 9% increase. The increase in the average TC rain rate with translation speed is mainly caused by the TC net inflow in the lower troposphere, as well as vertical wind shear. These findings have important implications not only for a deeper understanding of rain rate changes in a translating TC but also for short-term forecasts of TC rainfall and disaster preparedness.

This study assesses the relative impacts of model resolutions, tropical cyclone (TC) trackers, and ocean coupling on simulating TC climatology over the western North Pacific (WNP) based on six Coupled Model Intercomparison Project phase 6 (CMIP6) High-Resolution Model Intercomparison Project (HighResMIP) models from 1979 to 2014. The HighResMIP multimodel ensemble (MME) analysis shows that the high resolution has a higher Taylor skill score II (S₂) in both temporal and spatial patterns of TC genesis frequency and accumulated cyclone energy (ACE) than the low resolution. In contrast, the TempestExtremes tracker (coupled run) results in a higher S₂ in temporal patterns but a lower S₂ in spatial patterns than the TRACK tracker (uncoupled run). Among the three factors, increased resolution leads to the greatest improvement in S₂ in both temporal and spatial patterns. Furthermore, this study investigates the projections of future TC activity over the WNP by HighResMIP under the SSP5–8.5 scenario. Overall, HighResMIP MMEs project a decrease in the genesis frequency, track density, and ACE of all TCs, with the high-resolution, TRACK tracker, and uncoupled run showing greater magnitude. The high-resolution MMEs, using both trackers, project an increase in the genesis frequency and ACE of intense TCs in the coupled run. Moreover, TC track density and ACE show a larger poleward migration in the coupled run than in the uncoupled run, consistent with the significant surface warming in the northern WNP.

Land cover changes and anthropogenic emissions cause by megacities can modulate convection and precipitation of the landfalling storms. In this study, we investigate the influence of megacity (the Pearl River Delta) on Typhoon Nida (2016) using the Weather Research and Forecasting model coupling with Chemistry (WRF-Chem). By changing the landuse type and initial anthropogenic emissions, the results suggest that megacity significantly enhances rainfall in outer-rainband in downstream area of the Pearl River Delta (PRD), with the land effect more significant than anthropogenic emissions. Urban land modifies the convection in the downstream area of PRD by enhancing vertical velocity, generating more sensible heat fluxes, and further promoting ice phase microphysical processes. The effect of anthropogenic emissions becomes evident after intensifying vertical convection, accelerating the conversion of hydrometeors into precipitation and further increasing precipitation in this region.

In 2017, Super Typhoon Hato was characterized by pronounced convective asymmetries in its inner core immediately before landfall in China. Together with the asymmetric convection, quasi-periodic lightning activities occurred concurrently, with a period of approximately 3 hours. This study discussed the characteristics of short-cycle lightning activity and corresponding polarimetric quantities in the inner core of Super Typhoon Hato (2017) before its landfall. Although the strongest inner-core convection was located persistently in the downshear and downshear-left quadrants, the lightning bursts behaved quasi-periodically with a cycle of about 3 hours. The analysis revealed that wavenumber-2 vortex Rossby waves (VRWs) were present in the inner core, propagating cyclonically and intensifying downshear left. When the positive perturbation of the VRWs was coupled with the VWS-forced convective enhancement, the local convection was significantly reinforced. Moreover, the slanted updrafts intensified by the phase-locking between VRW activity and convection strengthened by VWS can invigorate the growth of graupel, probably through riming processes, and further enhance the charge separation and lightning production immediately outside the eyewall. The invigorated graupel growth accompanied by lightning outbreaks can enhance surface precipitation and more efficient warm-rain growth below the melting layer. The radar observations demonstrated that when other VRWs were excited and propagated to the left side of the shear, the abovementioned phase-locking occurred again, leading to increased volumes of graupel and updrafts near the eyewall and short-cycle lightning bursts.

An idealized WRF model is used to investigate the track evolution of a tropical cyclone approaching westward at different departure positions toward an elongated mesoscale mountain (mimicking Taiwan) at different orientation angles (denoted by A0 and A90 for south-north and east-west orientations herein). The track deflection is primarily controlled by the meridional departure position as well as the ratio of the vortex size and effective terrain length (the aspect ratio, or the nondimensional vortex size). For a lengthy mountain range (A0) with the smallest aspect ratio, the cyclone track will be significantly deflected counterclockwise (northward) when the cyclone is closing to the terrain. The counterclockwise deflection ahead of the mountain is much enhanced as the aspect ratio is much larger as in A90 (a shallow mountain range). However, the counterclockwise deflection is stronger as the mountain range of A0 is rotated by 45 degrees counterclockwise rather than clockwise despite the same aspect ratio. Due to such counterclockwise paths, the tracks at later times may turn southward to the leeside of the terrain and thus clutter near the same location. When departing south of the terrain, the upstream cyclone may somewhat be deflected southward before taking a counterclockwise path (thus northward). The degree of the track deflection is also reduced by increased meridional departure and basic flow intensity. Only when the terrain blocking becomes significant as in A0, the cyclone can be deflected northward but then rapidly southward near landfall, due to the channeling effect of the induced strong northerly jet along the mountain base. The wavenumber-1 vortex flow and potential vorticity (PV) budget analysis has helped explain the track deflection that is dominated by horizontal PV advection in comparison to both vertical PV advection and differential diabatic heating that somewhat modulate the tracks in the vicinity of the mountain.

Predicting tropical cyclone activities has been a topic of great interest and research. Nonetheless, most of the existing forecast models are closed-source, and the prediction performance is fair, especially the early prediction. Therefore, we employ statistical methods to develop “simpler, more open and accurate” (compared with the existing) seasonal forecast models and build an open-source operational platform for providing seasonal forecasting service, focusing on the western North Pacific (including South China Sea, East Asia and Southeast Asia). In this conference, we will promote 3 statistical seasonal forecasting models (namely SYSU models) to predict the number of tropical storms over the western North Pacific, and number of tropical storms that could make landfall on South China and East China by every mid of May using preseason factors as the predictors.

Chlorophyll can significantly affect climate systems through changes in atmospheric and oceanic circulation, and therefore could impact tropical cyclone (TC) activity. Here, we use future ensemble simulations performed by an Earth system model to investigate the biological feedback by future chlorophyll change on the large-scale environmental conditions that affect TCs over the western North Pacific (WNP). For this investigation, two sets of global warming simulations are conducted with activated and inactivated marine ecosystem models. Compared to the inactivated simulation, the increased chlorophyll concentration over the tropical eastern Pacific is a crucial feature in the activated simulation. The increased chlorophyll concentration over the tropical eastern Pacific strengthens the upper ocean’s stratification, inducing shoaling of the mixed layer and intensification of the equatorial upwelling. These changes lead to a La Niña-like SST warming pattern, hence intensifying Walker and local Hadley circulations. The intensified local Hadley circulation derives upper-level westerly and lower-level easterly wind anomalies in the southeastern part of the WNP (SE-WNP), resulting in a significant decrease in the vertical wind shear and a significant increase in the genesis potential index (GPI) in that region. The enhanced GPI in the SE-WNP and further TC properties are reproduced using a high-resolution atmospheric model simulation, which supports our Earth system model results. Our results can contribute to the understanding of the projection of TC activity in other basins because the tropical Pacific SST variability can influence the climate conditions in other basins. Acknowledgments: This work was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean government (No. 2020R1A4A3079510).

One of nature's most violent and destructive occurrences, tropical cyclones (TCs) have a significant negative influence on society and the economy. Reducing TC damages requires precise TC intensity estimation and forecasting. The Geo-KOMPSAT-2A (GK2A) satellite images are used in this study to create the Convolutional Neural Network (CNN) model, which estimates the TC intensity in the western North Pacific (NWP). Given the insufficient GK2A data, the present study adapts a transfer learning technique, which uses information learned from available Communication, Ocean, and Meteorological Satellite (COMS) satellite images, to develop the model. When the two data are similar, the transfer learning technique performs well because it can enhance learning a new task by transferring knowledge from a task that has already been trained. The present CNN model based on transfer learning approaches lowered mean absolute error (MAE) by up to 27% compared to GK2A-only learning because COMS and GK2A satellite data use similar infrared channels. Our findings imply that transfer learning, combined with additional satellite data or artificial intelligence methods, will represent a significant advance in the estimation of TC intensity. Acknowledgement. This research was supported by Korea Institute of Marine Science & Technology Promotion(KIMST) funded by the Ministry of Oceans and Fisheries, Korea (20180447, Improvements of ocean prediction accuracy using numerical modeling and artificial intelligence technology) and the National Research Foundation of Korea(NRF) grant funded by the Korea government(Ministry of Science and ICT)(No. RS-2022-00144325)

Each year, there are significant variations in both the activity of tropical cyclones (TC) and the damage they do. The reduction of wear and human loss brought on by tropical cyclones (TCs) depends heavily on the longer-term forecasting of TCs. In this study, the primary development regions and precursors responsible for the genesis and intensification of TC were identified using a Causal-network-based approach. However, there are numerous worldwide links that connect all of the severe occurrences. Therefore, using this Causal Effect Network (CEN) based algorithm, it is examined how tropical cyclone teleconnection and correlation with El Nino Southern Oscillation (ENSO), Indian Ocean Dipole (IOD), and North Atlantic Oscillation (NAO) during the satellite era (1980-2020) over the North Indian Ocean (NIO) basins. The most appropriate metric for cyclone energy is Accumulated Cyclone Energy (ACE); its correlation with the various factors is investigated. We examined the variation in TCs activity during all three phases (positive, negative, and neutral phases). The findings indicate an upward trend in ACE over the NIO region during that particular time period. The most intense cyclones last longer at this time, but they are less frequent overall. After 1997, ACE begins to change, and it continues to climb sharply. Analysis of the sea surface temperature (SST), and vertical wind shear (VWS) between 850 and 250 hPa is conducted, and it demonstrates ACE's positive changes and variability. Causal graphs show the contributing regions with a good significance level. These findings could lead to a better understanding of the teleconnections between the events in the atmosphere or the ocean, and better tropical cyclone forecasting could lessen the damage brought on by TCs and shows how other phenomena affect the different parameters.

The outbreak of intense convection in the inner core of tropical cyclones (TCs) may promote the rapid intensification of TCs. It is well known that TC inner-core convection varies diurnally and usually peaks in the early morning. Recent studies found that TC intensification rate (INTRATE) also exhibits a diurnal cycle that is in phase with the inner-core deep convection. These studies further proposed that the nocturnally enhanced inner-core convection may simultaneously promote the TC INTRATE. However, the previously reported diurnal amplitude of INTRATE is just marginal (5-10%). It is unknown whether and how the diurnal cycle of the INTRATE varies among different TC intensifying periods. Also, whether the initial time of the TC intensification exhibits an evident diurnal signal? Based on the analysis of 30-year TC track data, this study investigates whether and how the INTRATE and initiation time of intensifying TCs vary diurnally. TC intensifying periods are classified into slowly intensifying (SI) and rapidly intensifying (RI) events. RI events last ～42 h on average, much longer than SI events. Interestingly, TC intensification including RI is the most likely to initiate at 00-06 LT. The INTRATE shows limited diurnal variations, especially for SI and RI events (amplitude of 6-8%), and the peak INTRATE time shifts from the morning in non-RI events to the late afternoon for RI events. Inner-core convection of all intensifying events maximizes in the early morning, in phase with the peak initiation time. However, the INTRATE and inner-core convection in RI TCs are diurnally out of phase, suggesting that the nocturnally enhanced inner-core convection may play a role in triggering TC intensification (e.g., RI), but not maximizing the INTRATE.

Typhoons are one of the deadliest natural disasters causing damage to human life and property so accurate forecasts of typhoons are crucial to reducing the risks. For the prediction of typhoons, forecasts from numerical weather prediction models (NWPs) are fundamental, and operational centers largely rely on NWPs. In the meantime, NWP techniques for the course and intensity of typhoons have been developed over many years, but prediction errors still exist so it is necessary to verify errors and analyze biases of the NWPs.
In this presentation, the prediction errors of the typhoon information from various kinds of NWPs were calculated compared to RSMC Tokyo's best track, and the tendencies of each NWP were analyzed for recent four years. The NWP used in the analysis were the global and ensemble models of KMA, NCEP, US Navy, ECMWF, and JMA.
As a result, DPEs show different validation scores every year due to the characteristics of typhoon phenomena. For example, there are many typhoons with unusual paths in 2021, so the accuracy is relatively low in all models. In this case, it is hard to predict both tracks and intensity so the model consensus didn't agree well. Also, the spatial distribution of biases such as along-track biases (ATBs), cross-track biases (CTBs), and biases of maximum wind speed are analyzed. NWPs tend to simulate typhoons' track leftward and slower, and intensity weaker. Moreover, some distinct biases are shown for the different regions in the western North Pacific area.

The decreasing sea surface temperature with increasing latitude is one of the causes of the weakening of the intensity of typhoons moving to mid-latitudes. In 2016, the track of Typhoon Chaba passed over the East China Sea where the sea surface temperature (SST) was abnormally high; hence, Chaba maintained a relatively high intensity. The Changiang River discharge peaked in early July and gradually decreased before typhoon Chaba approached. Conversely, the salinity in the East China Sea was minimal in early August and increased. Moreover, the positive anomaly area of sea surface temperature in the East China Sea matched the negative anomaly area of sea surface salinity related to Changiang river discharge. Changiang freshwater inflow and sea surface temperature in the East China Sea had 30 days lagged positive correlation. Therefore, based on these, we determined that the increase in the East China Sea's sea surface temperature is associated with the outflow of the Changiang River. In this study, we investigated the sea surface temperature warming due to ocean stratification caused by Changiang diluted water and following typhoon-ocean interaction after warming and ocean stratification using a coupled atmosphere-ocean modeling system. The freshwater inflow caused ocean stratification in the East China Sea and formed a barrier layer, and this barrier layer inhibited the vertical mixing and energy transport between the surface and the thermocline and led to sea surface warming. As a result, the stabilized ocean structure restricted sea surface cooling induced by typhoon-forced upwelling and turbulent mixing. In conclusion, this study discovered the mechanism that maintains the intensity of TCs moving northward using ocean-atmosphere couple modeling, which may be used to improve the performance of ocean-atmosphere couple modeling in predicting the intensity of TCs.

Super typhoon Lekima (2019) is the 5th strongest typhoon to make landfall in mainland China since 1949. After its landfall, typhoon Lekima moved northward along the coastline, resulting in an extreme rain event in Shandong Province that makes the largest precipitation within the available meteorological records. A WRF model simulation that well produces the track and intensity of typhoon Lekima and the spatio-temporal evolution of the rainfall is used to analyze the multi-scale characteristics of the extreme rain event. Different from the typhoon precipitation occurred at low latitudes, the extreme rain event occurred in midlatitudes were influenced by the interactions of midlatitude synoptic systems and typhoon circulation, especially with five mesoscale rainbands. The midlatitude synoptic systems, mainly including the upper-tropospheric jet, the Western North Pacific Subtropical High, the mid-latitude trough, the low-level jets and typhoon Krosa (2019), allow typhoon Lekima to maintain its intensity after landfall and provide favorable kinematic, thermodynamic and moisture conditions for the heavy rainfall in Shandong. Based on the evolution of the mesoscale rainbands, the extreme rain event can be divided into three stages. The first stage can be classified as a distant rainfall, which was affected by two convective rainbands associated with boundary layer processes. The second stage had the largest precipitation, featured the formation of a frontal zone in Shandong interacting with typhoon Lekima. The third stage had weakened rainfall and was directly influenced by the spiral rainband of typhoon Lekima.

AMOC is a typical density-driven thermocline circulation, featuring a clockwise basin-scale meridional overturning in the present climate. A freshwater hosing over a north Atlantic leads to slowing down or even collapse of AMOC, while in order to recover the typical state of AMOC, over-reduction of freshwater than that used for a collapse of AMOC is required. This indicates an existence of multiple equilibria state, which can be visualized by a hysteresis diagram showing a relationship between freshwater forcing (FWF) and AMOC intensity. Here, we explored AMOC hysteresis dependency on a frequency of FWF using a Stommel’s box model. Results showed that the abrupt transitions between typical state and collapse state were lagged as FWF changes faster. Furthermore, as the complexity of system increases, the detailed hysteresis feature was modified. Especially, the abruptness in transition during decreasing FWF phase was more distinct than that during increasing FWF phase. This asymmetric hysteresis response to FWF was attributed to nonlinear salt-advection feedback, which is mainly operating on a typical state but does not on a collapse state of AMOC.

ENSO (El Nino – Southern Oscillation) is the most dominant mode on Earth, having tremendous impacts on nature and human society. Since the Earth’s climate is changing continuously by anthropogenic greenhouse gas emissions, understanding of the ENSO teleconnection in changing climate is needed. In this study, based on an idealized CO₂ ramp-up and ramp-down climate model experiment, which increases CO₂ concentration by 1% per year (ramp-up) and decreases symmetrically (ramp-down), we show how the ENSO teleconnection responds in the CO₂ removal experiment and what factors induce the changes in ENSO teleconnection. ENSO teleconnection patterns in midlatitude 500hPa geopotential height anomalies show the systematic eastward-shift tendency during both the ramp-up and -down periods, and thus the ENSO teleconnection change shows significant hysteresis behavior. To investigate main cause of this hysteresis, we analyze the correlation between the changes in ENSO characteristics including standard deviation, skewness, phase shift, and its convection center, and the NINO3.4-500hPa geopotential height regression. The results reveal that the change in ENSO teleconnection in each region was correlated with the different factors. Especially, the ENSO skewness change was related to changes in ENSO teleconnections near North and South America, and the phase change in ENSO induces the change of ENSO teleconnection near Africa, the Aleutian region, and Australia. Through this study, it is expected that we could predict changes in ENSO teleconnection in each region according to changes in ENSO characteristics due to climate change.

Arctic amplification refers the fastest warming of Arctic region due to Anthropogenic increase of greenhouse gases. The main cause of this fast warming is known as so-call ‘ice-albedo feedback’ especially associated with the reducing sea ice area over the Arctic. Therefore, in near future, the emergence of seasonal sea ice free would be expected. The prediction of sea ice free timing is challenging because of complexity in atmosphere-ocean-ice system. In particular, it is known that the fast disappearing of Arctic sea ice in early 21 centaury was not well simulated in most of the advanced climate models. Therefore, the correction of the future projection in Arctic sea ice using observations is necessary. Here, using 1979-2020 analysis data, we calibrate the present-day melting rate of Arctic sea ice of climate models participating CMIP6 and apply for the future projection. Results show that the sea ice free year obtained from this calibration is earlier than that without calibration. This study may provide more accurate information in future Arctic sea ice change.

Variability in near-surface wind speed (NSWS) has significant impact for understanding the hydrological cycle, environmental governance, and renewable energy development. Global NSWS shows the stilling and reversal phenomenon before and after 2011, but lacks the attention on the discrepancy of monthly NSWS annual changes. Our study employed daily observational data from the China Meteorological Administration to investigate changes in monthly NSWS across China. We found a consistent decline in NSWS before 2011, after which March was marked by a hiatus in increasing wind. The hiatus phenomenon was mostly attributed to the southern displacement and weakening of the East Asian subtropical jet in March, which weakened the meridional temperature gradient and decreased transient eddy activity at mid-latitudes over China. Furthermore, the decline in atmospheric baroclinity suppressed anticyclonic anomalies, leading to a breakdown in NSWS. 

The Southern Ocean plays a crucial role in regulating global temperature through its interactions with the atmosphere, ocean, and sea ice. However, its thermostatic impact on greenhouse gas forcing is not well understood. This study conducted idealized CO₂ ramp-up and -down experiments and found that the hysteretic behavior of the air-sea interaction through declining sea ice has a substantial impact on the hysteresis of the Southern Ocean temperature. The reduced sea ice due to increased CO₂ supplies heat to the ocean through the sea-ice albedo feedback. In the Weddell Sea, where a large amount of sea ice exists, the reduction in sea ice makes open water, which leads to continuous heat release from the ocean to the atmosphere, thereby raising the temperature of the atmosphere. The loss of heat at the ocean surface is supplemented by accumulated heat in the ocean subsurface during the period of increased CO₂, thus inhibiting the formation of sea ice. These feedback processes keep the Southern Ocean warm for a prolonged period, suggesting that once human-induced warming starts, it may take a long time for the Southern Ocean to return to its initial state, or irreversible changes may occur. Acknowledgments: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-2018R1A5A1024958). The CESM simulation was carried out on the supercomputer supported by the National Center for Meteorological Supercomputer of Korea Meteorological Administration (KMA), the National Supercomputing center with supercomputing resources, associated technical support (KSC-2021-CHA-0030), and the Korea Research Environment Open NETwork (KREONET).

In this study, the relationship between atmospheric angular momentum's mass term (M_Ω) and water vapor has been investigated using four reanalysis datasets for the years 1950–2020: ERA5, ERA-20C, JRA55, and NCEP/NCAR. The four reanalysis datasets revealed a consistent positive long-term trend in global M_Ω anomalies, while also identified shifts in the trend from negative to more positive around 1970s. Meanwhile, analysis of global water vapor trends shows that there is an additional mass of water vapor, which results in variations in water vapor surface pressure (p_w) of about ~0.3 mbar. Changes in the contribution of water vapor to M_Ω after the 1970s were calculated based on the anomaly of M_Ωw (estimated from p_w) relative to the climatology value of 1970-1979. It can be shown that the M_Ωw anomalies tend to be positive after the 1970s period. In addition, the ratio of the M_Ωw anomalies relative to M_Ω suggests a change in the pattern of the contribution of water vapor to M_Ω, which is indicated by the fluctuation of this percentage every ten years. The highest contribution of water vapor to M_Ω occurred in the equatorial region during 2010-2020, with a ratio of M_Ωw to the global M_Ω at 25-50%. The stability of the relationship between M_Ω and water vapor is examined using the 20 year shifting window correlation. The relationship between M_Ω and water vapor is more consistent on annual and interannual scales, especially in the equatorial region, but there are inconsistencies for longer time scales. Inconsistency in the interdecadal scale is characterized by a lower statistical coherence from the mid-1980s to the 2000s, which indicates a decrease in the contribution of water vapor to M_Ω. These results suggest that water vapor alone is insufficient to account for the increase in M_Ω over the period.

Chemical reanalysis data, produced by assimilating observations into numerical model backgrounds, can provide systematically best-estimated and spatiotemporally continuous information on atmospheric chemical components. Nowadays, the chemical reanalysis data have been actively released by several institutes such as Copernicus Atmospheric Modeling and Monitoring Service (CAMS) reanalysis from European Centre for Medium-Range Weather Forecasts (ECMWF) and Modern-Era Retrospective Analysis for Research and Applications Aerosol Reanalysis (MERRAero) from National Aeronautics and Space Administration Global Modeling and Assimilation Office (NASA-GMAO). However, because such datasets are produced by using global chemistry transport models (e.g., C-IFS and GEOS-5/GOCART), they could not be suitable for air quality studies on a regional scale due to low spatial resolutions and poor parameterizations for regional atmospheric environments. Therefore, in this study, a chemical reanalysis system has been developed based on the Korean Air Chemistry Modeling system (hereafter ‘K_ACheMS RA’) for regional air quality study over northeast Asia and South Korea with high resolutions of 15×15 km² and 5×5 km², respectively. The K_ACheMS RA mainly consists of the following: i) the Community Multi-scale Air Quality version 5.2.1 (CMAQ v5.2.1) model updated for atmospheric conditions over northeast Asia by Gwangju Institute of Science and Technology (GIST) science team (i.e., CMAQ vG model); ii) the Weather Research and Forecasting (WRF) model with high resolutions of topography, land cover, and forest type datasets over South Korea; and iii) an advanced data assimilation system based on an ensemble Kalman Filter (EnKF). Ground-based observations, such as PM_2.5, CO, O₃, NO₂, and SO₂ in South Korea, China, and Japan, are assimilated into the CMAQ vG model every 3 hours. The evaluations of the produced reanalysis data by comparison with AIR KOREA observations in South Korea will be shown in this presentation.

Water vapor is one of the greenhouse gases which has strong impacts on the weather and climate. Despite of long-term observation by meteorological satellites, the physical/chemical process related to water vapor is not well understood. Therefore, continuous, accurate monitoring of atmospheric water vapor is essential for understanding its interaction with weather and climate. From this purpose, water vapor retrieval algorithm is newly developed in the visible channels (435.0 – 467.0 nm) for the Geostationary Environment Monitoring Spectrometer (GEMS) onboard the GEO-KOMPSAT 2B satellite. The algorithm is based on the direct fitting method, which is originally developed by Smithsonian Astrophysical Observatory (SAO). The GEMS Total Column Water Vapor (TCWV) algorithm conducts the two-step approach, which are comprised of spectral fitting of slant column densities (SCDs), and conversion of the retrieved SCDs to vertical column densities (VCDs) using air mass factors (AMFs). The TCWV retrieval algorithm is applied to GEMS observations for all months in 2021. The retrieval results show not only the monthly variations of water vapor but also its diurnal cycle over Asia. TCWV derived from GEMS observations are validated by comparing with AERONET precipitable water data. The results agree well with the reference data and the linear regression parameters tend to vary with time. If high quality water vapor product can be produced from this study, it is expected to improve our understanding of water vapor-related chemical and physical processes.

Air-suspended particulate matter under the size of 2.5 μm (PM_2.5) has been known to be a potential cause of asthma and increase of mortality even when exposed in the short-term (U.S. EPA, 2019). East Asian countries, where concerns on PM_2.5 has been high, have provided ground-based PM_2.5 observation data, but is spatially confined to the monitoring site. To overcome this limitation, aerosol optical properties (AOPs) derived from space-borne satellites have been used to estimate PM_2.5 concentrations over a broad area. In this study, we used a near-real-time random forest model with AOPs from a geostationary satellite, GEO-KOMPSAT-2B, to nowcast the ground-level PM_2.5 in East Asia. Hourly AOPs of the second generation Geostationary Ocean Color Imager-II (GOCI-II) retrieved by Yonsei aerosol retrieval algorithm were used as a major input, while meteorological and atmospheric chemical composition data from numerical models, and spatially and temporally inverse-weighted PM_2.5 were used as ancillary variables. AOPs of GOCI-II are provided in 2.5 km spatial resolution and 10 times hourly in the daytime, for which ancillary variables were collocated in terms of time and space. The synthetic minority over-sampling technique (SMOTE) was adopted to compensate for the low occurrences of high-PM_2.5 cases. Validation against ground-based PM_2.5 networks has shown a good model performance in terms of R² and RMSE. Hourly validation yielded a high R² in the daytime, and a higher R² in the winter. The predicted hourly PM_2.5 of this study can be used as an real-time indicator for outdoor air quality and public health.

Crop residues open burning has significant adverse effects on regional air quality, climate change and human health. Accurate crop residues open burning emission inventory can provide an important basis for objective and comprehensive analysis of biomass combustion, strengthening biomass combustion control and improving air quality. In this study, based on the VIIRS 375 m thermal anomaly products, land use data and high-resolution remote sensing images of sky maps, the spatiotemporal characteristics of crop residues open burning in central China were extracted and analyzed. Based on statistical models and satellite data, the high-resolution emission inventory in central China was estimated. The results showed that the number of crop residues open burning points showed a significant downward trend from 2012 to 2020, and the number of fire points remained at a low level after 2015. The peak period of straw burning is mainly concentrated during the harvest season. However, under strict control measures, staggered burning in some areas has led to several scattered small peaks of straw burning in spring and winter. The black carbon (BC), organic carbon (OC), sulfur dioxide (SO2), nitric oxide (NOX), carbon monoxide (CO), carbon dioxide (CO2), fine particulate matter (PM2.5), coarse particulate matter (PM10), ammonia-ammonia (NH3), methane (CH4) and non-methane volatile organic compounds (NMVOC) emitted by crop residues open burning were 34.84, 149.72, 41.06, 90.11, 2640.97, 78094.91, respectively. 485.17, 481.05, 35.21, 246.38 and 499.59 Gg. Among them, the largest contributor was rice, followed by wheat, rapeseed and maize, with a contribution rate of 35.34–64.07%, 15.78–34.71%, 9.12–25.56% and 5.69–14.06%, respectively.