A WMO Task Team on Tropical Cyclones and Climate Change was tasked with providing an updated assessment of the science of tropical cyclones and climate change. The assessment was published in the Bulletin of the American Meteorological Society in 2019 and 2020.The assessment team reviewed a series of case studies to assess whether detection and attribution of anthropogenic influence on tropical cyclones can be supported by historical observational data and other evidence. The majority of the author panel had low confidence that detection and attribution has been established, with one case (poleward migration of the latitude of maximum intensity in the western North Pacific) having low-to-medium confidence. They further concluded that "the balance of evidence" suggests that there are some cases where there may be detectable (unusual compared to natural variability) or attributable (to anthropogenic influence) changes emerging in observed tropical cyclone activity, even though confidence is currently low by conventional assessment standards. A few more recently published studies not assessed by the WMO team will also be discussed. Projections of tropical cyclone activity under a 2 degree Celsius global anthropogenic warming scenario include the following at the global-mean scale:  greater storm surge inundation levels due to rising sea level (high confidence); +14% increase in tropical cyclone precipitation rates; +5% increase in tropical cyclone wind speeds; and +13% increase in the proportion of category 4 & 5 tropical cyclones (the latter three projections each with medium to high confidence). The authors' confidence was lower for projections of: decreased global tropical cyclone frequency; and a reduction of tropical cyclone translation speed.

Many studies have suggested that the number and percentage of rapidly-intensifying (RI) tropical cyclones (TCs) appear to be increasing in the western North Pacific. A more important question is how many of these RI TCs make landfall in the East Asia coast (EAC), which is the objective of this study. In the presentation, it will be shown that along the EAC, both the number and percentage of landfalling TCs that experienced RI during their lifetime increased in the last two decades. Because of such increases, the annual power dissipation index (PDI) along the EAC also increased. Such increases result from the TC formation locations being further north and west so that if the TCs undergo RI, they are more likely to make landfall. Indeed, the location of lifetime maximum intensity of RI TCs has also shifted to nearer the coast. The main cause of the RI is an increased tropical cyclone heat potential and a reduced vertical wind shear, but the sea-surface temperature is not a contributor. Regionally, although the annual PDI increases along all sections of the EAC, those near the South China coast, East China coast, Korean Peninsula and Japan islands are all statistically significant but those for the Philippines and Vietnam are not.

Observations of lightning are commonplace worldwide and recent satellite instruments provide enhanced coverage. Lightning can be used for monitoring severe convection and precipitation, improving estimates of severe storm development, evolution and intensity, and hence provide early warnings for severe weather phenomena. In addition, lightning itself impacts the global climate by producing nitrogen oxides (NO_X), a strong greenhouse gas. For climate variability and change monitoring, lightning has been shown to follow trends and extremes in convective storms (convective mode, flash rate, flash extent) and track global surface air temperature on many natural time scales.  Given this relevance and potential as a climate variable, lightning has been added to the list of Essential Climate Variables (ECV) in the 2016 Global Climate Observing System Implementation Plan (IP), including a first attempt to define the requirements for climate monitoring with lightning measurements. A Task Team on lightning observations for climate applications completed an initial study that summarizes the work done and covers key aspects of needed lightning observations for climate applications. It explains the relevance of lightning observations, describes the current status of observations, discusses gaps and open research questions and provides suggestions for monitoring requirements for lightning, including metadata requirements. The Task Team is seeking relevant data sets to address climate questions using information about lightning. For instance, we started an initiative to extend lightning data into the distant past (many decades) using “thunder day” data. Identifying these and related data sets (e.g., Schumann Resonances) with regional or global coverage is an important aspect of our on-going work. Some of the data are naturally in the public domain, but most of the ground-based lightning network data are privately owned, sold commercially, and copyrighted. In this presentation on behalf of the task team we present a summary of our findings and discuss next steps.

The Australian Bureau of Meteorology (BOM) issues operational tropical cyclone (TC) seasonal forecasts for the Australian and South Pacific Ocean (SPO) regions in October each year, ahead of the Southern Hemisphere TC season (November to April). Various supervised statistical learning models have been developed to derive these TC seasonal forecasts, in which an important task is to determine, from many available predictors, the ones most informative to TC activity. Analysis of these models shows that the El Nino-Southern Oscillation (ENSO) related predictors such as the Dipole Mode Index, 5VAR index and the Southern Oscillation Index etc. are the most often used ones for seasonal TC forecasts in all regions. In particular, the 5VAR index is the first principal component of the monthly Darwin mean sea level pressure (MSLP), the Tahiti MSLP, the NINO3 sea-surface temperature (SST), the NINO3.4 SST and the NINO4 SST variates; and was developed at the BOM’s National Climate Centre (NCC). In this study, we investigate the use of sliced inverse regression (SIR) method, based on all available predictors excluding the 5VAR index, for seasonable TC activity modelling and forecasting. Statistical properties of the SIR method ensure that the predictor derived from the first SIR-based principal direction captures more information on TC activity than that derived from the first principal component score. Our applying the SIR method to the 1970/71 to 2014/15 Australia and SPO TC data gives better seasonal TC predictions than the principal component regression (PCR) method.

Tropical cyclones (TCs) significantly impact people, economies, and the environment in coastal areas. The intensity, tracks, and occurrence frequency of TCs are necessary for estimating the risk of damage. However, the probability statistics methods for risk assessment fail to conclude reliable results because a shortage of observation data of TCs exists at both spatial and temporal scales. This study presents a synthetic algorithm Tropical Cyclone Stochastic Model (TCSM), based on the Monte Carlo Markov Chain approach, developing a dataset representative of 10,000 -100,000 years of TC activity over the western North Pacific (WNP) under present climate conditions from 1979 to 2018. To estimate TC intensity precisely, this study employs a simple statistical method to correct the intensity bias of the simulated TCs and shows that model results preserve the statistics as found in the historical dataset. The conditional pdfs used in TCSM are presented in a modular and flexible way so that it is easy to analyze the influence of different environmental variables on the TC simulation. The environmental variables are determined by historical averages and random terms. In this study, three sets of experiments were designed to compare four environmental variables, including steering flow, geopotential height, sea surface temperature, and vertical wind shear. In brief, the TCSM dataset not only can be used for TC hazard assessments and risk modeling in TC-prone regions but also for discussing the possible effects of environmental factors on TCs.

Information Entropy is a statistics tool that helps us to explain the amount of uncertainty a variable has across its given domain, with values more uniformly distributed achieving larger entropy and more concentrated values achieving smaller entropy. For a given location the amount of precipitation it receives in a year can be placed in different categories (months/seasons) based on when it occurred in the year. Each category for a given location will have its own distribution with spatially specific parameters, the joint distribution of these categories can be considered a proportion distribution. For a given location how concentrated its yearly precipitation is distributed over the different categories can be measured by information entropy. We can then obtain an approximation of the distribution for the entropy of precipitation based on the proportion distribution. We can also look at the long term behaviours of precipitation entropy for Australia and how it has changed for certain locations and what are the consequences of this changing entropy.

Since November 2017, the Southeast Asia Regional Climate Centre-Network (SEA RCC-Network) has been providing climate information for Southeast Asia. As part of this network, Meteorological Service Singapore is leading the Long-range Forecasting (LRF) Node, which delivers regional seasonal climate outlooks. On top of the standard rainfall tercile and anomaly plots, the LRF node also provides outlooks for more extreme rainfall, such as the probability of exceeding the 90^th percentile, with the goal to provide useful information on upcoming climate related hazards. This work will review the products provided, along with their associated skill assessment. In particularly the extreme dry conditions in 2019, as well as the heavy rainfall in January 2021 will be examined to highlight the benefits and some of the limitations of the seasonal outlooks for climate related hazards in Southeast Asia.

The eastern coastal area of China is one of the most densely populated areas in the world, sufferring from serious air pollution. Trace metal elements (TMEs), especially toxic metals, can cause substantial damage to the environment and human health. 66 atmospheric aerosol samples and 42 atmospheric precipitation samples were collected from May to November 2020 in Qingdao, a megacity in Eastern China. The total and soluble concentrations of nine TMEs (Al, Fe, Zn, Pb, Cu, Ni, As, V, Cd) in the samples were analyzed by using Inductively Coupled Plasma-Mass Spectrometry (ICP-MS). The total concentrations of TMEs in atmospheric aerosol samples were found to be in an order Al > Fe > Zn > Pb > Cu > Ni > As > V > Cd, which has a similar trend over the soluble concentrations. Among these elements, Al has the lowest solubility about 5%, while As, Pb, Zn has the highest solubility about 60%. Except for metals (As, Cd, V) that are not detected because of their very low concentrations, the concentrations of TMEs in the atmospheric precipitation samples is generally 1~2 orders of magnitude lower than those in aerosol samples. Take aluminum for example, the total and soluble concentrations of Al in atmospheric aerosol samples were 3.63±3.6μg/m³ and 0.14±0.15 μg/m³, respectively, while the concentrations of Al in atmospheric precipitation samples were 0.35±0.45μg/m³ and 0.01±0.02μg/m³, respectively. Significant seasonal differences were showed for the concentrations of TMEs in aerosol and precipitation samples, and generally showed higher concentrations in autumn than in summer. The solubility of TMEs is controlled by many environmental factors, such as air mass source, meteorological factors and precipitation acidity. And the acidity of rainwater was the major factor for controlling the solubility of TMEs in precipitation． 

Size distributions of atmospheric aerosols vary significantly in time and space through various aerosol dynamic processes such as new particle formation, condensation/evaporation of semi- and low-volatile gases, coagulation, dry deposition. Atmospheric aerosols affect climate directly by scattering solar radiation and indirectly by altering clouds’ properties and lifetime. In addition, particulate pollution has been one of the most severe social problems due to its long list of adverse health effects. In this respect, it is important to understand the variations in aerosol size distributions because not only the sources and sinks but deposition ratios in the human respiratory tract of aerosol modes are different from each other. In this study, we extracted constituent aerosol modes from size distribution data obtained from long-term measurements (2008 – 2015) with scanning mobility particle sizer (SMPS, TSI model 3776) in Jeju Island, Korea by fitting size distributions with multiple lognormal distribution functions. Also, we investigated how the variations in aerosol modes are linked to related factors such as meteorology and precursor gases. Additionally, we used the international commission on radiological protection (ICRP) respiratory deposition model to calculate the fraction of aerosols deposited in the pharynx, bronchi, and alveoli for each mode.

Vehicular emissions (of both gaseous and particulate pollutants) have been reduced due to strict regulatory policies and advances in technology. However, traffic emissions are still one of the major sources of severe particulate air pollution (represented by high PM_2.5 levels) in urban areas. Numerous near-roadway studies have been conducted with multipoint-stationary or mobile measurements to investigate the impacts of roadway emissions on surrounding environments. However, multipoint stationary measurements were conducted mostly for criterion gas pollutants and with a somewhat small number of points due to limited resources. In addition, mobile measurements have limitations in that the snapshots of pollutants distributions are hard to be obtained, monitoring periods are limited, and sometimes certain vehicles ahead can affect dominantly the pollution levels detected. In this respect, intensive field campaigns were conducted combining mobile, multipoint stationary, and vertical measurements for gaseous and particulate pollutants near one of the busiest highways in Korea. Here, we present the preliminary results focusing on multipoint stationary measurements. The multipoint stationary experiments (5 points) include criterion pollutants (CO, NO, NO₂, O₃, PM_2.5, CO2) measured with custom-built sensor nodes, particle size distributions (10 – 300 nm with custom-built scanning mobility particle sizers and 0.3 – 10 mm with custom-built optical particle counters), in-situ weather conditions with two 3D sonic anemometers, and traffic rates and fleet compositions. Besides, PM_2.5 chemical compositions were measured with Aerosol Chemical Speciation Monitor (ACSM, Aerodyne).

The primary source of air pollution in urbanized areas is traffic emissions. Air pollutants emitted from vehicles are major precursors of ozone and fine particle formation. Recently, low-cost air quality sensors have been widely evaluated in the real atmosphere to obtain highly resolved air quality data in space. In this study, we built a low-cost air quality sensor network in a tunnel to estimate vehicular emissions under real road conditions, which was challenging in the past with expensive conventional instruments due to limited resources. A Tunnel provides excellent experimental environments for investigating vehicular emissions of air pollutants in that traffic-related pollutants are trapped in tunnels, minimizing dilution with ambient air, and fleet compositions and driving conditions reflect the real roadways. Experiments were conducted with a mobile platform equipped with CO₂ and CH₄ analyzer (G2201-I, PICARRO) and Selected Ion Flow Tube-Mass Spectrometry (SIFT-MS) for VOCs speciation. Simultaneously, continuous stationary monitoring for criterion pollutants (CO, NO₂, NO, O₃, PM_2.5, and PM₁₀) and temperature and humidity was also conducted at 7 locations inside the tunnel with low-cost sensor nodes. In the middle of the tunnel (~700 m and 1600 m from the entrance in winter and summer, respectively), a 3D-sonic anemometer was installed to examine the effects of the wind speeds and direction as well as turbulence characteristics on pollutant distributions. Here, we present preliminary results of emission characteristics of various fleet compositions, focusing on stationary sensor network measurements.

To investigate fine particles' physical and chemical characteristics for the winter season in Korea's coastal megacity, we conducted a month-long intensive field campaign in January of 2021 in Busan (the 2^nd largest city in Korea). The monitoring site can be affected from various source areas because it is located 2 km northeast of the port and 1.5 km east and southeast from the dense small-sized industrial areas. Besides, commercial urban areas are widely distributed north of the site, and relatively clean marine air comes from the east. The measurements include particle size distributions (105 channels, 12.6 – 532.8 nm with a scanning mobility particle sizer (SMPS3938L52, TSI) and 0.3 – 10 mm with an optical particle sizer (Model 3330, TSI)), near real-time inorganic ion speciations of PM_2.5 (PILS-IC, Metrohm 883), black carbon (MA 350, Aethalo Inc.), PM_2.5 mass concentration (DustTrak 8530, TSI), gaseous precursors (NO/NO₂, NH₃, SO₂ ; Thermo Scientific gas analyzers), in-situ meteorology and atmospheric turbulence (3D sonic anemometer, Gill WindMaster Pro). Two to four aerosol modes were extracted from the observed particle size distributions with a lognormal fitting method. In this study, we investigate the characteristics of aerosol physical and chemical properties with respect to prevailing winds by quantitatively relating the variations in individual mode properties (like modal size, geometric standard deviations, number, and mass concentrations) to various factors like wind directions, chemical compositions, and precursor gases, atmospheric turbulence parameters, and others. Here we present the preliminary results of those analyses.

In this study, we produced high-resolution (1km) daily data sets of surface air temperature and precipitation based on the PRIDE(PRISM-based Dynamic downscaling Error correction) model for South Korea during the 18-years from 2000 to 2017. The PRIDE model is a statistical downscaling model for the production of high-resolution climate change scenario data based on MK-PRISM(Modified Korean-Parameter elevation Relationships on Independent Slopes Model) with the QDM(Quantile Delta Mapping) method. The simulation data of five RCM(Regional Climate Model) were also used as input data for the PRIDE model, which are forced by the CMIP6 participating model UK-ESM as the boundary condition under historical period (2000-2014) and SSP5.8-5 period (2015-2017). Here, we compared RCM data and PRIDE data with MK-PRISM data in terms of ensemble mean and ensemble spread. Results show that the PRIDE model effectively eliminates systematic error in the RCM up to 63.0% for daily average temperature, 72.2% for daily maximum temperature, 68.2% for daily minimum temperature, and 28.7% for daily precipitation when evaluated from the RMSE perspective. Overall the ensemble spread of the PRIDE model is significantly decreased from 1.46℃ to 0.36℃ for daily temperature and from 2.0 mm/day to 0.72 mm/day for daily precipitation compared to the RCM ensemble spread. These indicate that the largest systematic error of the RCMs is effectively removed in the PRIDE model. Furthermore, we will present prediction results of climate change under the SSP scenarios.

This study examined historical and future tropical cyclone (TC) activities over East Asia using five regional climate models (RCMs) with 25 km horizontal resolutions, which were forced by three global climate models (GCMs). The experiments were based on the Coordinated Regional Climate Downscaling Experiment (CORDEX) framework. We investigated historical (1981-2005), mid-21 century (2026-2050), and late-21 century’s (2076-2100) mean of TC genesis, track, and intensity under RCP 2.6 and RCP 8.5 scenarios based on optimized TC tracking thresholds. Multi-model ensembles (MME) with equal weighting averaging (MME-EWA) and performance-based ensemble averaging (MME-PEA) methods were adopted to avoid models’ systematic bias and not to be significantly influenced by some poor models’ performances. The ensembles were conducted based on the same GCM forcing RCMs - HadGEM3-RA, CCLM, SNURCM, and RegCM forced by HadGEM2-AO, HadGEM3-RA, CCLM, SNURCM, and WRF forced by MPI-ESM-LR, and WRF and RegCM forced by GFDL-ESM-2M. In most ensembles, it was noteworthy that TC activities in the mid-latitudes seem to be increased due to the changes in environmental fields (e.g., vertical wind shear, subtropical high, and sea surface temperature).  

In the framework of the CORDEX-East Asia, evaluation simulations using high-resolution regional climate model (SNURCM) with ~25km (Phase2) grid scale have been conducted. In this study, we investigate whether the higher-resolution regional climate model (RCM) can generate added values for summer mean precipitation, large-scale circulation, and extreme precipitation compared to those with lower-resolution (~50km, Phase 1). In addition, added value index is used to quantitatively analyze the abilities of fine- and coarse-resolution RCM. Hence, sets of Phase 1 and Phase 2 simulations of the RCM are compared to observations in the East Asia region. In SNURCM simulations, positive (negative) added value of summer mean precipitation is reproduced over most ocean (land) region of East Asia in fine-resolution simulation. Extreme precipitation over Korea and Japan is reasonably reproduced in Phase 2 simulations because the simulations of typhoons and East Asia summer monsoon are improved. As a result, more positive added values for the intensity and spatial distribution of precipitation over East Asia are generated as the horizontal resolution of RCM increases.

This study carries out a multi-model inter-comparison project about evaluating the short-term (2-week) model bias under an active MJO event, and we will analysis one MJO event in 2017-18 boreal winter with hindcast approach for two climate models (TaiESM and SPCAM) and two weather models (NTUGFS and MPAS). For participating models, were initialized with daily ERA-Interim reanalysis data and prescribed SST, constructing 14-day hindcasts for each run in this MJO event period (2017/12/24 - 2018/01/21). The native spatial resolution for TaiESM and SPCAM are 0.9° × 1.25° in horizontal and with 30 vertical levels, SPCAM replaces the cumulus parameterization in each GCM grid by a 2-D (x-z) cloud resolving model with 4-km resolution and 28 vertical levels. For NTUGFS/MPAS, the horizontal solution is in 15 km × 15 km and with 72/55 vertical levels, and the model outputs will interpolate on to 0.9° × 1.25° grid for analysis. The results based on the day-3 hindcasts show that MPAS has better performance with simulated mean rainfall and OLR over the Indo-pacific region in this MJO event from the spatial pattern correlation coefficient with observation data. The observed lower OLR around the Maritime Continent area is overestimated in NTUGFS/SPCAM simulations, suggesting stronger convection over this area in those models. In contrast, MPAS/TaiESM produce positive OLR bias over this area. Models with negative/positive OLR bias also have higher/lower precipitation variability in this MJO event. Vertical profiles of moisture to the west of Maritime Continent (90°E-100°E, 10°S-10°N) in TaiESM/SPCAM show low-tropopause (1000 hPa-900 hPa) drying/moistening biases compared with observed profiles. We will examine the evolution of the biases after day-3 from the perspectives of convectively-coupled wave development.

The development and characteristics of the atmospheric boundary layer were investigated in a coastal areas using the wind profiler radars on the east coast of South Korea. For comparison of the atmospheric boundary layer height, we conducted a rawinsonde intensive observation from September 8 to 10, 2015, at the Uljin weather station operating wind profiler radar. Subsequently, through the wind profiler radar data collected in 2015, the development of the atmospheric boundary layer was continuously analyzed for one year. Finally, sea breeze was observed and its characteristics were investigated. The coastal zone is the area where the land meets the ocean, and the atmospheric boundary layer is affected by local winds such as the sea and land breeze. At the Uljin, the wind speed spectrum peaked at frequencies of 1 × 10^-5 Hz or 9 × 10^-4 Hz at altitudes of 0.26, 0.54, 0.83, 1.12, and 1.40 km corresponding to periods of 1 day and 12 h, respectively, which refer to the sea–land breeze. The atmospheric boundary layer height estimated by the refractive index structure constant C_n² of the wind profiler radar, fluctuated from 1.0 to 2.5 km between 0900 and 1800 LST during the daytime, and was in good agreement with that of rawinsonde. During one year in 2015, the mean atmospheric boundary layer height fluctuated from 0.8 to 2.3 km. When sea breeze was blown, the atmospheric boundary layer height was around 1 km during the daytime and did not change significantly.

To strengthen the effectiveness of disaster prevention and response to debris flow disaster, carrying out disaster prevention and preparation of debris disaster before, during and after the flood season. And to assist in the communication, management and education of various debris flow operations, and improve the effectiveness of disaster prevention and response operations.The effective disaster reduction is based on the correct debris-flow disaster preparedness and the real-time information communications between the disaster area and the rescue-control center in the typhoon or heavy rain attacked. This study first briefly proposed and designed a Real-Time Debris-Flow Disaster Network, which is composed of the emergency assessment system, the application servers, and the decision support server based on the disaster communications. Secondly, Unmanned Aerial Vehicle (UAV) are commonly used in disaster investigation and photographing in recent years for the first-line disaster investigation. Aerial images are generated as 3D models or numerical surface models to represent the characteristics of large-scale disasters, which has the advantage of intuition. Finally, this study concludes the countermeasures against the disaster, the database has collected the pre-analyzed nearly 100 potential disaster potential maps during 2014~2020 in the southwest area in Taiwan. According to the results, to achieve effective debris-flow prediction with high degrees of accuracy to protect human life and property.The authors are grateful for the financial support by the Tainan Branch, Soil and Water Conservation Bureau, Council of Agriculture, Executive Yuan, ROC under grant number 110WS030.

Northern Hemisphere extratropics recorded numerous extreme climate anomalies in boreal winter (January-March, JFM) of 2020. These anomalies, caused by record high positive Arctic Oscillation (JFM AO index of 2.8), significantly impacted the human lives and ecosystems. On the other hand, all of the well-established autumn precursors in autumn of 2019 pointed towards wintertime AO phase being negative. Indeed, the atmosphere was developing this way until late December, when sudden shift towards strong positive AO occurred in the troposphere and then spread into the lower stratosphere. These anomalies steadily enhanced throughout JFM, resulting in the extreme positive AO event. In this study, we show that the strong positive AO event was a result of the destructive interference of the anomalous planetary waves with the climatological ones. This interaction led to the flattening of the planetary waves, and the enhancement of the polar vortex.
This work was carried out with the support of the "Cooperative Research Program for Agriculture Science & Technology Development (Project No. PJ01489102)" Rural Development Administration, Republic of Korea. 

The cube root of energy dissipation rate (EDR) as a standard reporting metric for turbulence is estimated using 1-Hz quick access recorder data from Korean-based national air carriers including two different types of aircraft (B737 and B777) in 2012. EDR is estimated using zonal, meridional, and derived vertical wind components, and derived equivalent vertical gust (DEVG). The wind-based EDR is estimated by employing three different methods based on (i) second-order structure function, (ii) power spectral density (PSD) considering the Kolmogorov’s -5/3 dependence, and (iii) maximum-likelihood estimation using the von Kármán wind model. The DEVG-based EDR is estimated using (iv) lognormal mapping technique and (v) predefined parabolic relationship between observed EDR and DEVG. First, second-order structure functions were computed for each wind component within defined inertial subrange. For methods (ii) and (iii), Fast Fourier Transform is applied to each wind component, then the individual PSD is computed over a 2-minute time window. Finally, two different EDR estimates are computed by employing (ii) the Kolmogorov-scale slope or (iii) prescribed von Kármán wind model within the inertial subrange. Resultant EDR estimates from five different methods follow a lognormal distribution reasonably well. Mean and standard deviation of log-scale EDR estimates are somewhat different to those from a previous study using higher sampling rate (8Hz) of in situ aircraft data. However, 1-Hz EDR estimates show good agreement with four selected severe turbulence events with respect to turbulence intensity. This suggests that EDR estimates using relatively low sampling rate of flight data like the Automatic Dependence Surveillance-Broadcast can be useful for turbulence reporting, which will eventually provide more widespread database of aviation turbulence observations. Acknowledgement: This research is supported by the Korean Meteorological Administration Research and Development Program (KMI2020-01910), and by the National Research Foundation of Korea Research and Development Program (NRF-2019R1I1A2A1060035).

In line with rapid urbanization in the world, the urbanization rate in Korea is at least 80% and is expected to reach 86% in 2050, which expects more traffic congestion in urban areas in the future. According to the Korea Transport Institute (KOTI) analysis in 2015, the cost of the traffic congestion amount was about 33 trillion won. In this situation, Urban Air Mobility (UAM) is an emerging technology to solve this problem. In Korea, policy preparation and research and development are underway to commercialize drone deliveries and drone taxis (UAMs) by 2025. Drone delivery and drone taxi will be operational at altitudes between about 1 km and 50 m above the ground level. This altitude corresponds to the mixed layer (ML) in the planetary boundary layer (PBL). A large amount of micro-scale turbulence occurs in the ML due to convection and wind shear, which is a significant hazard to UAM operations. This study aims to analyze turbulence characteristics in the lower atmosphere in urban areas for safe UAM operation. For the analysis, the vertical profiles of wind (u, v, w) from windlidar and turbulent flux observation from sonic anemometers at six observation stations in the Seoul metropolitan area were used for the period from 2017 to 2019. In addition, the characteristics of turbulence occurrence according to the characteristics around the observation point (residential, commercial, agricultural) were also analyzed. The most important feature is that the frequency of strong turbulence in residential/ commercial areas was higher than that of agricultural areas. In the agricultural area, the lower atmosphere was stable at nighttime, while turbulence in residential/commercial areas was high at nighttime due to urban heat source. Details will be presented at the conference. Acknowledgement: This research is supported by the Korean Meteorological Administration Research and Development Program (KMI2020-01910).

Arctic amplification from global warming causes the zonal wind to reduce as does meridional temperature gradient because of the thermal wind relationship. It causes more extreme weather in mid-latitude, and also affects to the aviation industry, because the jet stream influences flight times, routes and fuel consumption. In this research, we study the changes of the upper-level jets in the Northern Hemisphere using the model outputs from phase 6 of the Coupled Model Intercomparison Project (CMIP6). The CESM2, NorESM2-MM, and MPI-ESM1.2-HR climate models are selected as they have the lowest absolute error of upper-level winds in the CMIP6 historical simulation. We calculate the trajectories of daily flights across the Pacific and North Atlantic Oceans using the historical run and four different Shared Socioeconomic Pathways (SSP)-based scenarios (SSPs; SSP1-2.6, SSP2-4.6, SSP4-6.0, and SSP5-8.5) from the selected models. It is found that the characteristics of upper-level jets are changed in different regions and scenarios: northward shift of the East Asian Jet, eastward propagation of the Eastern Pacific Jet, and northward or eastward shift of the North Atlantic Jet. Zonal winds locally increase in most flight routes, and there are statistically significant delays in round-trip flight times compared with the results of historical simulation. Consequently, it is expected to have significant increase of the Carbon dioxide emissions for long-distance flights. Acknowledgment: This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2019R1I1A2A01060035). JL’s work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore National Laboratory under Contract DE-AC52-07NA27344 via the Regional and Global Climate Modeling Program.

Recently, the use of infrasound data for atmospheric remote sensing has become possible. Infrasound waves generated near the surface penetrate the atmosphere where their propagation strongly depends on vertical wind and temperature gradients. These gradients create waveguides that duct the infrasound waves between the ground and different atmospheric layers, providing the ability to assess the atmospheric state from the infrasound propagation. This study investigates a velocity spectrum analysis (vespagram-based) approach providing a multi-direction comparison between simulated microbarom soundscapes and infrasound data recorded at ground-based stations. Discrepancies in directional spectra between vespagrams and modelled soundscapes would then indicate biases in atmospheric wind fields used for simulations. The microbarom radiation model accounts for both the source radiation dependence on elevation and azimuth angles and finite ocean-depth. The effects of the atmospheric ducting from oceanic source regions to stations are estimated using a semi-empirical attenuation law. This study uses infrasound data recorded in northern Norway during 2014-2019 to generate vespagrams presenting signal power depending on time and back-azimuth direction. The processed infrasound data and the modelled microbarom soundscapes at the station are compared in different aspects: i) signal amplitude, ii) azimuthal distribution and iii) signatures of extreme atmospheric events such as sudden stratospheric warmings (SSWs). The results demonstrate the ability of vespagrams to display the time-dependent microbarom azimuthal distribution, amplitude, and frequency on a seasonal scale, as well as changes during SSWs. The vespagram-based analysis is computationally low-cost and can uncover microbarom source variability also when infrasound from several competing microbarom sources impinge the station. There is also a potential for near-real-time diagnostics of atmospheric model products and microbarom radiation models, especially when applied to multiple stations, e.g., exploiting the CTBTO International Monitoring System network.

Meteor wind observations of the migrating semidiurnal tide (SW2) made by a longitudinal chain of high-latitude SuperDARN radars are compared against simulations using a mechanistic primitive equation model. The model is a three-dimensional, non-linear and time-dependent spectral model having a model top at ~430 km altitude. The modeled background zonal winds and temperatures are nudged to daily mean data from the Navy Global Environmental Model - High Altitude (NAVGEM-HA) meteorological analysis system up to thermospheric heights (~95 km altitude). The SW2 tide is forced using 3-hourly temperature tendency fields from the Specified Dynamics Whole Atmosphere Community Climate Model With Thermosphere and Ionosphere Extension (SD-WACCMX). To compare model to observation, the model is sampled according to the meteor echo distribution of the SuperDARN radars at the locations of available measurements for the years 2014 and 2015. Our model accurately reproduces most of the key features of the observed seasonal cycle in the SW2 phase and amplitude. Experiments show that the SW2 tide forced in the troposphere is especially sensitive to the background zonal winds and temperatures. The NAVGEM-HA background atmosphere enhances the tropospheric SW2 component by as much as a factor of 8 compared to zero-wind simulations, while also giving rise to rapid phase transitions observed around the equinoxes. In addition, the inclusion of an idealized surface friction profile is found to significantly delay the phase of the tropospheric SW2 component, which can strongly impact the net simulated tide at mesosphere-lower-thermosphere altitudes through interference with the SW2 tide forced in the stratosphere.

We investigate the possible detection of charged dust in incoherent scatter observations with EISCAT. The mesosphere and lower thermosphere contain small nanometer-sized dust particles that form from the cosmic dust material that enters Earth atmosphere. These meteoric smoke particles form in the meteors that are observed at 80 to 120 km altitude. They are transported in the ambient atmosphere and grow in collisions. In addition, also ice particles form during summer at mid and high latitudes. Both, the smoke and the ice particles, denoted as dust particles, influence chemical processes through charge interaction. They also participate in the formation of radar echoes that form in the presence of turbulence at altitudes close to the mesopause. There are, however, little direct observational results on the number density of the dust particles and their surface charge. The incoherent scatter that is observed with high-power large-aperture radar is influenced by coupling behavior of electrons to ions as well as charged dust particles. Collisions with neutral atmosphere further influence the incoherent scatter and the collision frequencies are high in this region of the atmosphere. The presence of charged dust can be inferred from incoherent scatter observations under certain conditions and provided that electron number density is sufficiently high to generate back scatter signal that can be analyzed. A detection of charged dust in incoherent scatter observations with EISCAT was reported previously by another group. We present here comparisons of radar measurements using the EISCAT radars near Tromsø (69.58° N and 19.23°E) and on Svalbard (78.15° N and 16.02° E) with model calculations to examine the statistical presence of dust in these radar signals.

Kelvin-Helmholtz instability (KHI) is one of significant sources of atmospheric turbulence; KHI generates Kelvin-Helmholtz billows (KH billows) and subsequent turbulence, which mix up heat, momentum, and minor constituents. The purpose of this study is to examine the structure of the KH billows in the Antarctic troposphere and lower stratosphere using the PANSY radar, an MST radar, at Syowa Station (69.0°S, 39.6°E) and elucidate the formation mechanism. A frequency-domain interferometry (FDI) technique using five different frequencies of transmitted waves was employed. The FDI technique provides echo power and vertical winds with a resolution higher than the transmitted sub-pulse length. Two FDI observation campaigns were performed over 10 days each, that is, from March 14-24 and August 2-12, 2019. In this presentation, we will show results focusing on the KH billows detected at 1100 UT – 1730 UT on 21 March around z=9.0 km (A) and 1100 UT – 1300 UT on 22 March around z=7.0 km (B). In case A, a wave-like structure (~2.0 km) that is dominant in the horizontal winds was observed in association with the KH billows. This wave-like structure is likely due to an inertia-gravity wave (GW) propagating energy upward and having a northeast- or southwestward wavenumber vector (~150 km) and slow ground-based phase velocity (~2.3 m). These dynamical properties are consistent with the orographic gravity wave generated in the severe southwestward wind near the surface. In case B, strong vertical shear of meridional winds localized on z~7.0 km was observed. This structure corresponds to a strong wind part of a well-developed low-pressure system approaching Syowa Station. We will also discuss the mechanism of the low-pressure system development accompanying such a strong wind region in the upper troposphere, based on the numerical model simulations.

We present a study of migrating and non-migrating tidal winds observed simultaneously by three meteor radars situated in the southern equatorial region. The radars are located at Cariri (7.4° S, 36.5° W), Brazil, Kototabang (0.2° S, 100.3° E), Indonesia and Darwin (12.3° S, 130.8° E), Australia. Harmonic analysis was used to obtain amplitudes and phases for diurnal and semidiurnal solar migrating and non-migrating tides between 80 and 100 km altitude during the period 2005 to 2008. They include the important tidal components of diurnal westward-propagating zonal wavenumber 1 (DW1), diurnal eastward-propagating zonal wavenumber 3 (DE3), semidiurnal westward-propagating zonal wavenumber 2 (SW2), and semidiurnal eastward-propagating zonal wavenumber 2 (SE2). In addition, we also present a climatology of these wind tides and analyze the reliability of the fitting through the reference to Whole Atmosphere Community Climate Model (WACCM) winds. The analysis suggests that the migrating tides could be well fitted by the three different radars, but the non-migrating tides might be overestimated. The results based on observations were also compared with the Climatological Tidal Model of the Thermosphere (CTMT). In general, climatic features between observations and model migrating tides were satisfactory in both wind components. However, the features of the DW1, DE3 and SW2 amplitudes in both wind components were slightly different from the results of the CTMT models. This result is probably because tides could be enhanced by the 2006 northern hemisphere stratospheric sudden warming (NH-SSW) event.

We report an analysis of the response of tides in neutral atmospheric mesospheric winds to recurrent geomagnetic activity in 2005 over Antarctica. The mesospheric winds were observed by the Davis meteor and medium frequency radars (68.5°S, 77.9°E; magnetic latitude, 74.6°S). The zonal component of the daily prevailing winds showed a westward increase as the geomagnetic activity increased, while the meridional prevailing winds did not show a clear response. The semidiurnal tides responded greatly to geomagnetic activity, with significant increases in amplitude. The zonal and meridional semidiurnal tides both showed a clear upward propagating phase but responded differently to geomagnetic activity. The amplitude of the meridional component of the diurnal tides increased significantly, while the zonal diurnal tidal amplitude showed no apparent change. These results indicate that geomagnetic activity can significantly influence mesospheric dynamics.

A 31 MHz meteor radar at Svalbard has been used to study polar summer mesospheric echoes (PMSE) during summer 2020. Data from 19 July is presented for examples and there is good agreement for general layer morphology between the meteor radar and a co-located 55 MHz MST radar. New techniques have been developed to better characterize the PMSE layer and its internal dynamics using meteor radar. Key to this is the analysis of PMSE layer Doppler as a function of range, which enables fine structure to be observed, as well as zenith angle to different range/Doppler bins to be estimated. Wind speed in the layer has been determined from the range-Doppler profile of PMSE return, which is in good agreement with estimates obtained from meteor trail drift radial velocities. Combined with the antenna polar diagram of the radar’s individual antennas, the range-Doppler profile of PMSE return has also been used to estimate the aspect sensitivity of the layer, which yielded a value of 6.6 ± 2.8°. A comparison of meteor trails inside and outside PMSE showed the possibility of reduced meteor trail radar echo decay times, but further work is needed to conclusively prove a relation between the presence of PMSE and reduced echo decay times. The use of raw time-series data originally collected for meteor detections to study PMSE means that additional analysis functions can be performed without compromising the original purpose of meteor radars.

Since the mid-1990s, a simple form of Machine learning (ML), namely artificial neural networks (ANN), has been used extensively in climate forecasting. However, ML has often been criticized by forecasters and end-users as being a “black box” because of the perceived inability to understand how ML makes its predictions. Recently there is a strong interest to explore the use of interpretable ML methods for climate prediction. The generation of the multi-model ensemble (MME) is a well-accepted approach to improving the skill of climate forecasts from individual GCMs. This talk will explore the question: whether ML-based multi-model ensemble methods add value over existing methods to improve the skill of the seasonal predictions where the sample size is a big constraint to train the ML network? The primary goal of the current study is to create a guideline for the proper usage of ML in the multi-model ensemble and identify potential value-added over regular methods. A state-of-art single-hidden-layer feed-forward neural network (SLFN) based non-linear regression method viz., Extreme Learning Machine (ELM) employed on the coupled GCMs of the North American Multi-Model Ensemble (NMME) to make the multi-model ensemble. The skill and interpretability of the proposed method compared other regular MME methods, including simple arithmetic mean and singular value decomposition-based multiple linear regressions. The results will be discussed for the South Asian monsoon.

In the era of modern science, scientists have developed numerical models to predict and understand theweather and ocean phenomena based on fluid dynamics. While these models have shown high accuracyat kilometer scales, they are operated with massive computer resources because of their computationalcomplexity. In recent years, new approaches to solve these models based on machine learning have beenput forward. The results suggested that it be possible to reduce the computational complexity by NeuralNetworks (NNs) instead of classical numerical simulations. In this project, we aim to shed light upondifferent ways to accelerating physical models using NNs. We test two approaches: Data-Driven StatisticalModel (DDSM) and Hybrid Physical-Statistical Model (HPSM) and compare their performance to theclassical Process-Driven Physical Model (PDPM). DDSM emulates the physical model by a NN. The HPSM,also known as super-resolution, uses a low-resolution version of the physical model and maps its outputsto the original high-resolution domain via a NN. To evaluate these two methods, we measured theiraccuracy and their computation time. Our results of idealized experiments with a quasi-geostrophicmodel show that HPSM reduces the computation time by a factor of 3 and it is capable to predict theoutput of the physical model at high accuracy up to 9.25 days. The DDSM, however, reduces thecomputation time by a factor of 4 and can predict the physical model output with an acceptable accuracyonly within 2 days. These first results are promising and imply the possibility of bringing complex physicalmodels into real time systems with lower-cost computer resources in the future.

In this study, idealized experiments are simulated by a cloud-resolving model based on the vector vorticity equation (VVM). By controlling the horizontally uniform large-scale moistening and the presence of the wind shear, the aggregated convection can be recognized by statistical characteristic of the distribution of cloud size. A deep learning approach are applied to identify the characteristics of convective organization with only the limited area of a snapshot from the model output, which contains only the spatial distribution of the cloud water path. The results suggest that given a limited area of a snapshot of the cloud water path, the cloud coverage and perimeter of clouds are the important factors for diagnosing the convective convection. In addition, the importance of the perimeter is greater (smaller) as the cloud coverage are small (large). A direct application of this study is the diagnosis of the organized convection with the limited parameters and areas of satellite images. Also, the approach of exploring and selecting the hidden variables in the deep neural network in this study can be applied as a framework in the future study of convective aggregation.

Hirosaki University has installed an X-band weather radar (known as “Hirodai-Shirakami radar”) on the roof of its eleven-story building and has been observing precipitation since 2014. Tsugaru Plain, where Hirosaki University is located, is in the northernmost part of Honshu, Japan, and known for its heavy snowfall. This area is surrounded by mountains and is located near the boundary between the two observation ranges of the C-band Doppler radar of Japan Meteorological Agency, making it difficult to detect low-level snow clouds. Therefore, Hirodai-Shirakami radar is expected to contribute to the understanding of the development and progression process of snow clouds around this area. With this background, we are attempting to use Machine Learning to build a model for predicting quantitative hourly winter precipitation from Hirodai-Shirakami radar observations. The data for five winters from 2014 to 2019 were trained using Deep Learning. For the teacher data, we used hourly precipitation data (Masuda et al., 2019) with rain gauge capture rate correction applied to APHRODITE daily precipitation data APHRO_JP (Kamiguchi et al., 2010). First, we set ten elements to the input layer: reflection intensity, Doppler velocity (four elevation angles, respectively), latitude, and longitude, and the results did not lead to a model that could improve precipitation prediction (the error between actual and predicted values did not decrease). Next, we added hourly estimates of ERA5 reanalysis dataset (temperature, specific humidity, wind component, and vertical velocity) at the same time to the input layer, which improved the model (the error decreased). In the future, we will switch the reanalysis data used for the input layer to the forecast values from the JMA numerical forecast model, and verify the feasibility of the winter precipitation forecast model by combining the numerical forecast model output with Hirodai-Shirakami radar observations, and will also evaluate Recursive Neural Network.

In Japan, heavy rainfall disasters happen almost every year. Predicting rainfall events accurately is essential for people to evacuate in advance of the disaster happening. Many researchers have tried to forecast the rainfall but found it hard because of its chaotic characteristics. Recently, the application of “Deep learning” outperforms the conventional approaches in various fields as the size of data is larger and computational power increases. This study aims to forecast rainfall events ranging from weak to heavy by using deep learning technology. We used the spatial distribution of the Radar AMeDAS Precipitation (RAP) data by the Japan Meteorological Agency (JMA) covering all over Japan. We divided the data into 13 regions of 256 km square. We trained the deep learning model called U-Net, pixel-wised prediction architecture. The input data are the continuous variables of rainfall intensity for 6 hours up to the current time (30-minute intervals), and the teacher data are the categorized rainfall intensity for the next 6 hours (1-hour intervals). We divided the 13 years’ precipitation data as follows: The data of 2006 - 2012 were prepared for training, 2013 - 2015 was used as validation for checking the learning progress. The data between 2016 and 2018 were for a test in which the prediction accuracy was evaluated. For many rainfall events, the model successfully predicted their occurrences and movements in 1 hour prior. However, the model tended to overestimate the rainfall intensity, and its trends became noticeable with a longer prediction time. For preventing this problem, the method of data augmentation has to be explored. 

The well-known machine learning technique of artificial neural networks (ANN) is used for predicting the surface air temperature anomalies over Japan in the winter season. Lag correlation analysis is used to derive the input attributes for the ANN model. Comparison of the ANN predictions with the North American Multi-Model Ensemble (NMME) predicted surface air temperature anomalies shows the ANN to outperform the NMME predictions with high ACC skill score and low root mean square error values.

Over 50 years since Lorenz’s 1963 study and a follow-up presentation in 1972, the statement ``weather is chaotic’’ has been well accepted. Such a view turns our attention from regularity associated with Laplace’s view of determinism to irregularity associated with chaos. In contrast to single type chaotic solutions, recent studies using a generalized Lorenz model (GLM, Shen 2019a, b; Shen et al. 2019) have focused on the coexistence of chaotic and regular solutions that appear within the same model using the same modeling configurations but different initial conditions. The results, with attractor coexistence, suggest that the entirety of weather possesses a dual nature of chaos and order with distinct predictability. In recent studies (Shen et al 2021a, BAMS), based on the GLM, we illustrate the following two mechanisms that may enable or modulate two kinds of attractor coexistence and, thus, contribute to distinct predictability: (1) the aggregated negative feedback of small-scale convective processes that can produce stable non-trivial equilibrium points and, thus, enable the appearance of stable steady-state solutions and their coexistence with chaotic or nonlinear oscillatory solutions, referred to as the 1^st and 2^nd kinds of attractor coexistence; and (2) the modulation of large-scale time varying forcing (heating) that can determine (or modulate) the alternative appearance of two kinds of attractor coexistence. Recently, the physical relevance of findings within Lorenz models for real world problems has been reiterated by providing mathematical universality between the Lorenz and Pedlosky models, as well as amongst the non-dissipative Lorenz model, the Nonlinear Schrodinger, and the Korteweg–de Vries equations (Shen 2020, 2021). Based on our results obtained using  global models and Lorenz models, we then discuss new opportunities and challenges in predictability research with the aim of improving predictions at extended-range time scales, as well as sub-seasonal to seasonal time scales.

We investigated the vertical profile of CO and six hydrocarbons (C₂H₆, C₂H₂, HCHO, CH₃OH, HCOOH, and HCN) related to the Australian bushfire. For this analyses, we used satellite data: hydrocarbon mixing ratio retrieved from the ACE-FTS measurement and fire information (fire count, burned area, and fire radiative power) retrieved from the MODIS measurements. Considering the typical dominance of westerly in this region, we examined the CO and hydrocarbon pattern over the Pacific, located in the eastern part of Australia; We confirmed that the air mass from the Australian bushfire region can be transported to the east. During the period showing large fire count and burned area, we can see the obvious enhancement of CO and hydrocarbons, except HCHO, implying that the HCHO amount in the air may have multiple source of emission and production. The altitude showing the enhancement almost reaches to the tropopause, meaning that we also carefully look into the variation of atmospheric composition in the upper troposphere and even lower stratosphere when the large scale bushfires occur. While this study was performed based on the sparse measurement of hydrocarbon species, we could find several interesting features about the variation of multiple hydrocarbon quantity related to the characteristics of wildfire, which can be significantly considered for the feedback analyses between the biomass burning and climate change. Thus, the design of geostationary satellite for the hydrocarbon measurement looks necessary in the future.

The Geostationary Environment Monitoring Spectrometer (GEMS), launched on 19 February 2020, aims to monitor the daily air quality information for trace gases and aerosols in the Asia-Pacific region. For the mission goal, GEMS was designed to measure the direct sun light and backscattered sunlight from the Earth in ultraviolet and visible ranges with significantly higher spectral resolution. In order to retrieve the geophysical products such as ozone, nitrogen dioxide, formaldehyde with required accuracy, the accurate understanding of instrument spectral response functions (SRFs) are important. The SRFs of two-dimensional detectors, like the GEMS instrument, depend on the spectral and spatial dimension, and hence, the main challenge is to determine each SRF over the 1033 pixel spectral range for each 2048 field of view. On the other hand, the characteristics of prelaunch SRFs investigated using the analytical functions were found to be very similar in both wavelength and field of view directions, thanks to highly stable and symmetrical hardware specification of GEMS. In addition, the initial on-orbit SRFs retrieved using the observed solar irradiances show that their characteristics are consistent with the prelaunch ones, indicating that no significant change occurred during the harsh launch process. However, the on-orbit characteristics of SRFs may vary over time due to the thermal changes, sensor degradations and switching modes of instrument. Thus, it is necessary to monitor the temporal variations of on-orbit SRFs during the mission. Here, this study retrieves daily on-orbit SRFs using GEMS solar irradiances measured from November 2020 to July 2021 and assuming analytical function form. The detailed results of the on-orbit characteristics and behavior of SRFs are to be presented.  

The Geostationary Trace gas and Aerosol Sensor Optimization (GeoTASO) instrument is an airborne hyperspectral spectrometer measuring backscattered solar radiation in the ultraviolet (290–400 nm) and visible (415–695 nm) wavelength regions. This study presents high-resolution sulfur dioxide (SO₂) maps over the Korean Peninsula, produced by SO₂ retrievals from GeoTASO measurements during the Korea–United States Air Quality Field Study (KORUS-AQ) from May to June 2016. The highly sensitive GeoTASO instrument with a spatial resolution of ~250 m × 250 m can detect point emission sources of SO₂ within its fields of view, even without merging multiple overlapping observations. To retrieve SO₂ vertical columns from the GeoTASO measurements, we apply an algorithm based on principal component analysis (PCA), which is effective in suppressing noise and biases in SO₂ retrievals. The retrievals successfully capture SO₂ plumes and various point sources such as power plants, a petrochemical complex, and a steel mill, located in South Chungcheong Province, some of which are not detected by a ground-based in situ measurement network. Spatial distributions of SO₂ from GeoTASO observations in source areas are consistent with those from the Stack Tele-Monitoring System reports and airborne in situ SO₂ measurements. Comparisons of SO₂ retrievals from GeoTASO and low Earth orbit satellite sensors demonstrate the significance of high-resolution SO₂ observations by indicating that GeoTASO detects small SO₂ emission sources that are not precisely resolved by single overpasses of satellites. To evaluate the dilution effect of SO₂ columns occurring at the Geostationary Environment Monitoring Spectrometer (GEMS) spatial resolution, we upscale the GeoTASO SO₂ retrievals to the GEMS ground pixel size. Since the upscaled GeoTASO retrievals also detect SO₂ plumes clearly, we expect GEMS to identify even small SO₂ emission sources over Asia.

Dallas-Fort Worth (DFW) area is currently non-compliant with Environmental Protection Agency’s (EPA) National Ambient Air Quality Standards (NAAQS) for ozone. A long-term trend analysis of the measured concentrations of ozone, oxides of nitrogen (NO_x), and total non-methane organic compounds (TNMOC) from three ambient air quality monitoring stations (Dallas Hinton (DAL), Fort Worth Northwest (FWNW), and Denton Airport South (DEN)) was conducted. Positive matrix factorization (PMF) was employed on VOC concentrations in order to understand the long-term changes in apportioned sources. Vehicle exhaust and other combustion-related emissions have steadily decreased since 2000 as shown by the decrease in oxides of nitrogen (NO_x) concentration by -3.87%/year, -2.69%/year, and -1.21%/year, respectively. At the same time, DEN showed a +9.97%/year increase in TNMOC concentrations, while DAL and FWNW decreased by -1.62%/year and -0.63%/year, respectively. The TNMOC concentrations at all three sites were largely composed of n-alkane species due to the proximity to the Barnett shale gas region, which saw increasing activities through 2014. The isopentane/n-pentane ratios suggest a higher influence from vehicular emissions in DAL and FWNW, and DEN was influenced by emissions from natural gas activities. Highly reactive TNMOC species (1-butene, m/p-xylene, propylene and toluene) had a stronger influence on the ozone formation potential (OFP) calculated for DAL and FWNW, whereas slow reacting n-alkane species (ethane, isopentane and propane) dominated in DEN. A large disconnect between the county-wide emissions inventory trends when compared to the trend of TNMOC concentrations was also noted. Based on our findings, we recommend continued reduction of NO_x and reactive TNMOC to improve ozone conditions at DAL and FWNW, and additional reduction of TNMOC species closely related to natural gas emissions in DEN.  Policy planners must employ region-specific approaches to emissions control strategies to meet ozone compliance.

Coronavirus Disease 2019 (COVID19) pandemic poses extreme threat to public health and economy, particularly to the nations with higher population density. The disease first reported in Wuhan, China; later, it spreads elsewhere, and currently, India emerged as COVID19 hotspot. In India, we selected 20 densely populated cities having infection counts higher than 500 (by 15 May) as COVID19 epicenters. Daily COVID19 count has strong covariability with local temperature, which accounts approximately 65– 85% of the explained variance; i.e., its spread depends strongly on local temperature rise prior to community transmission phase. The COVID19 cases are clustered at temperature and humidity ranging within 27–32°C and 25–45%, respectively. We introduce a combined temperature and humidity profile, which favors rapid COVID19 growth at the initial phase. The results are highly significant for predicting future COVID19 outbreaks and modeling cities based on environmental conditions.
On the other hand, CO2 emission is alarmingly high in South Asia (India) and entails high risk of climate change and extreme hot summer. Zoonotic viruses are sensitive to warming induced climate change; COVID19 epicenters are collocated on CO2 emission hotspots. The COVID19 count distribution peaks at 31.0°C, which is 1.0°C higher than current (2020) and historical (1961–1990) mean, value. Approximately, 72% of the COVID19 cases are clustered at severe to record breaking hot extremes of historical temperature distribution spectrum. Therefore, extreme climate change has important role in the spread of COVID19 pandemic. Hence, a strenuous mitigation measure to abate greenhouse gas (GHG) emission is essential to avoid such pandemics in future.