Back to Program


Search By:

Presentation Mode : All
Conference Day : 05/08/2021
Time Slot : AM1 08:30 - 10:30
Sections : AS - Atmospheric Sciences










Atmospheric Sciences | Thu-05 Aug




AS37-A001
Comparison of Chemical Fingerprint and Source Apportionment of Pm2.5 at Two Seaports in the Philippines and Taiwan

Yu-Lun TSENG1+, Chung-Shin YUAN1#, Gerry BAGTASA2, Jian-Xing WU1
1National Sun Yat-sen University, Taiwan, 2Institute of Environmental Science & Meteorology, University of the Philippines, Philippines


This study aims to clarify the spatiotemporal variation, chemical fingerprints, and source apportionment of atmospheric fine particles (PM2.5) in the Kaohsiung and Manila Seaports, and further identify the potential sources of PM2.5 and their contributions. Two representative sampling sites located at Kaohsiung and Manila Seaports were respectively selected for simultaneous sampling of PM2.5 from May 2018 to January 2020. The sampling of 24-h PM2.5 was conducted for continuous 7 days in each season. Water-soluble ions, metallic elements, carbonaceous content, anhydrosugars, and organic acids in PM2.5 were then analyzed to characterize their chemical fingerprint. Furthermore, chemical mass balance (CMB) receptor modeling and backward trajectory simulation were applied to resolve the source apportionment of PM2.5 at both sites. The results indicated that both seaports were highly influenced by polluted air masses from the north, resulting in elevated PM2.5 concentrations in winter and spring. Secondary inorganic aerosols (SO42–, NO3-, and NH4+) accounted for 34.1–76.0% of water-soluble ions (WSIs). Crustal elements dominated the metallic content of PM2.5, but trace elements were mainly originated from anthropogenic sources. More organic carbon (OC) than elemental carbon (EC) was found in PM2.5, with secondary OC (SOC) contributing 23.9–38.9% to the former. The concentrations of levoglucosan were in the range of 21.78-154.50 ng/m3. The results showed that the air quality of Manila Seaport was worse than that of Kaohsiung Seaport. The major sources of PM2.5 were soil dust, mobile sources, sea salts, and biomass burning. The contribution of anthropogenic sources (i.e., industrial boilers, waste incinerators, and secondary aerosols) to PM2.5 rose up in winter and spring, suggesting that the northerly transport of PM2.5 highly influenced the particulate air quality for both seaports.

AS37-A007 | Invited
Impact of Chemistry and Transport Processes Above Atmospheric Boundary Layer on Surface PM2.5 in South Korea

Hyo-Jung LEE+, Hyun-Young JO, Yu-Jin JO, Juseon BAK, Cheol-Hee KIM#
Pusan National University, Korea, South


The long-range transport of air pollutants frequently occurring in East Asia affects the air quality at both surface and upper level over neighboring regions along the elevated mixed layer above atmospheric boundary layer. In addition, a shallow stable boundary layer forms near surface at nighttime, separating the residual layer above it from the surface, which brings out vertical redistribution of air pollutants. Therefore, the understanding the chemical transformation processes within these elevated mixed layer and residual layer is critical for predicting the air quality over downwind regions. In this study, we explore the influence of chemical conversion occurring above planetary boundary layer on surface PM2.5 concentrations through analyzing the vertical structures of particulate- and gas-phase air pollutants measured by aerosol lidar measurements and simulated by three-dimensional atmospheric chemical transport model (WRF-Chem). Our results show how the behavior of primary and secondary air pollutants such as sulfur dioxide, nitrogen oxides, sulfate, nitrate, and organic aerosols differs within the residual layer and elevated mixed layer. In addition, the intrusion process of secondary air pollutants generated in the upper levels at nighttime approaching ground level on the following day is also identified. <This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A1A03044834).>

AS37-A005
Optimizing the Empirical Relationship Between Aerosol Optical Thinness and Fine Particle Matter Concentration Over East Asia

Juseon BAK#+, Hyo-Jung LEE, Cheol-Hee KIM, Wonbae JEON, Hyun-Young JO
Pusan National University, Korea, South


Aerosol particles should be regularly monitored to protect the public health as well as to understand the effect of aerosols on climate change and cloud microphysics. Space-based remote sensing is suited to detect atmospheric air pollutants at regular spatial/time intervals with wide coverage. Satellite measurements of aerosols are provided as an aerosol optical thickness (AOD), a measure of the amount of light that aerosols scatter and absorb through the entire atmosphere. The satellite-derived AOD is empirically converted to concentrations of fine particle matter concentration (PM2.5) using the regression relationship between AOD and PM2.5 based on chemical transfer models. In this work, the sensitivity analysis will be presented to see how the relationship between AOD and PM2.5 differ with chemical compositions and optical properties using WRF-Chem, and CMAQ, GEOS-Chem, respectively. This chemical model-based relationship is evaluated against that driven from in-situ PM2.5 measurements and AEORNET based remote sensing measurements. The final objective of this work is to establish the optimized AOD- PM2.5 relationship over the East Asia to retrieve PM2.5 from polar-orbiting MOIDS measurements and then extend to geostationary GEMS measurements. This work is funded by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A1A03044834). 

AS37-A002
Investigation on Chemical Characteristics of Marine Fine Particles Long-range Transported Via Easterly Offshore Channel of Taiwan Island Towards South China Sea

Po-Hsuan YEN+, Chung-Shin YUAN#, Yu-Lun TSENG, Jun-Hao CENG, Ker-Yea SOONG
National Sun Yat-sen University, Taiwan


The purpose of this paper is to investigate the spatiotemporal variation, chemical fingerprints, and transport routes of marine fine particles at three remote sites in the West Pacific Ocean (Green Island), Bashi Channel (Kenting Cape), and South China Sea (Dongsha Island). Day and night 12-h PM2.5 samples were collected for continuous seven days in each season during the intensive sampling periods, and further analyzed their chemical composition including water-soluble ions, metallic elements, carbons, anhydrosugars, and organic acids. Moreover, the potential sources of PM2.5 and their contributions were further identified by backward trajectory simulation and chemical mass balance (CMB) receptor model, which explored the transport phenomena and interaction between the continents and the marginal seas.The concentrations of PM2.5 at the three sites in winter and spring were higher than those in summer and fall. Chemical analysis results showed that abundant secondary inorganic aerosols (SO42-, NO3-, and NH4+) accounted for 58.2-69.6% of water-soluble ions. Crustal elements (Mg, Ca, Fe, and K) are the dominant metallic content of PM2.5. The mass ratios of OC and EC (OC/EC) were generally higher than 2.0 for PM2.5 in winter and spring. High concentrations of levoglucosan were observed in winter and spring with the annual average concentrations of 15.4 ng/m3 and the range of 3.8~28.7 ng/m3. During the Asian northeastern monsoon periods, a superimposition phenomenon of local emissions and long-range transport significantly increased the concentration of PM2.5. Receptor modeling results showed that the major sources of PM2.5 were fugitive dust, sea salts, and secondary inorganic aerosols at three remote sites. 

AS37-A004
Impacts of N2o5 Heterogeneous Chemistry Over Korean Peninsula by Regional Air Quality Modeling

Hyun-Young JO+, Hyo-Jung LEE, Juseon BAK, Yu-Jin JO, Wonbae JEON, Jaehyeong PARK, Cheol-Hee KIM#
Pusan National University, Korea, South


This study investigated the potential of fine nitrate (NO3- in PM2.5) formation from Nocturnal dinitrogen pentoxide (N2O5) heterogeneous chemistry under relatively high humidity winter days in the Seoul Metropolitan Area. The heterogeneous hydrolysis of dinitrogen pentoxide (N2O5) plays an important role in both nocturnal particulate nitrate formation and photochemistry on the following day through the photolysis of nitryl chloride (ClNO2), yet these processes in highly polluted urban areas remain poorly understood. As a modeling study, the WRF-CMAQ which has improved the parameterizations for N2O5 heterogeneous uptake and ClNO2 production and gas-phase chlorine chemistry was employed, and several numerical tests are carried out with base case simulation and two sensitivity tests while turning off N2O5 heterogeneous uptake or ClNO2 chemistry or both. The results showed that the model simulations including N2O5 heterogeneous uptake and ClNO2 chemistry was able to capture the temporal patterns and the magnitudes of NO3- and PM2.5. Compared with the base simulation, two sensitivity tests showed lower NO2 concentrations and higher O3 concentrations. It demonstrates the impact of N2O5 heterogeneous uptake and ClNO2 chemistry on NO3- and O3 formation. Our study suggests the heterogeneous N2O5 uptake or ClNO2 chemistry are needed to improve our understanding of NO3- and PM2.5 formation and better forecast PM2.5 pollution levels over SMA.  *This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A1A03044834).

AS37-A003
Analysis of Meteorological Patterns of High PM2.5 Episodes by Employing PCA/MLR Statistics

Daniel CHOI1,2+, Cheol-Hee KIM1#, Hyo-Jung LEE1, Hyun-Young JO1, Yu-Jin JO1, Shin-Young PARK3, Geum-Hee YANG4
1Pusan National University, Korea, South, 2National Air Emission Inventory and Research Center, Korea, South, 3Korea Institute of Atmospheric Prediction Systems, Korea, South, 4National Institute of Environmental Research, Korea, South


Statistical analysis is one of the most intensive techniques to describe the meteorological and atmosphere-chemical features. This study was performed to figure out meteorological features related with the high-PM2.5 pollution phenomena in two major cities (Seoul and Busan) of Korea by applying statistical approaches: Principal Component Analysis (PCA) and Multiple Linear Regression (MLR). Firstly, PCA was conducted to extract the meteorological modes through combining weather variables from surface and 850hpa. The periods of PCA were focused on high-PM2.5 cases during 2018-2019 to find the corresponding weather pattern of recent PM2.5 pollution. The meteorological variables adopted at two levels were five surface-observed variables: temperature (T2m), sea level pressure, wind speed, west-east wind component of surface wind, south-north wind component of surface wind, and three variables observed at 850hPa level: west-east wind component at 850hPa (u850), south-north wind component at 850hPa (v850), and vertical potential temperature gradient between 500hPa and 850hPa. Next, we conducted MLR analysis and confirmed the relationships between PM2.5 daily mean concentration and PCs. The results of currently employed two-step statistical approaches revealed that prevailing dominant factors influencing the high PM2.5 days in Seoul are mainly composed of upper wind characteristics in winter including positive u850 and negative v850, implying that continental anti-cyclone had a strong influence. In Busan, however, dominant factors in high PM2.5 concentration were correlated with the several major components such as high T2m and positive v850 in warm season high PM2.5 case, indicating North Pacific anti-cyclone had a substantial effect on Southeast region. Thus, our results are implying that the high-PM2.5 cases in Korea have the region-based seasonality and characteristics that are explainable by two layers (surface and 850hPa) with several dynamic and thermodynamic variables that can be considered for the region-specific PM2.5 reduction policies. [Acknowledging support by NRF funded by the Ministry_of_Education(No.2020R1A6A1A03044834)]

AS37-A006
Vertical Profile of the Organic Aerosol in Southeast Korea During the KORUS-AQ Campaign and Interpretation of Its Influencing Factors

Jong-Min KIM+, Hyo-Jung LEE, Hyun-Young JO, Yu-Jin JO, Cheol-Hee KIM#
Pusan National University, Korea, South


In this study, we comparatively analyzed Particulate Matter (PM) characteristics based on emissions, weather characteristics, and ground and aircraft measurements of KORUS-AQ campaign for two regions of Korean Peninsula: Metropolitan and Southeastern areas. Although the concentrations of nitrate and sulfate varies quantitatively, the ratio of the average concentration of nitrate to sulfate was relatively higher in Metropolitan area than in Southeastern area. The levels of OC and OA concentrations differed depending on the periods and altitude of the measurement in two areas. We also selected two cases: stagnant and external inflow dominant cases during the KORUS-AQ campaign periods and analyzed the differences between two areas. The results shows that OC concentration has some correlation with meteorological factors; a positive and negative correlations with temperature and wind speed, respectively. In the external inflow dominant case, pollutants in the Korean Peninsula were affected by the transport from the west area of the Korean Peninsula. Meanwhile, the intercomparison of non-surface or upper layer air quality during the KORUS-AQ campaign periods indicated that nitrate and sulfate concentrations were found to be relatively lower than that of the ground, while OA concentration was high in the upper layer. This is presumably due to the generation of SOA induced by vigorous photochemical reactions with high temperature and low horizontal wind velocity which are in turn resulting in both low horizontal ventilation effect and the higher accumulation of OC's high concentrations on the ground. With the results of active vertical mixing process due to the developing mixing height, higher surface OC concentration accordingly influence OA concentration at the upper layer measurements. This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1A6A1A03044834).



AS32-A026 | Invited
Making the Invisible Visible: the WMO Integrated Global Greenhouse Gas Information System

Philip DECOLA1#+, Oksana TARASOVA2, Jocelyn TURNBULL3, Mario PIERO ESPI2
1University of Maryland, United States, 2World Meteorological Organization, Switzerland, 3GNS Science, New Zealand


Accurate and precise atmospheric measurements of greenhouse gas (GHG) concentrations have revealed the rapid and unceasing rise of global GHG concentrations due to human activity. The resulting increases in global temperatures, sea-level, glacial retreat, and other negative impacts are clear. In response to this evidence, nations, states, cities, private-companies and individuals have been accelerating GHG reduction efforts while meeting the needs of global development. The urgency, complexity and economic implications of GHG reductions demand strategic investment in science-based information for targeting and tracking emission reduction policies and actions. In response, the World Meteorological Organization (WMO) Global Atmosphere Watch Program (GAW) and its partners have initiated the development of an Integrated Global Greenhouse Gas Information System (IG3IS). IG3IS combines atmospheric GHG concentration measurements and human-activity data in an analysis framework to better identify opportunities to reduce emissions of greenhouse gases and pollutants that reduce air quality, and to track the effectiveness of these policies and investment decisions. In order to successfully create and deliver information that results in valuable, verifiable emission reductions, IG3IS methods are co-developed with the user communities it hopes to serve. The IG3IS Science Implementation Plan documents its guiding principles, objectives and methods and was adopted by the WMO Executive Council in June 2018. We will present the high-level policy and action-oriented project achievements of IG3IS, as well as its future directions across its four primary objectives: 1) informing the quality control and assurance of national emission inventories, 2) informing GHG emission reduction actions in large urban source regions, 3) detecting and quantifying large hot-spots of methane emissions in industrial, agricultural and waste-management sectors, and 4) applications at the global scale intended to support the UNFCCC Paris Agreement global stock take.

AS32-A020 | Invited
Global to Local Impacts on Atmospheric CO2 from Reduction in Fossil Fuel Emissions Caused by COVID-19 Lockdown

Ning ZENG1#+, Pengfei HAN2, Di LIU2, Zhiqiang LIU2, Tomohiro ODA3, Cory MARTIN4, Zhu LIU5, Bo YAO6, Wanqi SUN7, Pucai WANG8, Qixiang CAI2, Russell DICKERSON1, Shamil MAKSYUTOV9
1University of Maryland, United States, 2Institute of Atmospheric Physics, Chinese Academy of Sciences, China, 3Universities Space Research Association, United States, 4RedLine Performance Solutions LLC and National Weather Service of National Oceanic and Atmospheric Administration, United States, 5Tsinghua University, China, 6Meteorological Observation Center of China Meteorological Administration, China, 7Meteorological Observation Centre of China Meteorological Administration, China, 8Chinese Academy of Sciences, China, 9National Institute for Environmental Studies, Japan


The world-wide lockdown in response to the COVID-19 pandemic in year 2020 led to economic slowdown and large reduction of fossil fuel CO2 emissions, but it is unclear how much it would reduce the trend of atmospheric CO2 concentration, the main driver of climate change, and whether this impact can be observed. We used a state-of-the-art atmospheric transport model to simulate CO2, driven by a new daily fossil fuel emissions dataset and hourly biospheric fluxes from a carbon cycle model forced with observed climate variability. Our results show 0.21 ppm decrease in atmospheric column CO2 anomaly in the Northern Hemisphere latitude band 0-45°N (NH45) in March 2020, and an average of 0.14 ppm for the period of February-April 2020, the largest in the last 10 years. A similar decrease was observed by the carbon satellite GOSAT, albeit with lower statistical significance. Using model sensitivity experiments, we further found that COVID and weather variability are the major contributors of this CO2 drawdown, and the biosphere gave a small positive anomaly. The ability of our current carbon monitoring systems in detecting the small and short-lasting COVID signal on the background of fossil fuel CO2 accumulated over the last two centuries is encouraging. Its impact suggests that to keep atmospheric CO2 at a climate-safe level will require sustained effort of similar magnitude and improved accuracy and expanded spatiotemporal coverage of our monitoring systems.

AS32-A029 | Invited
Toward Multi-scale Greenhouse Gas Monitoring System for Supporting Global Stock Take

Akihiko ITO1#+, Yosuke NIWA1, Tomohiro HAJIMA2, Nobuko SAIGUSA1
1National Institute for Environmental Studies, Japan, 2Japan Agency for Marine-Earth Science and Technology, Japan


Achievement of the carbon-neutral society is one of the overarching tasks for the sustainability of humanity. Under the Paris Agreement of the United Nations Framework Convention of Climate Change, it becomes an important task for research community to establish a comprehensive, high-precision, and transparent system of greenhouse gas (GHG) budget estimation. In April 2021, a new task-force project will be launched in Japan to develop a GHG monitoring system, provisionally called the Comprehensive Observation and Modeling for Multi-scale Estimation of Greenhouse Gas budgetS (COMM-EGGS), funded by the Ministry of the Environment, Japan. This is a joint project of National Institute for Environmental Studies, Japan Agency for Marine-Earth Science and Technology, Chiba University, and Meteorological Research Institute. The project is composed of three research components: 1) observation and top-down estimation of GHG budget, 2) evaluation of GHG mitigation with an Earth system model, and 3) bottom-up estimation of GHG budget. These activities cover different spatial scales spanning from major city to national and global emissions, by using ground observatory, aircraft, and satellite observations and fine-mesh atmospheric and surface emission models. In the project, we put emphasis on the Asia-Pacific region, in which a comprehensive GHG monitoring system is deficient to date. Through mutual comparison and validation, we will make attempt to improve confidence of the estimation system for GHG budget verification. Finally, the system aims at contributing to the Global Stocktake of the Paris Agreement by providing scientific evidence for GHG emission reduction.

AS32-A008
Regional Inverse Modeling System for Urban Carbon Flux in Megacity Co2-Seoul Project

Eunsil OH#+, Sujong JEONG
Seoul National University, Korea, South


Exact assessment of urban carbon cycle is a critical step to achieve carbon neutral for mitigating climate change impact. As Lagrangian transport models for high spatial and temporal resolution applications have been developed, regional inversion studies are actively conducted to verify carbon fluxes at a fine scale. In Seoul, Korea, as for the “Megacity CO2-Seoul” project, the regional inverse modeling system employing WRF-STILT which re-estimates urban fossil fuel carbon dioxide (FFCO2) flux was established. Under the Bayesian inverse model framework, the inversion system modifies the FFCO2 inventory of the Seoul Capital Area (SCA), which is produced by the bottom-up approach, by paring the ground CO2 measurement constraints from 5 different sites in the SCA. In addition, urban biospheric flux generated by the CArbon Simulator from Space (CASS) is utilized. In this presentation, we introduce the regional inverse modeling system optimized in SCA. By having the newly developed urban inversion system, we found our prior FFCO2 inventory was overestimated in the comparison with results of the inverse model. More detailed methods and results will be presented at the meeting. This study was supported by the National Research Foundation of Korea (NRF) grant funded by the Government of Korea (NRF-2019R1A2C3002868) and the Creative-Pioneering Researchers Program through Seoul National University (SNU).

AS32-A015
Components for Future Monitoring of Fossil Fuel CO2 Emissions

Yilong WANG1#+, Philippe CIAIS2, Gregoire BROQUET3, François-Marie BRÉON3
1Institute of Geographic Sciences and Natural Resources Research, China, 2Institut Pierre Simon Laplace, France, 3Laboratoire des sciences du climat et de l'environnement, France


The monitoring of fossil fuel CO2 emissions is the biggest challenge for the deployment of a policy-relevant integrated carbon observation system. Atmospheric inversions can use CO2 measurements to provide an independent way of quantifying the CO2 fluxes. An operational observing system requires in situ and remotely sensed data at much higher resolution and density than currently observing system, as well as the inclusion of fossil fuel CO2 proxy measurements such as radiocarbon in CO2. In this study, we use two inversion systems to separately assess the potential of space and in situ components to monitor anthropogenic CO2 emissions. The space component follows the specifications from the European Spatial Agency (ESA) for the Copernicus Anthropogenic Carbon Dioxide Monitoring (CO2M) concept, and the in situ component assumes a network of about 40 sites measuring radiocarbon in CO2 over each main continent. The results show that the space component can only constrain the emissions from large (in terms of emission rates) cities and point sources, while the in situ networks could help to constrain emissions at regional to national scales, thus raising a need for sustained efforts to improve the calibration of each component of the system and to develop integrated inversion systems.  

AS32-A021
Tracing Atmospheric Fossil Fuel Co2 Based on Δ14co2 Observation in Chinese Cities and Background Sites

Zhenchuan NIU#+
Institute of Earth Environment, Chinese Academy of Sciences, China


 As a significant fossil fuel CO2 (CO2ff) emitter, how much of the increased atmospheric CO2 derived from fossil fuel emissions is not only an environmental issue, but also an important scientific question in China. Radiocarbon (14C) is a unique tracer to quantify atmospheric CO2ff. Atmospheric Δ14CO2 were measured in Chinese cities and background sites, with the objectives to trace temporal atmospheric CO2ff variations and to determine the factors influencing them. Significantly (p < 0.05) high CO2ff values were observed in winter in inland city. There are no significant differences in CO2ff values between weekdays and weekends. Diurnal CO2ff variations were plainly evident. Δ14CO2 values at the regional background sites that close to economic region were found to be significantly (p < 0.05) lower than those at Waliguan (WLG), indicating different levels of CO2ff inputs at those sites. Regional sources contributed in part to the CO2ff inputs at Shangdianzi (SDZ) and Lin’an (LAN), while local sources dominated the trend observed at Luhuitou (LHT). The variations of CO2ff were controlled by a combination of emission sources, topography, air mass trajectories and atmospheric boundary layer. These data provide a preliminary understanding of atmospheric Δ14CO2 and CO2ff inputs in Chinese cities and background sites.

AS32-A003
Chinese Hydrofluorocarbons Emissions Inferred from Atmospheric Measurements

Bo YAO1#+, Xuekun FANG2, Martin VOLLMER3, Stefan REIMANN3, Liqu CHEN1, Ronald G. PRINN4
1Meteorological Observation Center of China Meteorological Administration, China, 2Zhejiang University, China, 3Laboratory for Air Pollution and Environmental Technology, Empa, Swiss Federal Laboratories for Materials Science and Technology, Switzerland, 4Center for Global Change Science, Massachusetts Institute of Technology, United States


Hydrofluorocarbons (HFCs) have been widely used in China to replace ozone-depleting substances (ODSs) that are required to be phased out under the Montreal Protocol regime. There are limited studies, which report HFC emissions in China, especially for recent years and using top-down approaches based on atmospheric measurements. Here we used flask and in situ measurements for nine HFCs from seven sites across China over the period 2011−2017, and FLEXPART-model-based Bayesian inverse modeling, to estimate HFC emission magnitudes and changes in China. We found that emissions of HFC-32 (CH2F2), HFC-125 (CHF2CF3), HFC-134a (CH2FCF3), HFC-227ea (CF3CHFCF3), and HFC-245fa (CHF2CH2CF3) have been increasing fast over this period, while emissions of HFC-143a (CH3CF3), HFC-152a (CH3CHF2), HFC-236fa (CF3CH2CF3), and HFC-365mfc (CH3CF2CH2CF3) were relatively stable. Total CO2-equivalent emissions of the nine HFCs increased from ~60 Tg yr-1 in 2011 to ~100 Tg yr-1 in 2017. Among these nine HFCs, HFC-134a (39%) and HFC-125 (35%) are the biggest contributors to the national total HFC CO2-equivalent emissions. Cumulative contributions from China’s HFC emissions to the global total HFC mole fractions and their related radiative forcing increased from 1.0% in 2005 to 10.7% in 2017. When comparing global emissions with the sum of emissions from China and the developed countries, an increasing difference is observed over recent years, which points to substantial additional HFC emissions from other developing countries under the Kyoto Protocol.



AS36-A007
Plans And Status For IMERG Version 07

David BOLVIN1#+, George J. HUFFMAN2, Eric NELKIN1, Jackson TAN3,2
1Science Systems and Applications, Inc., United States, 2NASA Goddard Space Flight Center, United States, 3University of Maryland, Baltimore County, United States


As part of the upgrade of Global Precipitation Measurement (GPM) mission data products to Version 07, the U.S. Science team’s merged satellite product, the Integrated Multi-satellitE Retrievals for GPM (IMERG), is nominally scheduled for full release as V07 in early 2022. The changes being implemented are driven by community feedback, optimization of existing components, and the introduction of new features.   Preliminary evaluation and user feedback show that intense precipitation (as defined locally) is frequently mis-estimated in the current V06 IMERG, usually with positive bias.  Therefore, extensive work has been done to improve the intercalibration process for individual satellite products.  User feedback raised another issue, namely that the averaging in the morphing (Kalman filter) process distorts the regional PDF of precipitation rates.  This problem is being addressed by implementing the Scheme for Histogram Adjustment with Ranked Precipitation Estimates in the Neighborhood (SHARPEN), a regional adjustment to the PDF of morphed precipitation estimates.  A third issue is that the IR retrieval scheme being used in V06 IMERG requires upgrading.  As well, V06 IMERG does not use passive microwave precipitation estimates over snowy/icy surfaces due to difficulties in retrieval.  Upgrades to the Goddard Profiling (GPROF) algorithm are being evaluated for skill over snowy/icy surfaces, with the expectation that such retrievals will be useful for some satellites, at least for surface temperatures that are not extremely cold.  Finally, we have refined the motion vector computation to use multiple model/reanalysis moisture variables to improve performance around orography. We will report the plan for V07 processing as of the conference time, and provide some examples of the changes in algorithm performance between V06 and V07.

AS36-A005
Evaluation of IMERG Oceanic Precipitation

Jackson TAN1,2#+, Jianxin WANG3, David MARKS3, David BOLVIN3, David WOLFF4, George J. HUFFMAN2, Eric NELKIN3
1University of Maryland, Baltimore County, United States, 2NASA Goddard Space Flight Center, United States, 3Science Systems and Applications, Inc., United States, 4NASA Wallops Flight Facility, United States


This study evaluates the performance of the Integrated Multi-satellitE Retrievals for GPM (IMERG) V06B satellite precipitation product from the Global Precipitation Measurement (GPM) mission against two distinct sets of surface observations in the Pacific Ocean. Specifically, we compare the IMERG precipitation in the open-ocean environment at the half-hourly 0.1° scale against the ground radar on the Kwajalein Atoll and at the monthly point-scale against gauges from 38 low-lying atolls. Such evaluations over the ocean are particularly important as in-situ precipitation observations are sparse, hindering the identification of possible errors and the improvement of the algorithm.  The evaluation against both sets of ground reference indicates an overall overestimation, likely stemming from a positive bias in the estimation of higher intensities. The evaluation against the atoll gauges revealed a seasonally varying performance—possibly due to the presence of tropical cyclones—and a consistent underestimation of light precipitation and overestimation of intense precipitation. The evaluation against the radar demonstrated an improvement over the previous version, but also uncovered a double-peak in the distribution of precipitation from cross-track sounders. These results provide a framework for understanding the performance of IMERG over the oceans and improving the next version of IMERG.

AS36-A001
The Development of Malaysian Convective Rainfall Algorithm (MCRA)

Noor Azam SHAARI+, Ambun DINDANG#
Malaysian Meteorological Department, Malaysia


Heavy rainfall from the convective clouds is a weather phenomenon that frequently occurs in Malaysia. The scientific approach to measure the rainfall amount from the convective cells using the satellite platform is urgently needed to understand the phenomenon and to improve the now-casting techniques. This new rainfall algorithm development considers the local factor that contributes to the phenomenon such as the topography and land-sea atmospheric interaction within an observation domain of 90oE to 130 oE and 10 oS to 10 oN. Precipitation rates are derived from a simple regression method between the brightness temperature of IR (Tbs) from Himawari-8 with the instantaneous convective rainfall type of the Dual-frequency precipitation radar (DPR) from the Global Precipitation Measurement (GPM) core observatory spacecraft. The regression of the mean precipitation radar at each mid-point interval 0.1K of Tbs is established and the equation representing the regression is used for making estimation. Several rainfall cases are used to validate the algorithm’s accuracy with other satellites’ derived rainfall products of GPM-IMERG and GSMaP. The concern of this study is to determine the best rainfall threshold amount that gives the ideal estimates statistically. The result shows a stronger R-squared value of almost 80% in the relationship between the Tbs and the derived estimates when the threshold value of 0.8mm is applied in the algorithm. The correlation coefficient values for the relationship are found to be negatively stronger, which is nearly 0.85 at the same threshold value. The correlation coefficient value of the derived rainfall between the model and GPM-IMERG is also found stronger; nearly 0.9. The rainfall plots can easily be discriminated when compared with the actual image by Infrared Enhanced_2 (10.4) of satellite Himawari-8.

AS36-A003
OSSE for a Hyperspectral IR Sounder on the Himawari Follow-on Geostationary Satellite

Kozo OKAMOTO1#+, Hiromi OWADA1, Tadashi FUJITA1, Masahiro HAYASHI1, Masahiro KAZUMORI1, Michiko OTSUKA1, Yoshifumi OTA1, Naotaka UEKIYO1, Hiromu SEKO1, Haruma ISHIDA1,2, Hiroshi ISHIMOTO1, Kotaro BESSHO1, Hironobu YOKOTA1
1Japan Meteorological Agency, Japan, 2 Meteorological Research Institute, Japan


 JMA started discussing the next Geo satellite program in 2018, keeping in mind the Vision for WIGOS in 2040. We consider a hyperspectral IR sounder (HSS), an enhanced VIS/IR imager and/or lightning imager as candidate instruments. Thus, we are conducting Observing System Simulation Experiment (OSSE) for a hyperspectral IR sounder in the global and regional data assimilation system to assess impacts of GeoHSS on tropical cyclone in the global system and heavy precipitation in regional system as well as representative meteorological parameters. The GeoHSS was simulated from the ECMWF re-analysis 5 (ERA5) assuming the spectral characteristics of MTG/IRS and scanning function of full disk every hour at 30km resolution as the initial trial. We assimilated clear-sky radiances in the global OSSE and temperature and relative humidity profiles in regional OSSE. Assimilation procedures such as observation screening and observation errors assignment were employed from HSS radiance assimilation in the operational global system and sounder retrieval assimilation in the previous operational regional system. The cycle assimilation experiments showed improvement in forecasts of Typhoon track and heavy precipitation that devastated wide regions in west part of Japan. Assimilation experiments with different observation settings (scanning frequency and observation instrumental error) also presented overall positive impacts of GeoHSS. We also developed a one-dimensional variational (1D-Var) scheme to derive temperature and humidity profiles from GeoHSS radiances. The profiles will be used for nowcasting and more realistic impact assessment in the regional data assimilation for GeoHSS.

AS36-A002
Impact of Aeolus Dwl Data Assimilation on Forecasting Skills for a Very Strong Typhoon Named Haishen

Izumi OKABE#+, Kozo OKAMOTO
Japan Meteorological Agency, Japan


The horizontal line of sight (HLOS) wind data from Aeolus Doppler Wind Lidar (DWL) is available from the European Space Agency (ESA) Earth Online Portal. Assimilation experiments of this HLOS data were conducted using JMA’s global data assimilation (DA) system. The experiment term is from July to November 2020 including 20 typhoon events. The result of the assimilation experiments showed positive impacts of Aeolus HLOS wind data on the analysis accuracy and forecasting scores of not only wind speed but also geopotential height, temperature and specific humidity. Improvement of typhoon track forecasting was found suggesting the improved environment forecasting. This presentation focuses on a very strong typhoon named Haishen. After the system developed into a typhoon near Ogasawara Islands on 1st September, it headed west-northwest and developed into a large and very strong typhoon. The typhoon approached to Okinawa and Kyushu region from 5th to 7th keeping its intensity and brought human and physical damage. The track forecasting skill for Haishen was improved by assimilation of Aeolus data. The details of Aeolus data observed near Haishen and assimilation experiments will be shown in the presentation.

AS36-A006
Inter-comparisons of Geostationary Infrared Observations Using Simulated Radiance from Two Numerical Weather Prediction Models

Su Jeong LEE#+
Ewha Womans University, Korea, South


Observations from geostationary (GEO) satellites have been playing an important role in improving global weather monitoring and prediction systems. Particularly, the increased number of channels and finer spatial/temporal resolutions of the new GEO imagers, such as the Advanced Himawari Imager (AHI) on board the Himawari-8/9, Advanced Baseline Imager (ABI) on the Geostationary Operational Environment Satellite-16/17, and the Spinning Enhanced Visible and InfraRed Imager (SEVIRI) on the Meteosat series, can contribute to the prediction of small scale and fast-evolving weather phenomena. For such applications, however, an accurate calibration of imagers on board such GEO satellites is required. For the calibration of the Advanced Meteorological Imager (AMI), the main payload of the 2nd geostationary satellite of Korea, GEO-KOMPSAT-2A, this study inter-compares observations from AMI infrared channels with measurements from AHI, ABI, and SEVIRI over clear-sky ocean. To that end, the simulated radiances, converted to brightness temperature, with two different Numerical Weather Prediction (NWP) model fields, ECMWF reanalysis (ERA5) and the analysis from the global prediction model in Korea Meteorological Administration, which is based on the U.K. Met Office Unified Model, are utilized and the results are analyzed as a function of space, time, and satellite angles. The study reveals that the infrared channels of the four GEO imagers have very similar bias characteristics with a few exceptions such as a striping issue in AHI, ABI, and AMI. The study also highlights the benefits of using NWP model simulations for the inter-calibration by revealing not only the instrument-specific features but also the uncertainties in the NWP models and the radiative transfer model (Radiative Transfer for TOVs in this study). Detailed results will be presented during the conference.

AS36-A009
Incremental Learning with Neural Network Algorithm for Monitoring Pre-convective Environments Using Geostationary Imager

Yeonjin LEE1+, 2
1Ewha Womans University, Korea, South, 2,


Early warning of severe weather caused by convection is challenging. Monitoring instability trends along with water vapor variation from meteorological satellites has been utilized to help such activities. The current study proposes a neural network (NN) algorithm for a quick and efficient retrieval algorithm of convective available potential energy (CAPE) and total precipitable water (TPW) from Korea's new geostationary satellite imagery measurements (GEO-KOMPSAT-2A/Advanced Meteorological Imager (AMI)). This method is to fully utilize the high spatio-temporal resolution of the observation data as well as to be independent of the NWP models. However, conventional NN learnings using a static dataset have limitations such as exhaustive learning, impractical, and not matching in time series data. Here, we introduce incremental learning using a dynamic dataset which begins with the existing weight information transferred from a previously learned model when new samples emerge. This method uses a sliding window approach which moves along the data with a small contiguous portion to prevent sudden changes in the distribution of training data. For the preparation of learning datasets of the algorithm, nine infrared brightness temperatures of AMI, six dual-channel differences, time/geographical information, and a satellite zenith angle are used as input variables and the CAPE and TPW calculated from ECMWF's latest global atmospheric reanalysis (ERA5) are used as the corresponding target values. The final collocated dataset is prepared over the clear sky in the Northeast Asia area for one year (August 2019 to July 2020) and three months (August to October 2020) as the learning and test dataset, respectively. Incremental learning results show fast convergence to a steady-state, more stable error statistics over time, and better performance (bias decreased from 0.76 to 0.17 mm for TPW and from 27.32 to 8.56 J/kg for CAPE) compared with static learning.



AS07-A004
Recent Migration of Tropical Cyclones Toward Coasts

Ralf TOUMI#+, Shuai WANG
Imperial College London, United Kingdom


Poleward migrations of tropical cyclones have been observed globally, but their impact on coastal areas remains unclear. We investigated the change in global tropical cyclone activity in coastal regions over the period 1982–2018. We found that the distance of tropical cyclone maximum intensity to land has decreased by about 30 kilometers per decade, and that the annual frequency of global tropical cyclones increases with proximity to land by about two additional cyclones per decade. Trend analysis reveals a robust migration of tropical cyclone activity toward coasts, concurrent with poleward migration of cyclone locations as well as a statistically significant westward shift. This zonal shift of tropical cyclone tracks may be mainly driven by global zonal changes in environmental steering flow.

AS07-A002
Three Dimensional Aspects of the Fujiwhara Effect

Jaedeok LEE1, Kosuke ITO2#+, Johnny CHAN3,4
1Kongju National University, Korea, South, 2Kyoto University, Japan, 3Asia-Pacific Typhoon Collaborative Research Center, China, 4City University of Hong Kong, Hong Kong SAR


A set of f-plane idealized numerical simulations at 15° N was conducted to understand the fundamental three-dimensional characteristics of the binary interaction of tropical cyclones (TCs). When the vortices are zonally separated by more than or equal to (less than) 10° of longitude, they tend to repel (rotate or merge). Binary interacting TCs have two maximum diabatic heating regions: one in the boundary layer and the other in the mid-troposphere. The maximum diabatic heating in the boundary layer is related to the supergradient wind with large radial inflows. As the vortex intensifies, the upper-anticyclonic circulation is also strengthened. These two upper-anticyclonic circulations form an extensive single anticyclonic circulation. In the absence of a background flow, this single anticyclonic circulation results in a vertical wind shear that contributes to a substantial asymmetry in diabatic heating, subsequently modifying TC motion. This shear-induced diabatic heating results in repelling storm trajectories. In contrast, experiments without microphysical heating show distinctly approaching trajectories. These two different experiments demonstrate the importance of diabatic heating in storm movement. The time series of the potential vorticity (PV) budget shows a well-balanced feature between the local change term and the sum of other PV diagnostic terms. Above the boundary layer, the PV budget is balanced between the diabatic and advective terms. The local change in PV shows that the subsequent TC motion coincides with the positive PV area. In addition, the positive PV explained by the PV advection term accounts for the same scenario. Meanwhile, When two vortices interact in closer proximity, vortex Rossby waves form inward of the vortex. Since the horizontal shear could generate these vortex Rossby waves, this result may account for an increase in the horizontal shear as a result of the direct binary interaction of TCs.

AS07-A030
Improvement of Tropical Cyclone Track Simulation Over the Western North Pacific Using Machine Learning Method

Kyoungmin KIM#+, Dong-Hyun CHA, Jungho IM
Ulsan National Institute of Science and Technology, Korea, South


A tropical cyclone (TC) causes a significant amount of social and economic damage with strong wind and heavy rainfall. To mitigate and prepare for damages, it is important to improve the accuracy of the TC track simulation. In previous researches, the track of TC has been predicted with the numerical models and the statistical models. Recently, machine learning methods also have been used for the TC track forecast with historical data and satellite images without forecast data from numerical models. In this study, we simulated TC track by a numerical model and corrected the bias of simulation with artificial neural network (ANN). All categories of TCs except for tropical depression (i.e., tropical storm, severe tropical storm, and typhoon) from June to November were hindcasted by the Weather Research and Forecasting (WRF) model, and TCs that occurred from 2006 to 2018 over the western North Pacific were included. The simulated positions of TCs as well as TC information and atmospheric characteristics were used as input variables. Since the error of simulation increases with the forecast lead time, the positions of TCs at 72-hour forecast lead time were set as target variables for ANN. The number of neurons in ANN was optimized by TCs in 2006-2015 and TCs in 2016-2018 were corrected by the optimized ANN. Also, the result of bias correction is analyzed by K-means clustering. The ANN reduced the error of WRF by 8.81% in four of the eight clusters. For the other four clusters, the output selection method based on the simulation error of the WRF was applied to exclude the outlier of ANN results. The error of ANN was further reduced than that of WRF by applying the output selection. Consequently, the error of ANN with output selection was 4.34 % less than that of WRF.

AS07-A015
Assimilating Doppler Radial Velocity and Dual-doppler Synthetic Winds on Landfalling Tropical Cyclone Nesat (2017) in Taiwan

Wei-Ting FANG1#+, Ming-Jen YANG2, Pao-Liang CHANG3
1Central Weather Administration, Taiwan, 2National Taiwan University, Taiwan, 3Central Weather Bureau, Taiwan


In this study, improvement of the inner-core structure of tropical cyclones (TCs) by assimilating Doppler radial velocity and dual-Doppler synthetic winds are examined to explore their impacts on the TC track and structure prediction. We first conducted ideal experiment to compare the wind analysis as assimilating the Doppler radial velocity and dual-Doppler synthetic winds. For both uniform flow and rotational flow experiments, the analysis increments in the area which more than two radars can cover simultaneously are identical. Results show that assimilating Doppler radial velocity with the proper consideration of radar beam width could more effectively correct the bias of the TC center and inner core structure than that without beam-width consideration when the background and observation were mismatched. We then applied this technique to a real case, TC Nesat (2017), to evaluate the forecast performance as assimilating two observation datasets. It was found that assimilating the extended Doppler velocity could forecast the moving speed and the inner-core structure more accurately than those of assimilating dual-Doppler synthetic winds or Doppler velocity without beam-width consideration. The result indicates that even the radial velocity only provides the radial component of actual horizontal wind, the more observations are used in radial velocity data assimilation experiment, the better model performance is revealed. Therefore, it is suggested that using Doppler velocity for radar data assimilation is preferable for landfalling TCs in Taiwan.

AS07-A008
Effects of the Assimilation of Relative Humidity Reproduced from T-PARCII and Himawari-8 Satellite Imagery Using Dynamical Initialization and Ocean-coupled Model. Part Ii: a Rapid Weakening and Neutral Phase of Typhoon Trami (2018)

Jaedeok LEE1#+, Doo-Sun R. PARK2, Kosuke ITO3, Chun-Chieh WU4
1Kongju National University, Korea, South, 2Kyungpook National University, Korea, South, 3Kyoto University, Japan, 4National Taiwan University, Taiwan


As a consecutive study, this Part II study investigated the rapid weakening (RW) of Typhoon Trami (2018) using the same methodology as in the Part I study. During the RW period and neutral phase period, Trami stayed longer in a specific area and thereby producing substantial upwelling. To examine this ocean effect, two sensitivity experiments such as the coupled atmosphere-ocean experiment with HYCOM sea surface temperature (AO-HYCOM) and the uncoupled atmosphere experiment with ERA5 reanalysis sea surface temperature (UA-ERA5) were carried out with the same initial vortex at the initial time. The UA-ERA5 experiment remained strong up to 120 h simulation time without producing significant upwelling. In contrast, the AO-HYCOM experiment showed a substantial reduction in sea surface heat fluxes and suppressed convective activity from the inner-core area with substantial upwelling during the RW period. As a result, the AO-HYCOM experiment well predicted the RW of Trami, leading to substantial improvement in intensity forecasts. However, the AO-HYCOM experiment also showed some worse track forecasts associated with the cross-track errors. Given that other atmospheric large-scale environments are almost similar between the UA-ERA5 and AO-HYCOM experiments, substantial upwelling and gradual changes of the ocean state could be a major factor in modulating tropical cyclone track and intensity.

AS07-A028
Advanced Global Model Ensemble Forecasts of Tropical Cyclone Formation, and Intensity Predictions along Medium-Range Tracks

Russell ELSBERRY1, Hsiao-Chung TSAI2#+, Wei-Chia CHIN2, Timothy MARCHOK3
1University of Colorado, Colorado Springs, United States, 2Tamkang University, Taiwan, 3Geophysical Fluid Dynamics Laboratory, United States


Marchok vortex tracker outputs from the ECMWF ensemble (ECEPS) and NCEP ensemble (GEFS) are utilized to provide the Time-to-Formation (T2F of 25 kt or 35 kt) timing and positions along the weighted-mean vector motion (WMVM) track forecasts, and the weighted analog intensity Pacific (WAIP) technique provides 7-day intensity forecasts after the T2F. Example T2F(35) forecasts up to 5 days in advance of two typhoons and one non-developer in the western North Pacific are described in detail. An example T2F forecast of pre-Hurricane Kiko in the eastern North Pacific indicates that Hawaii would be under threat by the end of the 15-day ECEPS WMVM track forecast. Validations of the T2F (25) and T2F(35) timing and position errors are provided for all ECEPS and GEFS forecasts of the two typhoons and Hurricane Kiko. If the T2F timing errors are small (< 1 day), the T2F position errors along the WMVM track forecasts will be small (< 300 km). Although the primary focus is on the western North Pacific, the example from the eastern/central North Pacific indicates the potential for future application in other basins. 

AS07-A031
Impact of High-resolution Moving Nesting Domain on TC Track and Intensity Forecast

Jihong MOON#+, Dong-Hyun CHA, Jinyoung PARK
Ulsan National Institute of Science and Technology, Korea, South


In the last few decades, the predictability of numerical weather prediction (NWP) models to tropical cyclone (TC) track has improved dramatically. However, a significant error still exists for a medium-range (3 to 5 days) track forecast. The regional model with high-resolution horizontal grid spacing is widely used in predicting TCs for the accurate representation of TC intensity and structure. In this study, the general impact of high-resolution moving nesting domain on TC intensity and track forecast is verified for a total of 115 forecast cases of 37 TCs using the WRF model. The 1DM experiment with a horizontal grid spacing of 12 km tends not to capture the intensification process, especially for the maximum intensity of above 60 m/s. The intense TCs are better predicted by the 2DM experiment with a moving nesting domain of 4 km. The bias and RMSE of maximum wind speed and minimum sea-level pressure are decreased in the 2DM, and the forecast of LMI (Lifetime maximum intensity) is also improved in the 2DM. For the track forecast, 1DM tends to predict the TC track deviated right to the best-track data, and this rightward tendency is decreased in the 2DM experiment. In particular, the leftward shift in the 2DM is frequent for the intense TCs, and these TCs are generally distributed in lower latitude. It seems that the track forecast of lower latitude TCs are sensitive to the TC core dynamics other than large-scale features, while the TCs transited to mid-latitude are highly influenced by the large-scale features like the western North Pacific high and mid-latitude trough.

AS07-A010
Effect of Background Error Tuning on Assimilating Satellite Radiance: Evidence for the Prediction of Tropical Cyclone Track and Intensity

Gaurav TIWARI#+, Pankaj KUMAR, Alok Kumar MISHRA
Indian Institute of Science Education and Research Bhopal, India


The background error covariance is an essential component in data assimilation under a numerical weather prediction system, which largely dominates the error correlations between different analysis variables. This study aimed to assess the background error length and variance scale factors on the intensity and track prediction of tropical cyclone systems for radiance data assimilation over the North Indian Ocean. To examine the sensitivity of the background error variance scale and length scale, multiple data assimilation analyses were performed using the three-dimensional variational data assimilation method by varying the decorrelation length scale (LEN_SCALING) and the variance (VAR_SCALING) of the background error. In radiance data assimilation, generally smaller variance scale and length scale factors improve the cyclone track's prediction. Therefore, setting both VAR_SCALING and LEN_SCALING provides a better track than other combinations. The influence of the VAR_SCALE factor on the cyclone track prediction is more than that of the LEN_SCALE factor. Also, there is no direct impact of tuning variance scales and length scales on the forecast intensity for the first few forecast hours. Tuning length scale factors less than 40% may provide reduced forecast errors of maximum surface wind and minimum-sea-level pressure with increasing forecast lead time.



AS30-A003
The Hazardous Rainfall Event Associated with Three Consecutive South China Sea Tropical Cyclones in 2017-18 Boreal Winter and an Evaluation of Its S2s Predictability

Mong-Ming LU#+, Wayne Yuan-huai TSAI, Chung-Hsiung SUI
National Taiwan University, Taiwan


During the three weeks from December 13, 2017 to January 2, 2018 the Philippines was successively attacked by three tropical cyclones that caused sever flood, landslides, sea transportation interruption and high death toll in Mindanao. This high-impact extreme rainfall event was found to be closely associated with convectively coupled equatorial waves (CCEWs) in particular the westward moving equatorial Rossby waves and mixed Rossby-gravity waves. The MJO was not active during the active period of tropical cyclones. Our analysis of the S2S database suggest that only the best forecast model ECMWF can generate the positive skew forecast of the South China Sea (SCS) rainfall in 10 days before the occurrence of the subseasonal peak event. However, the forecasted skewness is weaker than observation. A subseasonal peak event, which was defined as the period with maximum 15-day accumulated rainfall amount within the boreal winter half year from November to April (NDJFMA) based on the CMORPH dataset, is chosen as the target for evaluating S2S prediction. The peak event is the most important wet episode of time scale within the range of 10~60 days, during which the accumulated rainfall shows distinctly contribution to the seasonal totals. The results support the hypothesis that the MJO can extend the forecast skill of the peak event for one more pentad over the off-equatorial region.

AS30-A012
Evaluation of Adaptive Thresholds for Sub-seasonal Extreme Weather Forecast in Southeast Asia

Wee Leng TAN#+, Ryan KANG, Thea TURKINGTON
Centre for Climate Research Singapore, Singapore


With hydro-meteorology disaster forming majority of natural disasters in Southeast Asia, there is increasing focus to improve the prediction of heavy rainfall events. However, due to the varied and complex topography of Southeast Asia, both the predictive skill and the impact of heavy rainfall events is not uniform over the region. While 90thpercentile is a commonly used threshold to highlight possible occurrence of extreme rainfall in sub-seasonal forecast, a lower percentile may be more appropriate and may give a better predictive skill for some regions, as the combination of rainfall duration and topology could amplify the effect of the heavy rainfall. Here we focus on case studies of heavy rainfall events that led to disasters in July and August 2020, using the ECMWF’s extended range model. Various rainfall percentile thresholds are applied for different regions in Southeast Asia, assessing both the skill of the model (based on the hindcasts) as well as the skill of the forecast for the specific events. The results will be used for a region-dependent adaptive threshold product for disaster management agencies.

AS30-A002
Subseasonal Forecast of the 2018 Summer Heatwaves over Northwestern Europe

Mien-Tze KUEH#+, Chuan-Yao LIN
Academia Sinica, Taiwan


This study investigated the subseasonal forecasts of the mid-July to early August 2018 heatwaves over northwestern Europe using the multi-model ensemble (MME) of real-time forecasts from the subseasonal to seasonal (S2S) database. The MME forecasts captured the evolution of the warm spells over the region up to 3 weeks in advance. However, the forecasts of the heatwave occurrence and signifcance are unsatisfactory. The successful forecasts of warm spells can be attributed to the S2S model's ability to capture the persistent blocking regime and Atlantic Low regime. Positive phase of North Atlantic Oscillation and enhanced Azores High were observed during the heatwaves, and were captured in the MME forecasts. However, their contribution to the subseasonal forecast of warm spells remains unclear. 

AS30-A008
Towards S2S-scale Large-ensemble Prediction Using Global Cloud/cloud-system Resolving Model Nicam on Fugaku

Tomoki MIYAKAWA1#+, Yohei YAMADA2, Tamaki SUEMATSU1, Masuo NAKANO2, Chihiro KODAMA2, Hisashi YASHIRO3, Ryusuke MASUNAGA1, Daisuke TAKASUKA2, Takanori KODAMA1, Takao KAWASAKI1, Takemasa MIYOSHI4,5, Masaki SATOH1
1The University of Tokyo, Japan, 2Japan Agency for Marine-Earth Science and Technology, Japan, 3National Institute for Environmental Studies, Japan, 4RIKEN Center for Computational Science, Japan, 5University of Maryland, United States


The world’s new #1 supercomputer Fugaku is now available. In theme 2 of the “Large Ensemble Atmospheric and Environmental Prediction for Disaster Prevention and Mitigation”, one of the programs for promoting researches of Fugaku, we target extended-range (from over a week to several months) predictions of high-impact weather events, using the global cloud/cloud-system resolving model NICAM and its ocean-coupled version NICOCO. It requires good model performance in producing largescale, long lasting fluctuations of MJO/BSISO, ENSO, etc., as well as the high-impact weather events themselves. In this talk, we provide an introductory overview of the activities and prospects of theme 2. Typhoon is the primary high-impact weather event targeted in the theme. A typhoon that hit Chiba prefecture, Japan, in 2019 is being investigated in relation with a Boreal Summer Intra-Seasonal Oscillation event, using a large-sized ensemble NICAM simulation (〜1000 members for 14 km mesh and 30〜100 members for 3.5 〜7 km meshes, initial conditions provided by NEXRA system). Sub-seasonal reproducibility and mechanism of MJO/BSISO are also being investigated as part of an intercomparison project for global storm resolving models (DYAMOND2), and through large-ensemble sensitivity studies (600 members for 14 km mesh). The increased computational power enables us to gain a clearer view of the model climatology, which is also a key requirement in producing extended-range prediction. The roles of ocean, e.g., local air-sea interactions, western boundary currents, and ENSO, are being taken into account using NICOCO. 

AS30-A004
Cloud-resolving Simulations of the Arabian Peninsula Winter Rainfall

Raju ATTADA1+, Thang LUONG2, Hariprasad DASARI2, Hsin-I CHANG3, Christoforus Bayu RISANTO3, L. Castro CHRISTOPHER3, Ibrahim HOTEIT2#
1Indian Institute of Science Education and Research Mohali, India, 2King Abdullah University of Science and Technology, Saudi Arabia, 3University of Arizona, United States


The spatio-temporal distribution of winter rainfall over the Arabian Peninsula (AP), a water-scarce region, is crucial for the management of various socio-economic sectors such as water resources, agriculture, and industrial development. High-resolution atmospheric models are often used to investigate the rainfall distribution over the AP because of the scarcity of the data in this region. However, the uncertainties in current models are tightly influenced by the representation of clouds, moist convection, and complex topography. Reducing grid spacing down to few kilometers (Cloud-System Resolving Model; CRM) is tested here for better description of clouds and associated precipitation over the AP. This CRM approach is expected to reduce some parameterization uncertainties and to increase the role of microphysical properties. We implemented a CRM based on the Weather Research and Forecasting (WRF) model to simulate the formation of winter clouds and precipitation over the AP during the period 2006-2016. We used the ERA-Interim reanalysis to drive the CRM, which we configured with a horizontal resolution of 2 km. We validated the CRM outputs with surface observations, reanalysis datasets, and an in-house high resolution (5 km) regional reanalysis (KAUST reanalysis). We further compared its simulations against with convective parameterized simulations. Our results suggest that the CRM simulated rainfall is in better agreement with the observed rainfall. The assessment of precipitation statistics suggest that the CRM approach adequately represents the diurnal cycles of precipitation over the AP. It also indicates that CRM describes better fine-scale surface precipitation characteristics and microphysical feedbacks. We further investigate the dominant precipitation-causing mechanisms during winter over the AP using the CRM outputs and compared its performance against the KAUST reanalysis. This work provides important knowledge towards the implementation of a high-resolution numerical modeling system for the prediction of the AP winter rainfall on seasonal and sub-seasonal scales.

AS30-A001
Arabian Peninsula Convective Events Sub-seasonal Forecast Capability Using Convective-permitting Regional Climate Modeling

Hsin-I CHANG1#+, Christopher CASTRO1, Christoforus Bayu RISANTO1, Thang LUONG2, Ibrahim HOTEIT2
1University of Arizona, United States, 2King Abdullah University of Science and Technology, Saudi Arabia


Severe weather associated with convective thunderstorms is becoming more intense globally and is also observed in the Arabian Peninsula (AP). AP convective extremes are often observed during winter season (October to March). Improvements in extreme weather forecast for sub-seasonal to seasonal forecast increase the preparedness of convective extremes and related hazards. We designed a series of ensemble forecast downscaling using the Weather Research and Forecasting model (WRF) at convective-permitting spatial scale. The driving global sub-seasonal reforecast is provided by the European Centre for Medium-Range Weather Forecasts (ECMWF). Sub-seasonal WRF simulations are performed on the AP’s top 20 extreme precipitation events reported in the last 20 years, downscaling from the 11 ECMWF hindcast ensemble members. Several aspects of the simulated events are evaluated: (1) Precipitation forecast capability: determine forecast window of opportunity in the regional climate model at sub-seasonal lead time, identify the value added using convective-permitting type modeling; (2) Teleconnection pattern forecast capability: determine forecast skill for the dominant large scale pattern related to the convective extremes in the driving ECMWF reforecasts and ERA-Interim reanalysis data; (3) Dominant synoptic patterns associated with the AP’s top 20 extreme events: identify forecast capability for different synoptic-driven extreme events. Historic data analysis identified 3 general synoptic patterns that lead to precipitation extreme. The extreme events are parsed into the 3 synoptic groups. Sub-seasonal forecast evaluations utilize statistical analysis tools commonly used in operational forecast evaluation, such as Probability of Detection (POD), False Alarm Rate (FAR) and Relative Operating Characteristics (ROC); (4) Mesoscale convective system (MCS) tracking: objectively tracking the MCS clouds in satellite observation and WRF downscaled reforecasts using cloud top temperature and precipitation. Through the designed analyses, we can collectively show the ensemble forecast skills for the largest convective events in the AP and advancement in forecast capability at sub-seasonal time scale.

AS30-A005
Evaluating 20 Years of Sub-seasonal Rainfall Forecasts Over the West Coast of the Red Sea

Thang LUONG1+, Hsin-I CHANG2, Christoforus Bayu RISANTO2, L. Castro CHRISTOPHER2, Raju ATTADA3, Hariprasad DASARI1, Ibrahim HOTEIT1#
1King Abdullah University of Science and Technology, Saudi Arabia, 2University of Arizona, United States, 3Indian Institute of Science Education and Research Mohali, India


Mesoscale convective systems (MCSs) occur between November and April over the west coast of the Red Sea. Typically in conjunction with intense extratropical systems, such MCSs often produce very heavy rainfall over short periods of time, which in turn can quickly become raging rivers due to the limited infiltration capacity of dry soils. Forecasts at lead times on the subseasonal-to-seasonal (S2S) timescale may potentially assist disaster risk reduction, and water resource management for the region. Here we demonstrate the predictability and forecast skill of the Weather Research and Forecasting Model (WRF) at convective-permitting resolution (4 km) by performing retrospective forecast simulations at sub-seasonal timescale (2 to 4 weeks ahead) over the 20-year period of 1998 to 2017. We dynamically downscale the European Centre of Medium-range Weather Forecasts (ECMWF) S2S ensemble reforecasts. The simulations are initiated weekly on Mondays of the 2018 calendar year from October 8th to November 26th for forecast lead times up to 30 days using 11 ensemble members. A total of 1,760 hindcasts for Novembers of 20 years has been generated to evaluate the predictability of rainfall over the west coast of the Red Sea. Most of the major rainfall events in November are associated with strong synoptic influence and are well reproduced by the convective-permitting model. The analysis of the WRF results demonstrates that the convective-permitting model lead to statistically significant improvements in the forecast of these events compared to ECMWF forecasts. Added values from the higher resolution model help create favorable thermodynamic conditions for rainfall, namely the retreating of the two semi-permanent Highs at 850mb and stronger low-level moisture fluxes from the Red-Sea towards land.

AS10-A007
Effect of Stratospheric Ozone on the Surface Temperature in the Northern Hemisphere During Boreal Winter

Yong-Cheol JEONG1#+, Sang-Wook YEH2, Seungun LEE3, Rokjin J. PARK3, Seok-Woo SON3
1Hanyang University, Korea, South, 2Hanyang university, Korea, South, 3Seoul National University, Korea, South


It is well known that ozone, which is one of the important atmospheric species, can affect surface climate by modulating atmospheric radiative processes and further atmospheric circulation. When the Antarctic ozone hole matters, many previous studies focused on the surface impact of ozone in the Southern Hemisphere. In recent years, however, it has been suggested that ozone may affect the surface climate in the Northern Hemisphere (NH). In this study, the effect of ozone on the surface temperature in the NH during boreal winter is investigated using an idealized model experiment. We conduct two experiments, which are referred to as LINOZ-on and LINOZ-off, using the Global/Regional Integrated Model system-Chemistry Climate Model (GRIMs-CCM). In LINOZ-on, ozone is calculated with the linearized ozone scheme in the GRIMs-CCM. In LINOZ-off, however, the ozone is prescribed. It is found that two experiments simulate different mean surface temperatures in the NH during boreal winter, which is associated with the Arctic Oscillation (AO)-like difference in the surface. Similar results are found in the comparison between CMIP5 AMIP models with relatively high and low mean stratospheric ozone concentrations. We argue that stratospheric ozone can exert a significant impact on the surface temperature in the NH. This result implies that stratospheric ozone is an important factor to correctly simulate surface temperature in the NH in the climate model.

AS09-A032
Long-term Trends of Surface Air Pollutants in Metropolitan Cities of South Korea

,
,


South Korea achieved economic expansion through rapid industrialization and urbanization, but socio-economic development caused severe environmental problems and increased health risk, especially in densely populated cities. Thus, air quality in the Korean peninsula has been a serious issue for the public and researchers. To figure out the situation of Korean air pollution, we need to examine the long-term trend of air pollutants which are highly variable in time and space. Here, we investigate trends of air pollutants (PM10, SO2, NO2, O3 and CO) in seven metropolitan areas, Korea from 2002 to 2018 using the surface measured data provided from the AIRKOREA data archive. Trends of PM10, SO2, NO2 and CO vary within cities (PM10: -1.5~-0.1 mgm-3/yr, SO2: -0.2~0.0 ppb/yr, NO2: -0.6~0.1 ppb/yr, CO: -21.8~-1.2 ppb/yr). Seoul metropolitan area shows the largest decrease of four pollutants. But Daejeon where is located near Seoul doesn’t have clear decline in concentrations of pollutants. O3 pollution gets worse (0.5~0.9 ppb/yr), especially in inland cities (Seoul, Daejeon, Gwangju). The results suggest importance of local scale emission in the long-term perspective. Comprehensively, we identified significant differences of trends according to pollutants and cities. Our study has its significance as a basic data for understanding and diagnosing air pollution. We anticipate our analysis to be used as a reference in establishing a policy for air quality. Establishing a long-term air quality improvement plan which is considering local characteristics will effectively reduce the damage caused by air pollution.

AS09-A037
Improving Validation Result of OMI NO2 Total Column with Dilution Correction

KangHo BAE1#+, Chang-Keun SONG1, Sangseo PARK1, 2, 2, Dong-won LEE3, 2, 2, Jong-Uk PARK4
1Ulsan National Institute of Science and Technology, Korea, South, 2, , 3National Institute of Environmental Research, Korea, South, 4Seoul National University, Korea, South


Collecting quality-assured and nearly real-time data from ground-based observation networks are in high need of geophysical validation for GEMS, the world first geostationary satellite to monitor the diurnal variation of trace gases and aerosols over the Asian region with high spatial/temporal resolution, international and domestic efforts to expand of ground-based observation networks have been going on. In these regards, several Pandora instruments started to be newly operated in Korea. Pandora products have been used to validate for gaseous products(O3, NO2 TCD) of LEO satellite data (i.e., OMI, OMPS, TROPOMI L2 products). The previous studies showed that the slope of the regression line for NO2 TCD of the urban region was very low due to (a) spatial resolution difference between LEO satellite and PANDORA(spatial inhomogeneity) and (b) loss of high surface concentration of NO2 a priori profile for polluted area. In this study, we especially try to improve the validation result of OMI NO2 TCD with Pandora by using the CMAQ model result for considering the horizontal spatial inhomogeneity of NO2Keywords: GEMS, PANDORA, remote sensing data, OMI, validationAcknowledgement : This subject is supported by Korea Ministry of Environment(MOE) as “Public Technology Program based on Environmental Policy(2017000160001)” by the FRIEND(Fine Particle Research Initiative in East Asia Considering National Differences) Project through the National Research Foundation of Korea (NRF) funded by the Ministry of Science and ICT (Grant No.: 2020M3G1A1114615)

AS09-A002
Impacts of Soil NOx Emission on O3 Air Quality in Rural California

Tong SHA1#+, Xiaoyan MA2, Jun WANG3, Huanxin ZHANG3, Nathan JANECHEK3, Yanyu WANG3, Yi WANG3, Lorena Castro GARCIA3, G. Darrel JENERETTE4
1Shaanxi University of Science and Technology, China, 2Nanjing University of Information Science & Technology, China, 3The University of Iowa, United States, 4University of California, Riverside, United States


Nitrogen oxides (NOx) are a key precursor in O3 formation. Although stringent anthropogenic NOx emission controls have been implemented since the early 2000s in the United States, several rural regions of California still suffer from O3 pollution. Previous findings suggest that soils are a dominant source of NOx emissions in California, however a statewide assessment of the impacts of soil NOx emissions (SNOx) on air quality is still lacking. Here we quantified the contribution of SNOx to NOx budget, and effects of SNOx on surface O3 in California during summer by using WRF-Chem with an updated SNOx scheme, the Berkeley Dalhousie Iowa Soil NO Parameterization (BDISNP). The model with BDISNP show a better agreement with TROPOMI NO2 columns, leading confidence in SNOx estimates. We estimate that 40.1% of the state’s total NOx emissions in July 2018 are from soils, and SNOx could exceed anthropogenic sources over croplands which accounts for 50.7% of NOx emissions. Such considerable SNOx enhance the monthly mean NO2 columns by 34.7% (53.3%) and surface NO2 concentrations by 176.5% (114.0%), leading to an additional 23.0% (23.2%) of surface O3 concentration in California (cropland). Our results highlight the co-benefits of limiting SNOx to help improve air quality and human health in rural California.

AS09-A011
Refinement of Anthropogenic Emission Inventory Based on Viirs-npp Nighttime Lights

Tang KEQIN#+
Nanjing University of Information Science & Technology, China


An accurate and timely air pollutant emission inventory is the premise and basis for effective prevention and control of air pollution. MEIC is one of the most advanced anthropogenic emission inventories in China and is widely used by many environmentalists and researchers. However, the spatial resolution of the MEIC is 0.25°×0.25° which cannot finely reflect the characteristics of air pollutant emissions at the urban scale. To further improve the spatial resolution of the MEIC, this study uses nighttime light data at 1×1 km spatial resolution from the NOAA (National Oceanic and Atmospheric Administration) VIIRS-NPP satellite to optimize the MEIC. The nighttime lights image of 2016 from satellite was inverted by GIS to get the light data which includes longitude and latitude of each grid. The results showed that the average light density in China is 0.49, and the light density in coastal areas (0.59-0.71) was stronger than that in inland areas (0.4-0.47) overall. The emission inventory optimized by satellite lighting data can more clearly reflect the distribution characteristics of emissions in different functional areas of the city. For instance, in Shanghai, the gradient of SO2 emission in the MEIC from the city center to the suburbs varied at 0.85 mol km-3 hr-1, while the gradient of SO2 emission variation in the optimized emission inventory is significantly enhanced by 22.6%. Further, the study will combine WRF-Chem, a regional atmospheric chemistry model, to quantitatively evaluate the improvement of the urban atmospheric pollutant simulation by using the optimized emission data. Key words: satellite nighttime light data; anthropogenic emission; air quality. 

AS09-A012
Impact of Land Cover Change on Summer Temperature in East China: an Integrated Assessment with Modeling and Satellite Observations

Yanyu WANG1#+, Jun WANG2, Huanxin ZHANG2, Nathan JANECHEK2, Yi WANG2, Qianshan HE3, Tiantao CHENG1
1Fudan University, China, 2The University of Iowa, United States, 3Shanghai Meteorological Service, China


Unprecedented land use/land cover change (LULCC) occurred in both urban and rural areas across China during the past two decades. While the urban heat effect is well documented, much less studied are the rural temperature variations and the overall LULCC impact on temperature during 21st century. By an integrated analysis with satellite observations and regional climate model simulations, the effect of LULCC on temperature is detected over Eastern China during 2003-2019. The WRF model is developed with the inclusion of real-time satellite data, to capture the LULCC’s continuous progression. Two sets of WRF simulations (land-constant and land-varying case) were performed for every two years in 2003-2019 and spatial pattern of temperature variations induced by LULCC was portrayed. It is found that an average of 0.41℃ warming occurs during urbanization and 0.05℃ rural cooling induced by vegetation increase. The urban temperature rises at a rate of 0.4℃ (every two years) along with the urban expansion and vegetation degradation. Much larger temperature increment with the mean value of 1℃ occurs in the new urban than the old urban. It is calculated that NDVI can explain ~8% of ΔT’s spatial variance in the old urban, ~12% in the new urban and ~35% in the rural. In addition, the urban warming can extend to the nearby non-urban areas in a proximity of 20 km. To decrease the urban warming and public health risk to extreme heat, a much larger green space fraction than before should be mandated in urban areas to alleviate urban-rural disparity of temperature as an inherent part of future LULCC policy.

AS09-A018
Extension of GOCI Yonsei Aerosol Retrieval Algorithm to GOCI-II

1, 1, Hyunkwang LIM2
1, , 2Yonsei University, Korea, South


Since GOCI (Geostationary Ocean Color Imager) was successfully launched as the world's first geostationary ocean color sensor in 2010, the GOCI Yonsei aerosol retrieval (YAER) algorithm has been continuously updated to retrieve hourly aerosol optical properties. The GOCI YAER products show good agreement with other satellite aerosol products and AERONET. The GOCI YAER products have been widely used for the analysis of air quality over East Asia and data assimilation to improve air quality forecast. GOCI-II was launched in February 2020 onboard the GEO-KOMPSAT-2B (GK-2B) satellite as the follow-up GOCI, and is going through in orbit tests. In terms of aerosol retrieval, GOCI-II has two major advantages compared to GOCI. First, the newly included UV band centered at 380 nm is advantageous to retrieve the radiative absorptivity of aerosols over the darker surface than a visible band. Second, the spatial resolution of GOCI-II is doubled to 250 m, compared to 500 m resolution of GOCI, which allows us to mask smaller-scale clouds and to retrieve the aerosol optical properties in better accuracy. In this study, the YAER algorithm is employed to retrieve AOD using GOCI-II data. The first result of GOCI-II AOD shows the consistent spatial distribution compared to GOCI AOD. GOCI-II AOD shows the correlation coefficient of 0.854 and the ratio within Expected Error (EE) of 70.4% against AERONET, which are comparable to the validation result of GOCI. It is expected that the GOCI-II data will facilitate the continuation of long-term aerosol recodes and address air quality issues over East Asia. 

AS09-A026
Identifying Regional Economic Level’s Impacts on Correlation Between So2 and Aerosols from Max-doas Measurements in the Yangtze River Delta, China

Yuhang SONG#+
University of Science and Technology of China, China


Ground-based multi-axis differential optical absorption spectroscopy (MAX-DOAS) observations were performed from 1 February 2019 to 31 January 2020 in Huaibei (in Anhui Province), Zhoushan (in Zhejiang Province), Nanjing (in Jiangsu Province), Shanghai (municipality directly under the Central government), which are all parts of the Yangtze River delta, a representative zone with high economic level in China. Through applying the optimal estimation method to the MAX-DOAS observations, tropospheric SO2 and aerosols vertical profiles and corresponding vertical column densities (VCDs) can be acquired. Based on the data above, the regularity of diurnal cycles of SO2 can be found. Specifically, the SO2 amount appears larger in the early morning and late evening and lower at noon, which can be explained with the diurnal variation of pollutant emissions and atmospheric stability. Combined with the meteorological data, the impacts of local meteorological conditions on SO2‘s VCDs can also be found out. From many meteorological factors, wind speed and direction are found to have a large effect on the level of the SO2-related pollution at cities of the Yangtze River delta, which specifically means the stronger wind will causes a more effective dispersion and then less SO2-related pollution. By comparing correlation coefficients between SO2 and aerosols in four different cities of different economic levels, it can be found that the area with higher economic level corresponds to a stronger correlation between SO2 and aerosols, which can be explained by the fact that sulfate aerosols’ formation plays a significant role in the total aerosol content in the developed regions. The retrieved data of MAX-DOAS agrees well with the data acquired from TROPOMI, proving our instruments’ robustness and reliability.

AS09-A027
Evaluate the Contribution of Secondary Formaldehyde to Ozone Formation Control in Four Metropolises Over China

Chuan LU1,2#+
1Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei 230031, China, 2University of Science and Technology of China, Hefei 230026, China


NO2 mainly comes from transportation, industrial residential, and power plant. HCHO is not only the main indicator of VOC, but also the major precursors of ozone and secondary organic aerosol. There are two sources of HCHO. The direct emission is its primary source, and the photochemical oxidation process is its secondary sources. Four typical metropolises of China were selected to analyze the similarities and differences in air pollutants. Four MAX-DOAS instruments were performed in four different metropolis of China (Guangzhou, Lanzhou, Xianghe and Shenyang) to investigate the vertical distribution of aerosol, NO2 and HCHO from Feb. 2019 to Jan. 2020, respectively. The surface concentration of NO2 and surface aerosol extinction observed from MAX-DOAS were compared with in situ NO2 and PM2.5, respectively, and they all showed good agreements. It was found that HCHO vertical structure in spring and autumn varied in the order of Guangzhou> Xianghe> Lanzhou> Shenyang; in summer varied in the order of Guangzhou> Xianghe> Shenyang> Lanzhou; in winter varied in the order of Guangzhou> Lanzhou> Xianghe> Shenyang. The NO2 vertical structure in spring varied in the order of Guangzhou> Shenyang> Lanzhou> Xianghe; in summer varied in order of Shenyang> Guangzhou> Lanzhou > Xianghe; in autumn and winter varied in order of Lanzhou> Guangzhou> Shenyang> Xianghe. The aerosol vertical structure in spring varied in the order of Guangzhou> Lanzhou > Shenyang> Xianghe; in summer varied in the order of Guangzhou> Xianghe> Shenyang> Lanzhou; in autumn and winter varied in the order of Guangzhou> Xianghe> Lanzhou> Shenyang. Moreover, CO and Ox were selected as HCHO source tracers, and the primary and secondary sources of HCHO in four metropolises are calculated. Through the research, the O3-NOx-VOCs sensitivities in vertical space is learned to analyze the sources of pollution in four metropolises and provide a strong guarantee for pollution control.

AS09-A034
Diurnal Variability of Urban Heat Island Intensity in Metro Manila, Philippines

John MANALO1,2#+, Jun MATSUMOTO3,4, Masato NODZU1, Lyndon OLAGUERA5,6
1Tokyo Metropolitan University, Japan, 2Philippine Atmospheric Geophysical and Astronomical Services Administration, Philippines, 3Department of Geography, Tokyo Metropolitan University, Japan, 4Japan Agency for Marine-Earth Science and Technology, Japan, 5Ateneo de Manila University, Philippines, 6Manila Observatory, Philippines


This study presents the first analysis of the diurnal variability of urban heat island (UHI) intensity in Metro Manila, Philippines. The two stations located in the two most populated cities in Metro Manila were investigated using the automatic weather station (AWS) measurements by the Philippine Atmospheric Geophysical and Astronomical Services Administration of the Department of Science and Technology (PAGASA-DOST) from 2014 to 2018. The highest UHI intensity (UHImax) was observed in the city of Manila (Port Area station) with a value of 4.03°C at 19:00 local time (LT), whereas, in Quezon city (Science Garden station), a UHImax of 3.02°C was observed at 18:00 LT. The lower building density and higher wind speed might have contributed to the lower UHI intensity observed in the Science Garden station. The results further show that the local characteristics of the location of the urban stations such as building density and green spaces seem to be the dominating factors that affected UHI intensity in the study area. 
The seasonal analysis revealed that highest UHI intensities occurred during the hot dry season (March-May), where lower values of NDVI, relative humidity, wind speed, and longer sunshine duration were observed. The results presented in this study were compared with other UHI studies conducted in Southeast Asia. This study serves as an initial step to further UHI research in the Philippines and to provide information that will help in mitigating the challenges brought by urbanization.

AS35-A023
Impact of Recent Warming in East Asian Marginal Seas on the Torrential Rainfall Event Occurred in Kyushu Island, Japan in July 2017

Atsuyoshi MANDA#+
Mie University, Japan


Torrential rainfall events tend to occur more frequently in Japan as well as many parts of the world. Attributing the events to the global warming is, however, a complicated task. Although the majority of the moisture that causes the torrential rainfall comes from the tropics and amount of moisture supply from the ocean surface is not large compared to the horizontal moisture transport from the south, this study has highlighted the importance of atmospheric moistening in the lower troposphere due to recent warming in the mid-latitude oceans. The torrential rainfall event that occurred in Kyushu Island in July, 2017 was selected as a typical test case in the torrential rainfall events during the Baiu rainy season in Japan. Numerical simulations in this study shows that fractional change of the amount of the precipitation due to the oceanic and atmospheric warming since 1980s is 6.8%, corresponding to 10.6 % increase per 1 K of sea surface temperature (SST) rise. It is larger than expected from the Clausius–Clapeyron relationship (~6%/K) and consistent with the previous data analysis study. The SST rise in the east Arian marginal seas plays a fundamental role in intensifying the conditionally unstable conditions in the lower troposphere and in turn increase the amount of precipitation. On the other hand, changes in precipitable water plays a secondary role.

AS35-A008
Impacts of Urbanization On Tropical Cyclone Rainfall Over the South China Great Bay Area as Inferred from Observations and Numerical Modeling

Chenxi HU1+, Francis Chi Yung TAM1,2#, Chi Lok LOI3, Kevin CHEUNG4, Yubin LI4
1The Chinese University of Hong Kong, Hong Kong SAR, 2Shenzhen Research Institute, China, 3National Taiwan University, Taiwan, 4Nanjing University of Information Science and Technology, China


In this study, the impact of urbanization on TC rainfall over GBA is investigated based on both station observations and meso-scale model simulations. By using the Hong Kong Observatory best track TC data and hourly precipitation data from 14 stations over GBA, 41 TC cases during the 2008-2017 period are studied. It is found that though the urban heat island (UHI) effect is highly suppressed during the approach of TC, TC extreme rainfall (defined as 99th percentile of hourly rainfall) is still stronger over urban compared to rural stations, with difference reaching 10mm/hr at the rainfall peak time averaged over all cases. Such enhanced TC precipitation is not induced by station location difference. The Weather Research and Forecasting (WRF) model with a single-layer urban canopy module (SLUCM) is also utilized to simulate TC cases making landfall over South China. TC Victor, Dujuan Mugknut, and Utor were dynamical downscaling by WRF-UCM to a 3km×3km resolution over the South China Sea area. Three experiments were conducted, the first one with urban land use replaced by cropland (NO-URBAN), normal land use with zero anthropogenic heat (AH) in urban grids (AH0), and normal land use with AH=300W/m2 for diurnal maximum (AH300). Results show that urban land use can strongly increase the TC precipitation over GBA megacity in all four cases, while the impact of AH itself is weak. Such rainfall enhancement can be attributed to stronger low-level moisture flux convergence (frictional convergence in PBL), as well as increased (decreased) convective available potential energy (convective inhibition) over the urban area. For TC Victor, urbanization can slightly decrease the TC intensity before landing; this may be induced by dry and hot airflow derived from the urban area, which interacts with the slow-speed TC over South China Sea before landfall.

AS35-A024
Two Types of Physical Systems Underlying Recent Summer Droughts in North China

Lan DAI+, Jonathon WRIGHT#
Tsinghua University, China


We investigate the physical processes behind summer drought in North China by evaluating moisture and energy budget diagnostics and linking them to anomalous large-scale circulation patterns. Moisture budget analysis reveals that summer drought in North China was caused dynamically by reduced vertical moisture advection due to anomalous subsidence and reduced horizontal moisture advection due to anomalous northeasterly winds. Energy budget analysis shows that reduced latent heating was balanced dynamically by decreased dry static energy (DSE) divergence in the middle-to- upper troposphere. Linking these results to previous work, we suggest that summer drought in North China was predicated on co-occurrence of the positive phases of the Eurasian (EU) and Pacific–Japan (PJ) teleconnection patterns, potentially modulated by the circumglobal teleconnection (CGT). In the typical case, the negative phase of the CGT intensified the positive EU-related upper-level cyclone. Resulting upper-level cooling and positive surface feedback imposed a cold-core surface anticyclone that weakened with height. By contrast, when the positive phase of the CGT occurred in tandem with the positive EU and PJ patterns, the anticyclone had a warm core and intensified with height. The two cases were unified by strong subsidence but exhibited opposite meridional advection anomalies. In the cold-core cases, meridional moisture inflow was reduced but meridional DSE export was enhanced, further limiting precipitation while maintaining negative thermal anomalies. In the warm-core case, which only occurred once, enhanced meridional inflow of water vapor supplied moisture for sporadic precipitation while reduced meridional DSE export helped to maintain strong static stability.

AS35-A010
Application of the Bias Correction and Spatial Downscaling Algorithm on the Temperature Extremes in China

Lianlian XU1#+, Aihui WANG2
1Sun Yat-sen University, China, 2Institute of Atmospheric Physics, Chineses Academy of Sciences, China


The Bias Correction and Spatial Downscaling (BCSD) is a trend-preserving statistical downscaling algorithm, which has been widely used to generate accurate and high-resolution dataset. We employ the BCSD technique to statistically downscale projected daily maximum temperature (DMT) over China from 13 GCMs in CMIP5 project to supplement the NASA Earth Exchange Global Daily Downscaled Projections (NEX-GDDP) dataset under the RCP2.6 scenario. We then compare the differences of DMT and four DMT-related indices (i.e., summer days (SU), annual maximum value of DMT (TXx), intensity and frequency of heatwave) between before and after downscaling over eight sub-regions of China. The results indicate that the BCSD method reduces the cool bias of the DMT over the whole China compared with original CMIP5 simulations, especially over the Qinghai-Tibet Plateau. The SU increases after downscaling for both China as a whole and most sub-regions except for South China. The BCSD also affects the mean value of TXx, intensity and frequency of heatwave at sub-regional scales, although it shows little impact on China as a whole. Besides, the BCSD reduces the temporal variability of most indices except for the heatwave frequency. The most striking finding is that the inter-model spreads of DMT, SU, TXx and heatwave intensity are dramatically reduced after downscaling compared with raw CMIP5 simulations. In summary, the BCSD method shows significant improvements to original CMIP5climate projections under RCP2.6 scenario.

AS35-A016
Interdecadal Variations in Extreme High–temperature Events Over Southern China in the Early 2000s and the Influence of the Pacific Decadal Oscillation

Baoyan ZHU1#+, Bo SUN2, Hua LI1, Huijun WANG1
1Nanjing University of Information Science and Technology, China, 2Nanjing University of Information Science & Technology, China


This study documents a sudden interdecadal variation in the frequency of extreme high–temperature events (FEHE) over southern China during summer in the early 2000s, which is characterized by a relatively small (large) FEHE during 1991–2000 (2003–2018). The composite analysis on the extreme high–temperature events (EHEs) over southern China indicates that the occurrence of EHEs is mainly influenced by increased downward surface net shortwave radiation, which is induced by the cloud–forced radiation anomalies associated with reduced cloud; the reduced cloud is attributed to anomalous descent motion and decreased water vapor content in the troposphere. Compared to the situation during 1991–2000, anomalous descent motion and decreased atmospheric water vapor content occurred over southern China in summer during 2003–2018, providing a more favorable climatic condition for EHEs. This interdecadal variation is associated with the strengthened Pacific Walker circulation after 2003. The Pacific decadal oscillation (PDO) is suggested to be an important driver for the above interdecadal variation, which shifted from a positive phase towards a negative phase after 2003. Numerical experiments demonstrate that a negative phase of PDO may induce a strengthened Walker circulation and anomalous atmospheric descent motion as well as water vapor divergence over Southern China.

AS35-A014
Improving the Dynamical Seasonal Prediction of Western Pacific Warm Pool Sea Surface Temperatures Using a Physical–empirical Model

Ping CHEN1#+, Bo SUN2
1Nanjing University of Information Science and Technology, China, 2Nanjing University of Information Science & Technology, China


The western Pacific warm pool (WPWP) has a profound impact on the global climate. In this study, the forecast skill of ENSEMBLES model for predicting the WPWP sea surface temperature (SST) for the period 1960–2006 is evaluated, where a WPWP index (WPWPI) is defined to represent the interannual variability of WPWP SST. The result indicates that the ENSEMBLES exhibit a poor skill in predicting the WPWPI during January–April (2- to 5-month forecasts starting on November 1). To improve the ENSEMBLES-predicted WPWP SSTs during January–April, a physical–empirical (PE) model is developed based on two predictors, using the year-to-year increment method and the linear regression method. The two predictors include the ENSEMBLES-predicted sea level pressure during January and the observed northern tropical Atlantic SSTs during the preceding August. The mechanisms associated with the two predictors are illuminated. The 1-year-out cross-validation and the independent hindcast indicate that this PE model may notably improve the WPWPI prediction of ENSEMBLES, with a correlation coefficient (CC) above 0.6 between the PE-model-predicted WPWPI and the observed WPWPI during January–April. The physical mechanisms expounded in this study and the PE model utilized in this study can be considered to improve the prediction of WPWP SST of numerical models in the future.

AS13-A001
An Assessment of Causality-guided Statistical Model for Extreme Mei-yu Precipitation Prediction Over China

Kelvin NG1#+, Gregor C. LECKEBUSCH1, Kevin HODGES2
1University of Birmingham, United Kingdom, 2University of Reading, United Kingdom


In 2020, record-breaking amount of Mei-yu rainfall around the Yangtze River has been observed.  Accurate prediction of extreme Mei-yu precipitation over China for the current and future climate is essential to decision and policy making in the region.  Many studies in the past have shown that large-scale modes, e.g. western north Pacific subtropical high and the south Asia high, play a role in controlling extreme Mei-yu precipitation over China. Although the spatial resolution of typical climate models might be too coarse to simulate extreme precipitation accurately, they may be able to simulate large-scale modes reasonably well.  One might therefore be able to construct a causality-guided statistical model based on those known large-scale modes to predict extreme Mei-yu precipitation.  In this presentation, we show a preliminary assessment of the performance of causality-guided statistical model which is constructed based on known large-scale atmospheric and oceanic modes and extreme Mei-yu precipitation in the two regions of China, i.e. Yangtze River Valley and Southern China.  Performance of the model varies throughout the Mei-yu season.  The highest performance has the adjusted coefficient of determination of 0.7.  Results of a sensitivity study are also discussed.

AS13-A006
Diagnostic Analysis of Heavy Rainfall Events Over the Yangtze River Valley During the 2020 Mei-yu Season

Qizhen SUN1#+, Yi-Leng CHEN1,2
1University of Hawaii at Manoa, United States, 2National Central University, Taiwan


During the 2020 early summer rainy season (Mei-Yu), a series of unprecedented heavy Mei-Yu precipitation events occurred over the Yangtze River Valley accompanied by widespread flooding over a prolonged period (62 days in total).  The moisture source for the excessive heavy rainfall in the Mei-Yu frontal zone over the Yangtze River Valley is mainly from the northern South China Sea, transported northward by anomalous strong low-level winds below the 900-hPa level, referred to as the marine boundary layer jet (MBLJ), for almost the entire Mei-Yu season (from 1 June to 31 July). The MBLJ (> 10 m s-1) is mainly related to the sub-synoptic scale pressure gradients between a stronger than normal WPSH and deeper than normal lee side low/trough on the lee side of the Tibetan Plateau and Yun-Gui Plateau.  The rainfall production in the Mei-Yu frontal zone is mainly caused by the secondary circulation associated with the subsynoptic low-level jet (SLLJ)/front system. The transport of the excessive low-level moisture from the subtropical ocean to the southern China plain is dominated by the anomalous mean flow rather than transient disturbances, contributing to the prolonged heavy precipitation period over the Yangtze River valley that lasted almost for the entire 2020 rainy season.

AS13-A010
Extreme Hot Characteristics and Their Relationships with Urbanization in Southern China During 2008-2017: A Statistical Analysis and Convection-permitting Modeling

Huanyan GAO1+, Yali LUO2#, Yuxing YUN3, Xiaoyu GAO4, Da-Lin ZHANG5, Yang CHEN3, Xiaoling JIANG3
1Nanjing University of Information Science and Technology, China, 2Nanjing University of Information Science & Technology, China, 3Chinese Academy of Meteorological Sciences, China, 4Department of Earth System Science, Tsinghua University, China, 5University of Maryland, United States


In this study, the total days, mean duration and intensity of extreme hot events over southern China during the 2008-2017 warm seasons are analyzed with daily maximum and minimum temperatures (Tmax and Tmin), by comparing the newly proposed independent hot day (IHD), independent warm night (IWN), and compound extreme (CMPD, i.e., the continuous occurrences of hot days and hot nights) to the traditionally defined hot day and warm night. Relationships between the hot extremes and urbanization are explored with 1-km resolution population density (P) data. Results show obvious differences in the spatial distributions between IHD, IWN, and CMPD over southern China. The three hot-extreme indices of CMPD are positively correlated with the population density, qualitatively similar to the traditional hot days and warm nights, reflecting the enhancement of hot extremes by the urban heat island effects. Negative correlations between the IHD and IWN indices and the population density are found, as their indices are more apparent over rural areas. In addition, IHD and IWN show less evident annual variations, while CMPD exhibits more meaningful annual variations resembling the traditional hot days and warm nights.     The analysis results obtained are used to evaluate convection-permitting climate modeling with the WRF model. The simulation underestimates IHD at most stations, with larger biases at the stations with lower population density (i.e., ln(P)<7). IWN and CMPD are also underestimated at the lower-population-density stations, but overestimated at the megacities with ln(P)>9. The observed relationships between population density and hot extreme are reproduced for CMPD, but not for the independent hot extremes. Sensitivity experiments suggest that use of a multi-layer urban canopy model significantly improves the Tmax simulation, while improved treatment of sub-grid cloud fraction and the aerosol effects could not. However, the improved treatment examined help little suppress the hot-extreme biases.

AS13-A012
Study on Key Characteristics and Main Trends of Precipitation in Thailand in Recent 32 Years

Wenting YANG1+, Jianhua SUN1#, Shenming FU2
1Chinese Academy of Sciences, China, 2Institute of Atmospheric Physics, Chinese Academy of Sciences, China


Based on the daily precipitation observation data from Thailand's meteorological department during 1981-2012, this study investigates key characteristics and main trends of precipitation in the entire Thailand and its five individual regions. It is found that, eastern and southern Thailand were the regions with the strongest precipitation, while northern Thailand was the weakest. In the focused 32 years, ~87% of the stations in Thailand featured an increasing trend in annual precipitation, with a total of 22 stations passed the significance of 95%. Southern Thailand was the region with the fastest increasing rate, and the Bangkok and Lokkun stations, which is located along the Gulf of Thailand, were the stations showed the largest increasing rates in annual precipitation. In contrast, eastern Thailand had the least number of stations that featured an increasing precipitation trend. The proportion of stations with abnormally high values (i.e., the 95 percentile and above) of annual precipitation, annual precipitation days and mean precipitation intensity increased significantly in Northern Thailand, implying that the range of extreme precipitation was enlarging in this region. For Southern Thailand, only the precipitation above heavy rainfall enlarged significantly in area. Precipitation above rainstorm showed significant differences in different regions of Thailand: in northern, northeastern and central Thailand, the non-persistent precipitation was dominant, whereas, in eastern and southern Thailand, the contributions of persistent and non-persistent precipitation was similar. The proportion of stations with positive anomaly of persistent precipitation (above rainstorm) showed a significant increasing trend in northern and southern Thailand, which indicates that the range of heavy precipitation affected by stable system in these areas had a significant expanding trend.

AS13-A019
Spatial and Temporal Characteristics of Abrupt Heavy Rainfall Events Over Southwest China During 1981–2017

Yangruixue CHEN1#+, Yali LUO2, Xiaofang WANG3, Ling HUANG4
1Chengdu University of Information Technology, China, 2Nanjing University of Information Science & Technology, China, 3China Meteorological Administration, China, 4Guangzhou Institute of Tropical and Marine Meteorology, China


The spatial and temporal characteristics of abrupt heavy rainfall events (AHRE: defined as 3-hr rainfall amount ≥50 mm and at least one 1-hr rainfall amount ≥20 mm during the 3 hr) over southwest China (SWC) between 1981 and 2017, including their occurrence frequency, intensity, and duration, were investigated based on hourly rainfall data collected from 468 rain gauge stations. The occurrence frequency of these AHREs exhibited large spatial variability among different regions. We distinguished three regions with high occurrence rates of AHREs: the Sichuan Basin (subregion A, SR-A), South Guizhou–North Guangxi (subregion B, SR-B), and West Hunan (subregion C, SR-C). Overall, the amount of rainfall generated by the AHREs was more closely related to event duration than event intensity. Analysis of the monthly variations showed that the frequency of AHREs over SWC mostly (ca., 86.8%) appeared in the warm season (May–August) and peaked in July. But the peak month exhibited distinct regional features, with it occurring in midsummer over SR-A, late spring or early summer over SR-B, and early summer over SR-C. With regards to diurnal variations, AHREs over areas with higher (lower) topography were initiated mostly in the afternoon (around midnight). In addition, short-duration (1–3 hr) AHREs began most frequently in the late afternoon, whereas the long-duration events (>6 hr) tended to begin around midnight. Composite analysis of the anomalies of moisture, equivalent potential temperature, and wind during extreme AHRE days revealed that AHRE occurrence coincided with an abnormal low-level cyclonic circulation, which enhanced wind convergence and the transport of moist and warm air, providing favourable thermodynamic conditions for AHRE formation. These results advance our understanding of rainfall characteristics over SWC and provide observation-based metrics for the evaluation of numerical simulations.