Objective synoptic pattern classification study on ozone pollution and diffusion transport characteristics in Dongguan City
Based on ground observation data and reanalysis data from 2019 to 2023,this study uses the Self-Organizing Map(SOM)neural network algorithm to perform synoptic pattern classification for ozone-polluted days in Dongguan,investigate the characteristics of ozone variations and their meteorological causes under different polluted weather patterns from a seasonal perspective.Results indicate that ozone pollution was primarily concentrated in the autumn over the five years,with 2019 and 2022 being the most severe.The SOM results demonstrate that influenced by seasonal characteristics,ozone pollution was more significant in three synoptic patterns:the Western Pacific subtropical high/typhoon periphery type in the summer(31.8%),the low-pressure trough type in spring,summer,and autumn(11.49%),and the transformed cold high ridge type in spring(8.43%).Conversely,pollution was lighter in three other synoptic patterns:the continental subtropical high/typhoon periphery during the summer-to-autumn transition(8.43%),the front of cold front type in autumn(7.28%),and the weak cold high ridge type across autumn,winter,and spring(32.57%).The airflow from the subtropical high/typhoon periphery substantially contributed to ozone pollution,with pollution days impacted in 2019 and 2022 accounting for 46%and 41%of the annual total,respectively.Vertical descending airflows from the east were present in all synoptic patterns,showing weaker tendencies in the warmer seasons and stronger in the colder seasons.The air masses affecting Dongguan's ozone pollution mainly originated from eastern Guangdong.Influenced by easterly airflows,compounded by emissions and diminished diffusion conditions on the western downwind side,likely significantly contributed to the characteristic west-high and east-low ozone distribution in Dongguan.
Dongguan Cityozoneobjective synoptic pattern classificationpollution characteristicsdiffusion conditionspotential source regions
NETL
NSTL
万方数据
