Study On Spatiotemporal Variation Characteristics And Influencing Factors Of Tropospheric Ozone In Three Urban Agglomerations In China

Tropospheric ozone（O₃）plays an essential and complex role in atmospheric chemistry and climate change and is thus one of the key research objects in the field of atmospheric science.Utilizing the monthly Tropospheric Column Ozone（TCO）data from OMI/MLS during 2005-2020,spatiotemporal variation characteristics of tropospheric O₃in three major urban agglomerations of China including Beijing-Tianjin-Hebei,the Yangtze River Delta and the Pearl River Delta were thoroughly analyzed.Convergent Cross Mapping（CCM）was employed to diagnose the driving effect of a large number of potential factors,which were divided into five categories:stratospheric-tropospheric exchange（STE）,precursors,meteorological conditions,East Asian monsoon,and atmospheric oscillations,on regional TCO in the study areas and thereby explained their local TCO spatiotemporal variation characteristics.Using CCM to screen explanatory variables,two Multiple Linear Regression（MLR）models were constructed to quantify the relative contribution of each main factor to regional TCO and to evaluate their long-term trend.The main conclusions are as follows:（1）The spatial distribution of TCO in China showed obvious centers of high and low values.High-value centers are controlled by anthropogenic emissions and concentrated in the Bohai Rim urban agglomeration.The low-value center is located on the Qinghai-Tibet Plateau,and the range is highly consistent with the large terrain.In addition,there is a small high-value center at the junction of Sichuan and Chongqing Provinces.There is significant seasonality in the variation of TCO in China with higher TCO in spring and summer than in autumn and winter.The annual cycle characteristics of TCO between northern and southern China are significantly different.From north to south,the seasonal fluctuation range and peak concentration decrease,the peak time gradually advances from July to April,and the single peak gradually transitions to double peaks.（2）The level of TCO in BTH is dominated by precursors and STE,with a cumulative relative contribution of 68.11%.The high concentration of local precursors and the downward transport of Brewer-Dobson Circulation（BDC）in the middle and high latitudes have an important contribution to the higher TCO in BTH than in the PRD.In addition,the summer peak of TCO in BTH also benefits from the meteorological conditions conducive to photochemical reactions created by the East Asian summer monsoon,the direct transport of pollutants to the north,and the abnormal cyclonic circulation which inhibits the outward diffusion of pollutants.The TCO in the PRD is dominated by dynamic processes including BDC,El Ni?o-Southern Oscillation（ENSO）,and the East Asian monsoon,with a total relative contribution of 80.32%.The strong upward transport of BDC from April to August,the dilution,northward transport of marine air masses,and the inhibition of photochemical reactions caused by the East Asian monsoon greatly account for the decrease of TCO in summer and the formation of a bimodal structure in the Pearl River Delta region.The situation in the YRD is between the BTH and the PRD and is generally closer to the BTH.Besides,this study found that the impact of ENSO on the long-term change of TCO in the three major urban agglomerations may be comparable to or even exceed the East Asian monsoon,and its mechanism deserves further in-depth analysis.The results of TCO influencing factors based on CCM are generally consistent with those of previous studies based on other methods such as models,which preliminarily prove the reliability and superiority of the CCM.（3）Previous studies generally used the Ni?o3 or Ni?o3.4 index as the proxy to analyze the relationship between ENSO and tropospheric O₃ in China and found their positive correlation.However,this study found that TCO in regions other than the PRD all had the strongest negative response to the Ni?o1+2 index,which was determined as the best proxy for ENSO by CCM.This means that ENSO may have a more complex and deeper impact on TCO in China,which needs to be further explored.According to CCM,in addition to ENSO,regional TCO also has stable causal relationships with the North Pacific Index（NPI）,Atlantic Multidecadal Oscillation（AMO）,and North Atlantic Oscillation（North Atlantic Oscillation,NAO）.Quasi-Biennial Oscillation（QBO）,which is generally used to assess the long-term trend of TCO,has no significant effect on TCO in most parts of China.Only TCO in the PRD has a weak response to the QBO index at 100h Pa.（4）The MLR model adjusted by CCM greatly improves the evaluation accuracy of long-term trends.TCO increased significantly from 2005 to 2020 in all study regions with positive trends of 2.6±0.3,2.8±0.3,2.4±0.4,and 2.3±0.1 DU/decade in BTH,the YRD,the PRD,and the whole China.