Aerosol Classification Method Based On AERONET And CALIPSO And Its Global Temporal-spatial Distribution

The types and characteristics of atmospheric aerosols have obvious regional and seasonal characteristics,which are mainly caused by the differences of emission sources and diffusion transport conditions in spatial-temporal.The dominant aerosols in different regions are different,and the dominant aerosols in the same region are also different in different seasons.Different dominant aerosols will cause different climate responses through their differences in radiative forcing,which are fed back to human activities and the natural environment.Therefore,strengthening the identification of aerosol types in different seasons and regions,would deepen understanding of the climate impact factors in this region,is conducive to improving the ability of atmospheric numerical models to simulate different regions and different seasons,and improving the accuracy of simulation.This paper used AE(?ngstr?m Exponent)data of global aerosol automatic observation network AEROENT(AErosol RObotic NETwork)Level 2 Version 3 and VDR(Volume Depolarization Ratio)data of polar orbit satellite CALIPSO(The Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation)Level 2 Layer to conduct global spatial-temporal matching from March 2018 to February 2019.Based on a large number of previous studies on the optical and microphysical properties of atmospheric aerosols,the threshold method was used to determine the types of atmospheric aerosols,which were divided into five types: natural dust type,pollution dust type,marine continental type,biomass burning type and other types of aerosols.This paper discuss the global aerosol type distribution,including the aerosol types in different seasons and the global distribution of different types of aerosols,and analyze the possible errors when matching satellite and ground-based aerosol data.The main results and conclusions are as follows:(1)Between March 2018 and February 2019,CALIPSO and AERONET had a total of 1014 matching points.After being classified by the threshold method,the largest and most widely distributed are marine continental aerosols,followed by polluted dust,accounting for 28.1% of the total number,the third is other types ofaerosols,accounting for 20.3% of the total number,the fourth is biomass burning type,accounting for 4.2% of the total number,and finally the natural dust type,accounting for 2.7% of the total number.(2)In spring,the proportion of polluted dust aerosols is the most,but the other three seasons are mainly marine continental aerosols.From spring to winter,the aerosol type of pollution dust gradually decreased.The natural dust aerosol also decreases with the change of seasons,more in spring,less in summer and autumn,and very few occurrences in winter.Biomass burning type is the least in summer,the most in spring,and the second in autumn and winter.The marine continental aerosol peaks in summer.The distribution range of the marine continental aerosols is basically unchanged with the seasons and can be observed in most stations.The seasonal variation has an important influence on the natural dust aerosols,but has little influence on the pollution dust aerosols and biomass burning aerosols,which mainly depends on the change of emission sources and diffusion transport conditions brought by the seasonal variation.(3)Marine continental aerosols are mainly distributed in the United States and European countries,even exist in island and Arctic ice stations,but rarely in desert stations.In addition to the sites in the United States and Europe where human activities are more frequent,polluted dust aerosols also occur frequently at the edge of the desert or in downwind areas of the desert.Biomass burning aerosols are mainly distributed in Mediterranean coastal countries in Europe,Central America,northern India,islands in South Asia,coastal areas in East Asia and South Africa.The emissions of biomass burning aerosols mainly come from local forest burning and agricultural burning,and have the contribution from cross-regional transportation.(4)The slope,intercept and correlation coefficient of the straight line are 0.473,0.097 and 0.477,respectively,which are obtained by AOD of 1014 matching points of CALIPSO and AERONET,and pass the significance test of 95%.After removing the matching points with larger absolute relative error(greater than 1.1),the fitting effect is optimized.The slope is increased to 0.714,the intercept is reduced to 0.042,and the correlation coefficient r is also increased from 0.477 to 0.717.In addition,the fitting effect at night is better than that at day.The reason for the large difference of AOD during the day may be that the background condition of strong sunlight makes the signal-to-noise ratio low.(5)Within the effective time and distance,the main reasons for the large error ofthe AOD observed by CALIPSO inversion and AERONET include: the discontinuity of aerosol distribution caused by terrain fluctuation;the cloud layer around 10 km above the surface,the error of cloud identification;the presence of mixed aerosol in the atmosphere,which will affect the inversion results;the optical thickness given by CALIPSO is inaccurate when biomass burning aerosols dominate;if the aerosol content is too small,the inversion is more difficult.