Correlation Of Arctic Aerosols With Clouds And Their Impact To Changes In Sea Ice

The Arctic is a highly climate-sensitive area.In recent years,the Arctic temperature has increased at a rate twice the global average,and aerosols have a significant impact on Arctic warming.To explore the optical characteristics of aerosols over the Arctic,we analyze the characteristics of the spatial and temporal distributions of the Arctic Aerosol Optical Depth(AOD)based on MODIS(Moderate-resolution Imaging Spectroradiometer)10 km and 3 km AOD products from March 2000 to October 2018 and validate these products via comparisons with the ground-based AErosol RObotic NETwork(AERONET).Then,using MODIS observational data,the month-by-month and seasonal mean changes of AOD and the various cloud properties(Cloud Fraction,CF、 Cloud Optical Thickness,COT、Cloud Liquid Effective Radius;CERL、Cloud Ice Effective Radius,CERI、Cloud Liquid Water Path,LWP、Cloud Ice Water Path,IWP)of Arctic aerosols from March 2001 to October 2020 were statistically analyzed,and the interaction between aerosols and clouds in different sea ice concentration was studied.Finally,the effects of Arctic aerosol-cloud interactions on changes in sea ice are discussed.The results showed(1)The correlation coefficients between the MODIS 10 km and 3 km AODs and the AERONET base station points are 0.840 and 0.852,respectively,and that the Root Mean Square Error is less than 0.1.Both the 10 km and 3 km AODs are in good agreement with the ground-based observations.The results show that,overall,the accuracy of the MODIS 10 km AOD increases as the latitude increases in the Eastern Hemisphere of the Arctic,although no consistent pattern is observed in the 3 km AOD.Additionally,the two AOD products typically exhibit greater accuracy in May,September,and October and are slightly more accurate in spring than in summer.Moreover,remarkable AOD differences are detected between low and high latitudes in the Arctic,where AOD increases unevenly as the latitude decreases.The Arctic MODIS AOD exhibits obvious seasonal variations.Spring AOD values generally range from0.1 to 0.3 and peak in the Nordic region in April.In summer,high AOD values extend to larger areas from June to August,while the AOD values are typically smaller than0.1 in autumn.Moreover,the two MODIS AOD products exhibit different patterns over Arctic land and ocean areas.The MODIS 10 km AOD is usually larger over the Arctic Ocean,while the MODIS 3 km AOD is generally larger over the Arctic land areas.(2)The seasonal correlations of Arctic AOD and cloud properties varied significantly across different sea ice concentration.Overall,in the marginal ice zone,the cloud properties of each season tend to have a positive correlation with AOD.In the open ocean,the spatial distribution of the correlation between cloud attributes and AOD has seasonal characteristics,negative correlation in spring and autumn,and positive correlation in summer,except for cloud attribute CF.CF and AOD were positively correlated under different sea ice concentration states.Overall,compared with the open ocean,cloud attributes and AOD are more correlated in the marginal ice zone,and COT,IWP,LWP,etc.show a more significant correlation with AOD in a longer range.(3)Overall,Arctic AOD and cloud attributes have significant negative correlations between month-by-month time series under different states of sea ice concentration.Compared with other cloud attributes,IWP,LWP,CERL,COT and AOD are more relevant.The correlation of AOD with the inter-month and seasonal scales of the cloud attributes is reversed in the marginal ice zone,meaning that the AOD of the marginal ice zone is inconsistent with the variation characteristics of the cloud attributes on different time scales.The peak of cloud attributes generally occurs at a maximum value of 1 to 2 months after the peak of AOD,indicating that there is a lag in cloud attributes and AOD in the Arctic,which may be the reason for the inconsistencies in the time scale of the marginal ice zone.Under different sea ice concentration states,COT and IWP passed the confidence test with AOD(-1,-2),which proved that COT and IWP lagged behind AOD by 1 to 2 months.In addition,in open ocean,CF,CERL,and LWP all lag behind AOD by 1 to 2 months.In the ice edge zone,LWP lags behind AOD by 1month.(4)The correlation between cloud properties and abnormal changes in SIC has significant spatiotemporal distribution characteristics.In different seasons,the anomalous positive correlations of cloud attributes COT,IWP,LWP,and SIC predominated in the marginal ice zone,while CF,CERI and CERL and SIC outliers showed a significant negative correlation in the marginal ice zone.Compared with liquid CERL and LWP,ice-like CERI and IWP are inversely correlated with SIC outliers.Combining the positive correlation between the cloud properties and AOD in the marginal ice zone,aerosols mainly affect sea ice changes by influencing cloud microphysical properties,and their impact on Arctic sea ice mainly depends on the balance between the cloud properties.