Effects Of Hygroscopicity On Aerosol Optical Properties And Direct Radiative Forcing In Beijing

The scattering hygroscopic growth factor f（RH）and backscatter hygroscopic growth factor f_b（RH）are important parameters for evaluating aerosol direct radiative forcing.The influence of relative humidity on aerosol properties and the direct radiative forcing of PM₁₀and PM₁were investigated in Beijing from January 2018 to December 2019.The annual mean scattering hygroscopic growth factor f（80%）of PM₁₀and PM₁were 1.60±0.24 and1.58±0.22,respectively.The annual mean values of f_b（80%）of PM₁₀and PM₁were 1.32±0.10and 1.29±0.08,respectively.The hygroscopicity of PM₁₀and PM₁aerosols were similar.The seasonal mean f（80%）of PM₁₀from spring to winter were 1.66±0.23,1.71±0.25,1.51±0.20and 1.49±0.16,respectively,which were higher in spring and summer,and lower in autumn and winter.f（80%）shows a strong positive relationship with both the scattering Angstr（?）m exponent（SAE）and the single scattering albedo（ω₀）under dry conditions;therefore,the scattering hygroscopic growth factor could be estimated using these two parameters.The upscatter fraction（β）and single scattering albedo,which are the key aerosol optical properties for the calculation of direct radiative forcing,are also RH-dependent.As RH increases,the upscatter fraction（backscatter fraction）decreases,andω₀increases.The ratio of aerosol radiative forcing at 80%RH to that under dry conditions was 1.48.The sensitivity experiment showed that the variation in the scattering coefficient with relative humidity had the greatest influence on radiation force,followed byβandω₀.The seasonal variation ofΔF（80%）/ΔF（dry）coincides with that of the aerosol hygroscopic growth factor.Our study suggests that aerosol hygroscopic growth has a significant impact on on aerosol properties and direct radiative forcing.Aerosol hygroscopicity is highly dependent on aerosol chemical components.We conducted synchronous measurements of the aerosol scattering hygroscopicity growth factor,chemical component and number concentration size distribution（PNSD）of PM₁aerosol in urban Beijing from August 20 to October 4,2019.During the measurements,the mass concentration of PM₁aerosol measured by AMS varied from 5.0 to 97.1μg m^-3,with the mean value of 30.3±17.0μg m^-3.The organic mass fraction is the largest,followed by sulfate and ammonium,and the nitrate mass fraction is relatively small.The mean values of f（80%）and aerosol hygroscopic parameterκ_f（RH）were 1.46±0.19 and 0.20±0.10,respectively.The variation of f（80%）andκ_f（RH）both decreased with the increase of organic mass fraction,and sulfate play an important role in aerosol hygroscopic growth during the observation period.The hygroscopicity parameterκ_orgof organic aerosol can be calculated fromκ_f（RH）and chemical composition.The mean value ofκ_orgwas 0.12±0.08,ranging from 0 to 0.45,which was related to the aging degree of organic aerosol.The ratio of PM_2.5to CO represents the aging degree of aerosol.The higher the PM_2.5/CO value is,the larger theκ_orgis.The enhanced hygroscopicity of pollution aerosols,on the one hand,is due to the increasing proportion of inorganic aerosols,on the other hand,it is also related to the enhanced hygroscopicity of organic aerosols.The aerosol liquid water content（ALWC）of PM₁aerosol was calculated in this study by three methods based on（a）ISORROPIA II model,（b）chemical composition,and（c）hygroscopicity growth factor f（RH）,and the results of ALWC calculated by these three methods are consistent.The contribution of different chemical components to ALWC was studied by volume mixing scheme andκ-K（?）hler theory.The results showed that the contribution of inorganic and organic aerosols to ALWC was 72%and 28%,respectively,with ammonium sulfate contributing the most.When the ambient RH is greater than 80%,the ALWC is even higher than the aerosol mass concentration under dry conditions.Under environmental conditions,the scattering hygroscopic growth factor f（RH）of aerosol at ambient relative humidity has an important effect on atmospheric visibility and aerosol radiative forcing.In this study,we developed random forest models by using meteorological data,air quality data,or optical parameters under dry conditions,which can distinguish aerosol hygroscopicity and can be easily obtained,as input variables for estimating f（RH）at ambient relative humidity in this study.In addition,the performance of the model is improved by optimizing the input variables.It was found that the model performed better when the absorption-related variables were removed,NO₂was removed from the air quality data,and the absorption coefficient,single scattering albedo,and submicron absorption ratio were removed from the optical parameters.When using meteorological data to develop a random forest,f（RH）is slightly underestimated,with a slope of 0.94.f（RH）estimated by optical parameters is overestimated,especially at high relative humidity.The random forest model based on air quality data had the best performance（slope=1.00,R²=0.94）.The contribution of O₃and PM_2.5is relatively larger,which reflects the level of atmospheric oxidation and pollution respectively.The random forest model can use easily available data,such as air quality data,to invert the aerosol hygroscopic growth factor f（RH）under ambient relative humidity,so as to obtain more information about aerosol hygroscopicity.It can also be applied to models to better simulate atmospheric visibility and assess the direct radiation of aerosols under environmental conditions.