Contents, Sources And Radiative Forcing Of Light-absorbing Particles In Seasonal Snow In Northern China

Light-absorbing particles?LAPs?,such as black carbon?BC?,organic carbon?OC?,and mineral dust?MD?,deposited on snow can effectively reduce snow albedo and enhance the absorption of solar radiation.They further contribute to alterations in snow morphology,accelerations in snowmelt,and reductions in snow cover,then affect local hydrological cycles and regional and global climate.In this study,based on the field campaigns across northwestern China in January and February,2012 and northeastern China in January,2014,respectively,we measured LAPs contents and analyzed their spatial-temporal distributions.Then we applied a Positive Matrix Factorization?PMF?model,the Hybrid Single Particle Lagrangian Integrated Trajectory?HYSPLIT?model and chemical analysis to estimate the potential sources of pollutants in snow.In addition,we analyzed the effect of LAPs in snow on snow albedo based on field observations as well as Snow,Ice,and Aerosol Radiation?SNICAR?model and Spectral Albedo Model for Dirty Snow?SAMDS?.Finally,we used MODIS observations to retrieve the radiative forcing by LAPs in snow(RF^LAPs_MODIS)across northeastern China in December-March from 2003 to 2017.In northeastern China,snow depth is spatially varied.In Inner Mongolia,snow cover was thin and patchy,and average snow depth was mostly less than 10 cm.Snow depth was maximum?46cm?inside a forest near the Changbai Shan.Snow density and radius?R_m?varied considerably from0.13 to 0.38 g cm^-3 and from 0.07 to 1.3 mm,respectively.Along the northeastern Chinese border,snow was cleanest with best BC concentrations(C_BC^est)from 30 to 260 ng g^-1.The snow in industrial regions across northeastern China was most heavily polluted and C_BC^est in surface snow ranged from510 to 3700 ng g^-1.In northwestern China,the cleanest snow was found in northeastern Xinjiang along the border of China,and it presented C_BC^est of approximately 5 ng g^-1.The dirtiest snow presented C_BC^est of approximately 450 ng g^-1 near industrial cities in Xinjiang.In Qinghai,OC dominated the 450 nm absorption,and the fractional contribution reached approximately 70%.Although dust accounted for the main portion of the particulate mass,the fractional contribution from Fe to absorption?approximately 5%?was not significant.In Xinjiang,the fractions of absorption from BC?45%?and OC?50%?were comparable,and the effect of Fe was limited.Chemical analysis indicated that SO₄^2-/NO₃^-ratios and concentration of BC were lower in the remote northeast.In contrast,in industrial regions,higher SO₄^2-/NO₃^-ratios and concentration of BC were observed due to more anthropogenic pollutions.HYSPLIT back trajectory cluster analysis implied that the snow in Inner Mongolia and the remote northeast was significantly affected by the transported dust from desert areas and the influence of human activities was limited.But in industrial regions,local anthropogenic pollution emissions were the dominated sources of pollutants in snow.In northwestern China,we applied PMF model to explore the fractional contributions of different sources to the particulate absorption from 650 to 700 nm.In Qinghai,biomass burning was the primary source?59%?,and soil dust accounted for 29%of the particulate absorption despite its high mass contribution.In Xinjiang,the source attributions varied by region.In addition,BC concentrations showed a negative correlation with altitude in Xinjiang,which likely resulted from gradient variations in the contributions of industrial pollution sources.An evaluation indicated that the predominant source of chemical components was soil dust in Qinghai,while industrial pollution in industrial regions of Xinjiang.The fractional mass of BC,which is regarded as an important light absorbing particle,ranged from 0.2 to 4.8%and presented an average of 1.3%.A new radiative transfer model?Spectral Albedo Model for Dirty Snow,or SAMDS?was applied to simulate the spectral albedo of snow based on the asymptotic radiative transfer theory.Given the measured BC,MD,and OC concentrations of 100–5000,2000–6000,and 1000–30000ng g^-1,respectively,in surface snow across northeastern China,the SAMDS model produced a snow albedo in the range of 0.95–0.75 for fresh snow at 550 nm,with a snow grain optical effective radius(R_eff)of 100?m.The snow albedo reduction due to spherical snow grains assumed to be aged snow is larger than fresh snow such as fractal snow grains and hexagonal plate or column snow grains associated with the increased BC in snow.For typical BC concentrations of 100 ng g^-1 in remote areas and 3000 ng g^-1 in heavy industrial regions across northeastern China,the snow albedo for internal mixing of BC and snow is lower by 0.005 and 0.036 than that of external mixing for hexagonal plate or column snow grains with R_eff of 100?m.These results also showed that the simulated snow albedos by both SAMDS and SNICAR agree well with the observed values at low LAPs concentrations but tend to be higher than surface observations at high LAPs concentrations.In northeastern China,MODIS retrieved radiative forcing by LAPs in snow(RF^LAPs_MODIS)presented distinct spatial variability,with the minimum(31.6 W m^-2)in west of northeastern China and the maximum(72.6 W m^-2)near industrial regions.The mean RF^LAPs_MODIS was⁵2.7±6.6 W m^-2in December-March from 2003 to 2017.Local emissions of BC dominated the spatial distribution of RF^LAPs_MODIS.Furthermore,RF^LAPs_MODIS exhibited a decreasing trend from⁶0 W m^-2 in 2003 to⁴8 W m^-2 in 2017,likely due to an increase in snowfall.We also validated RF^LAPs_MODIS using in situ radiative forcing estimates(RF_in^es _situ^timated).We found that the biases in RF^LAPs_MODIS were negatively correlated with LAPs concentrations and ranged from²5%to³00%in northeastern China.