Remote Sensing Inversion Of Surface Characteristic Parameters And Energy Fluxs Over The Taklimakan Desert

Due to its unique natural climate characteristics,the Taklimakan Desert plays a very important role in China's weather system and environmental system,and is also an ideal area for studying special land surface processes.Therefore,the study of land surface process parameterization and surface energy flux variation characteristics in this area can deepen the understanding of the surface-atmosphere interaction in this area.Domestic and international researches have been carried out on surface characteristic parameters and energy fluxes,and there are relatively more researches on surface characteristic parameters.However,with the continuous deepening of research and the expansion of the research area,traditional measurement methods can not meet the needs of regional research.Hence,the emergence of remote sensing technology provides a certain technical means for the study of regional scale,which is of great significance to the study of the change characteristics of surface characteristic parameters and energy fluxes as well as the dynamic monitoring of regional scale.In this paper,three kinds of remote sensing satellites,MOIDS,GLASS and FY-2E,and according to the measured data of FTIR and ASD spectrometers and the site observation data from the Taklimakan Desert Atmospheric Environment Observation Test Station were used to retrieve the surface characteristic parameters(surface emissivity,surface albedo,surface temperature),and surface energy fluxes of Taklimakan Desert,respectively.Based on the characteristics of the surface characteristics of the Taklimakan Desert and the spatial distribution of energy fluxes,the site data was used for verification and comparative analysis,below are the main conclusions:(1)The inversion equation of broadband surface emissivity of MODIS and GLASS were modified by the measured data of FTIR.The modified broadband surface emissivity and the measured value of R² have reached over 0.9.Based on the spectral response function of MODIS and FY-2E,the surface emissivity based on FY-2E was inversed.(2)Based on the albedo data measured by ASD spectrometer,the albedo of MODIS and GLASS in broadband was inversed and the coefficients of their empirical equations were modified.The modified broadband albedo and the measured data's R²were reached above 0.8.And obtained the broadband surface albedo spatial distribution characteristics based on MODIS and GLASS in the Taklimakan Desert.(3)According to the empirical algorithm,the land surface temperature from MODIS,GLASS and FY-2E were retrieved,and the spatial distribution characteristics of surface temperature based on MODIS and GLASS were obtained.The daily variation characteristics of surface temperature based on FY-2E were obtained and the inversion results of site observation data were used to verify that R² was above 0.9.(4)Using the surface characteristic parameters and site observation data obtained from the inversion,select the best three-day data of image quality in the June,July and August with the highest closure rate to inverse the surface energy flux of Taklimakan Desert,and the inversion results of the surface energy flux based on MODIS,GLASS and FY-2E in June,7 and 8 months of 2014 were obtained.The inversion results of three kinds of remote sensing satellite data show the rule of net radiation>sensible heat flux>soil heat flux>latent heat flux,and the station observation data were used for verification and comparative analysis.The average inversion errors of the net radiant fluxes of MODIS,GLASS,and FY-2E were 16.3 W·m^-2,50.04 W·m^-2,and 37.16W·m^-2,respectively.The average inversion errors of soil heat flux and sensible heat flux of MODIS and FY-2E were 36.88 w·m^-2,15.33 W·m^-2,and 46.36 W·m^-2,21.33 W·m^-2,respectively.The mean inversion errors of latent heat fluxes from three kinds of satellite data were 2.67 W·m^-2,1.13 W·m^-2 and 4.49 W·m^-2,respectively.