Research On Rapid Calculation Scheme Of Atmospheric Absorption Radiation And Remote Sensing Application

There are two main factors that affect the accuracy of gas radiative absorption.The first factor is the error of gas absorption scheme in fast radiation transfer model,and the second one is the uncertainty of line data in the spectral database.In this study,we expand the alternate mapping correlated k-distribution scheme?AMCKD?into satellite remote sensing field to improve the accuracy of radiative transfer calculation.The AMCKD method is applied to the Advanced Himawari Imager?AHI?on board the Himawari-8 and Medium Resolution Spectral Imager?MERSI?on board the Fengyun-3D to simulate the reflectance and brightness temperatures of the top of atmosphere?TOA?under clear sky.The accuracy of AMCKD is quantified by the comparison between rigorous line-by-line?LBL?result and AMCKD simulation.The relative errors of TOA reflectance are less than 0.7%for all shortwave channels,and the absolute errors of brightness temperatures at longwave channels are less than 0.4 K at any satellite zenith angle under the typical atmospheric profiles.Furthermore,the clear atmosphere brightness temperatures based on AMCKD are highly consistent with AHI and MERSI observations in a practical case,which strongly confirms the excellent performance of AMCKD in remote sensing application.The most widely used spectroscopic database in atmospheric sciences is the High Resolution Transmission?HITRAN?database.The HITRAN database has been updated recently.Therefore,it is necessary to analyze the effect of using different version of databases on radiative transfer calculations.This study evaluates the impact of database updates on climate models and satellite remote sensing applications.The radiative fluxes and cooling rates of the longwave region are calculated by the line-by-line radiative transfer model based on the HITRAN08,HITRAN12 and HITRAN16 molecule spectroscopic databases under four typical atmospheric profiles.The differences in the cooling rates based on different databases are mainly caused by the changes in the line parameters of H₂O and CO₂.The maximum difference in the upward fluxes at the top of the atmosphere between HITRAN08 and HITRAN16 is 0.6587 W m^-2,whereas that between HITRAN12 and HITRAN16 is only0.0406 W m^-2 which is much smaller than the maximum difference between HITRAN08and HITRAN16.Similar features are also observed in the downward fluxes at the surface and cooling rates.In climate model,these differences in radiative fluxes and cooling rates can be ignored as they are less than the errors of radiative transfer schemes themselves.However,the largest difference of simulated brightness temperature for the channel of hyperspectral measurements such as FY-3D Hyperspectral Infrared Atmospheric Sounder?HIRAS?is comparable to the noise equivalent temperature difference of the corresponding channels.Therefore,it is necessary to use the updated HITRAN molecular spectroscopic databases to reduce errors in hyperspectral satellite remote sensing applications.