Application Of The Rapid Radiative Transfer Model RTTOV In Simulating The Brightness Temperature Of FY-3D HIRA

As an important tool for the study of atmospheric state variables and satellite remote sensing,RTTOV as a fast radiative transfer mode with the advantages of fast computing speed,wide applicability and high simulation accuracy.HIRAS（High-Spectral Infrared Atmospheric Sounder）on board China FY-3D,which provides high resolution observations of the atmosphere in the vertical direction.This paper carries out a study of the sensitivity of the vertical stratification scheme to RTTOV when modelling radiation values and the upper atmospheric sensitive channels in HIRAS.RTTOV simulates the satellite mounted instrument with any number level of atmospheric state profiles,and RTTOV provides a total of five interpolation schemes.By comparison with LBLRTM,when the number of input profile level is greater than the number of coefficient file level,the interpolation schemes show little difference in the simulation results in the window area and water vapour channels,with less deviation in the carbon dioxide sounding channel for interpolation schemes 4 and 5.Interpolation schemes 1,2 and 3 show small oscillations in the temperature Jacobian profile in the vertical direction,and the water vapour obtained using interpolation scheme 2 Jacobian matrix exhibits strong oscillations.Conversely,when the number of input profile layers is smaller than the number of coefficient file layers,scheme 1 deviates slightly more than the other interpolation schemes in most channels,but the differences between the five interpolation schemes are small,and both the temperature Jacobian matrix and the water vapour Jacobian matrix are smoother and similar in shape.980-1070 cm^-1 and 700-800 cm^-1 bands in HIRAS are sensitive to ozone content in order to investigate the effect of ozone input on the simulated brightness temperature of RTTOV,the brightness temperature observed by HIRAS was used as the real value and compared with the simulated results,and it was found that the difference between the simulated results with and without ozone data and the real brightness temperature was larger in the channels in the strong ozone absorption band,and the difference could be more than 15K.It can be seen that the input ozone data greatly reduces the error of the brightness temperature simulation,so when The 660-700 cm^-1 band in HIRAS is sensitive to changes in the upper carbon dioxide content,but sensitivity tests on the carbon dioxide concentration show that whether or not carbon dioxide is input has essentially no effect on the simulated brightness temperature.In the carbon dioxide absorption band,the difference between the input carbon dioxide profile and the simulated brightness temperature using the default values in RTTOV is only 1 K.For other non-carbon dioxide absorption bands the resulting brightness temperature difference is almost 0 K.Therefore,the input of carbon dioxide profiles can be disregarded when assimilating upper level data.