[A-CG38] Satellite Earth Environment Observation

[ACG38-04] Effect of retrieval of land surface temperature derived from CSRs at window band on assimilation of CSRs at CO2 band

*Izumi Okabe¹, Kozo OKAMOTO¹, Toshiyuki Ishibashi¹ (1.Meteorological Research Institute of Japan Meteorological Agency)

Keywords:data assimilation, CSR, Land surface temperature

Clear-sky radiance (CSR) data derived from Himawari-8, GOES-16/-17, Meteosat Second Generation (MSG)-1/-4 are currently used in the JMA’s operational global data assimilation system. CSRs at water vapor (WV) bands of 6.2, 6.9 and 7.3 um are assimilated in order to obtain information about WV amount at the mid and upper troposphere for analysis of the system. One of the reasons for CSRs at other infrared band are discarded is accuracy of surface parameters used for radiative transfer (RT) calculation is not enough in the system. We tried CSRs at CO2 band of 13.3 or 13.4 um assimilation despite the difficulty to extract to the information about temperature at the low troposphere. Retrieval of land surface temperatures (LSTs) derived from CSRs at window band of 10.8 um was used instead of first-guess (FG) LST in RT calculation for CO2 and WV band CSRs. The validation using GCOM-C LST product showed that the retrieval has higher accuracy than FG LST. The result of assimilation experiments showed that use of CO2 band CSRs with the LST retrieval successfully improved analysis accuracies and forecasts scores of not only temperature but also relative humidity and wind speed. The improvement was especially found at the mid and low troposphere in the summer hemisphere and the tropics. On the other hand, degradation of those accuracies and scores were found when we assimilated CO2 band CSRs without the retrieval LST. These results confirmed the accurate LST is necessary for CO2 band CSRs assimilation. The details of the result will be shown in the presentation.