5:15 PM - 6:45 PM
[ACG36-P18] Development for assimilation of superobbed AMSR3 humidity sounding channels in the JMA’s global NWP system.
Keywords:Microwave radiance data, Data assimilation, Numerical weather prediction
Microwave radiance data including JAXA’s Advanced Microwave Scanning Radiometer 2 (AMSR2) have been assimilated in the JMA’s numerical weather prediction (NWP) systems. The assimilation of those data has significantly improved NWP skills. As a successor of AMSR2, JAXA plans to operate AMSR3 carried by the Global Observing SATellite for Greenhouse gases and Water cycle (GOSAT-GW) in this fiscal year. AMSR3 will have three additional high-frequency channels (165.5 GHz, 183±3 GHz and 183±7 GHz, V-pol) to measure solid particles and humidity profile in the atmosphere. JMA is preparing to assimilate these channels more effectively.
Duncan et al., 2023 indicated that averaging (superobbing) the humidity sounder radiances provide more representative observations for assimilation and a better utilization of total information content than thinning native resolution radiances. In the current JMA’s global NWP system, low frequency channels (19 GHz, 24GHz, 36GHz, V-pol) are superobbed with 4D-Var inner model grid spacing (approximately 55 km horizontal resolution) and thinned with 150 km grid box, but high frequency channels (183GHz, V-pol) are just thinned with 250 – 180 km grid box prior to assimilation. We attempted to apply superobbing to humidity sounder radiance data. We conducted an initial experiment to assimilate superobbed humidity sounder radiances of ATMS and MWHS-2 in JMA's global NWP system. Superobbing decreased standard deviation of FG departure (approximately 6 ~ 20 %) and increased the number of used data (6 times in conversion without superobbing) of ATMS and MWHS-2. In addition, assimilation of superobbed radiance decreased standard deviation of FG departure for CSR and MHS. These results indicate that short-range forecast of mid- to upper-tropospheric humidity fields were improved. The same effects are expected by applying this method to other humidity sounders (GMI, MHS, SSMIS) and the results will be presented in the conference.
Duncan et al., 2023 indicated that averaging (superobbing) the humidity sounder radiances provide more representative observations for assimilation and a better utilization of total information content than thinning native resolution radiances. In the current JMA’s global NWP system, low frequency channels (19 GHz, 24GHz, 36GHz, V-pol) are superobbed with 4D-Var inner model grid spacing (approximately 55 km horizontal resolution) and thinned with 150 km grid box, but high frequency channels (183GHz, V-pol) are just thinned with 250 – 180 km grid box prior to assimilation. We attempted to apply superobbing to humidity sounder radiance data. We conducted an initial experiment to assimilate superobbed humidity sounder radiances of ATMS and MWHS-2 in JMA's global NWP system. Superobbing decreased standard deviation of FG departure (approximately 6 ~ 20 %) and increased the number of used data (6 times in conversion without superobbing) of ATMS and MWHS-2. In addition, assimilation of superobbed radiance decreased standard deviation of FG departure for CSR and MHS. These results indicate that short-range forecast of mid- to upper-tropospheric humidity fields were improved. The same effects are expected by applying this method to other humidity sounders (GMI, MHS, SSMIS) and the results will be presented in the conference.