Assimilation of polar and geostationary satellite observations during RELAMPAGO using a WRF‐GSI‐LETKF system

CORRALES, PAOLA BELEN; CRAIG S. SCHWARTZ; JUAN RUIZ; GALLIGANI, VICTORIA

Simposio; 8th International Symposium on Data Assimilation (ISDA); 2022

We evaluated the impact of assimilating radiance observations from polar and geostationary satellitesover Argentina using a WRF-GSI-LETKF system with 10-km horizontal grid spacing. We conductedcontinuous hourly assimilation cycles from 0000 UTC 8 November to 0000 UTC 24 November 2018,where several convection cases developed over central and northeastern Argentina. The dataassimilation system had 60 ensemble members and used a multiphysics approach to represent modelerrors.The first assimilation experiment assimilated conventional observations, satellite-derived winds,radiances from different infrared and microwave sensors (i.e., AMSU, HIRS, MHS, ATMS, AIRS, andIASI) aboard several polar satellites, and radiances from GOES-16 ABI water vapor channels. Thesecond experiment assimilated everything from the first experiment except for infrared and microwaveradiances from the polar-orbiting satellites. Denial of these observations allows an assessment ofwhether they add value beyond the higher resolution GOES-16 ABI radiances for predicting thedevelopment of deep convection.Results from a case study during this period show a similar RMSE calculated against independentradiosonde observations measured during the RELAMPAGO field campaign for both experiments.This result indicates that the radiances from polar satellites do not provide new information to theanalysis when GOES-16 radiances are also assimilated. In particular, the assimilation of radiances fromGOES-16 modifies the humidity content in mid-levels and when the experiments are compared withthe observed precipitation (IMERG Final Run), precipitation forecasts were best in terms of arealcoverages and fractions skill score when only GOES-16 observations are assimilated.