The impact of hyperspectral radiances in the NCUM 4D-Var data assimilation system

Abstract

ABSTRACT This paper describes the combined impact of clear sky hyperspectral radiances from four space-based instruments in the National Centre for Medium Range Weather Forecasting (NCMRWF) Unified Model (NCUM) 4D-Var assimilation and forecast system through Observing System Experiment (OSE) during May 2018. Radiances from four hyperspectral instruments, Atmospheric Infrared Sounder (AIRS) onboard EOS-AQUA satellite, Infrared Atmospheric Sounding Interferometer (IASI) onboard both MetOp-A and MetOp-B satellites and Cross-track Infrared Sounder (CrIS) onboard NPP satellite along with other conventional and satellite observations are assimilated in the first simulation (EXP), while the radiances from the four hyperspectral instruments were denied in the latter simulation (CNTL). The simulations run four assimilation cycles per day centred at 0000, 0600, 1200, and 1800 UTCs. 5-day forecasts based on the 0000 UTC initial conditions are generated from both EXP and CNTL. Results show that both O-B (Observation-Background) and O-A (Observation – Analysis) of other satellite radiances assimilated in EXP improved (in this case biases reduced) due to the assimilation of hyperspectral radiances. The impact of hyperspectral data assimilation in various forecast fields is estimated by computing Forecast Impact (FI) parameter with respect to an independent dataset, ERA5. Hyperspectral radiance assimilation improved both mass and wind forecasts in the short-range, and the FI degraded with higher lead time. There is a slight improvement in the precipitation over the Indian landmass due to the assimilation of hyperspectral radiances. Verification of model forecasts with radiosonde observations over the Indian region shows marginal improvement in the humidity and temperature forecasts due to the assimilation of hyperspectral radiances. Assimilation of hyperspectral radiances slightly improved the characteristics of two cyclones, Sagar and Mekunu, formed over the Arabian Sea during the study period. Both the experiments predicted early onset of monsoon 2018 than the IMD date of monsoon onset.