Impact of Atmospheric Motion Vectors in an Operational Global NWP System : A Multi-Year Study
Geostationary and polar-orbiting satellite derived “Atmospheric Motion Vectors (AMV)” are one of the most important wind observations used in the Numerical Weather Prediction (NWP) systems and their assimilation have significantly reduced the forecast errors. National Centre for Medium Range Weather Forecasting (NCMRWF) global Unified Model (NCUM-G) employs hybrid 4D-Var data assimilation scheme for the assimilation of observations. Currently, NCUM-G uses AMVs from number of geostationary (INSAT-3D & 3DR satellites, Himawari satellite, GOES satellites, Meteosat satellites) and several polar-orbiting (NOAA and MetOp satellites and others) satellites. The impact of AMV products on the short-term forecast skill of NCUM-G has been evaluated using an adjoint based Forecast Sensitivity to Observation Impact (FSOI) method operationally. In this study, impact of AMV calculated using FSOI for the five Indian summer monsoon seasons (June to September), spanning 2018 to 2022 are considered. The study shows that AMVs from Geostationary platform have high beneficial contribution on forecast compared to that from Polar satellites, even though impact per observation is smaller than that from polar satellites. This may be largely due to high data volume and high temporal resolution of Geostationary AMVs. Beneficial impact of Meteosat satellites on the short-range forecast of NCUM-G has been the highest followed by GOES and Himawari observation whereas the impact of INSAT has been marginal yet have considerable contribution relative to other observation type. The forecast impact of the wind lidar observation (Aeolus satellite) has also been compared with that of AMV, wherever possible.
