Using microwave imager observations to improve the ocean temperature in a coupled NWP system
Knowledge of the ocean temperature is vital to the accurate use of satellite radiances in weather forecasting and it helps improve the quality of forecasts for the ocean and atmosphere. A recent upgrade to the European Centre for Medium‐Range Weather Forecasts (ECMWF) Integrated Forecasting System (IFS) has implemented outer‐loop coupling between the atmosphere and the ocean. The current work uses this method to pass ocean skin‐temperature estimates derived from satellite‐borne microwave imagers to the ocean data assimilation (DA) system; the updated ocean state is then passed back to the atmospheric DA system. It is shown that these skin temperature estimates improve the fit of the short‐range forecast ocean temperature profiles to Array for Real‐Time Geostrophic Oceanography (ARGO) float observations, resulting in up to 4% improvements in the Eastern Tropical Pacific. Improvements in the forecast skin temperature are also recorded indirectly through improvements in the fit of significant wave height observations and infrared channels sensitive to the surface on‐board geostationary satellites. In terms of the atmospheric forecast verification, including the mean fields, the most significant changes are apparent in the medium range (2 to 6 days). Statistically significant improvements are seen in the Tropics for relative humidity at the surface and in the Southern Hemisphere for temperature, vector winds and geopotential height.