Improvements in the Forecasts of Near-surface Variables in the Global Forecast System (GFS) via Assimilating ASCAT Soil Moisture Retrievals

Abstract Recent research has shown that assimilating satellite soil moisture (SM) retrievals into the land surface models (LSMs) improves simulations of land-atmosphere water and energy exchanges. With satellite SM retrievals becoming widely and continuously available, it is desirable to examine the impact of assimilating them into numerical weather prediction models in order to improve numerical weather forecast skills. Based on the development of the coupled system of National Centers for Environmental Prediction (NCEP)-Global Forecast System (GFS) and National Aeronautics and Space Administration (NASA)-Land Information System (LIS) in this paper, we designed an experiment to demonstrate the impacts of assimilating the Advanced Scatterometer (ASCAT) SM data products on the weather forecasts of GFS. With respect to the surface air temperature analysis product of National Oceanic and Atmospheric Administration (NOAA)-Climate Prediction Center (CPC) and CPC’s morphing method-based precipitation data, improvement from the ASCAT SM assimilation for probabilities of high quality forecasts can reach up to 1.7% for GFS precipitation, 3.1% for 2-meter minimum temperature, and 3.1% for 2-meter diurnal temperature range predictions, respectively. These results suggest that satellite SM data assimilation could be beneficial for GFS numerical weather forecasts of NOAA NCEP.