Assimilation of Screen-Level Variables in ECMWF’s Integrated Forecast System: A Study on the Impact on the Forecast Quality and Analyzed Soil Moisture

Abstract

Abstract In many operational numerical weather prediction applications, the soil moisture analysis is based on the modeled first-guess and screen-level variables; that is, 2-m temperature and 2-m relative humidity. A set of two global 61-day analysis/forecast experiments based on the Integrated Forecast System at the European Centre for Medium-Range Weather Forecasts (ECMWF) has been performed for June and July 2002. Analyses and forecasts based on the operational Optimal Interpolation (OI) scheme are compared against results obtained from an open loop system, in which soil moisture evolves freely. It is found that soil moisture assimilation or analysis has a significant impact on the model atmosphere. Temperature forecasts for the Northern Hemisphere up to a level of 700 hPa and up to nine days were significantly improved when the operational analysis was used. A comparison of volumetric soil moisture against in situ observations from the Oklahoma Mesonet reveals, however, that the operational OI system fails to improve both the analysis and the subsequent forecast of soil moisture itself. In addition, the system is not able to correct soil moisture for errors introduced through wrong precipitation in the background forecasts. Biweekly observations from the Illinois Climate Network support these findings. This study confirms the long assumed (but rarely proven) characteristics of analysis schemes using screen-level variables: The observations are efficient in improving the turbulent surface fluxes and consequently the weather forecast on large geographical domains. The quality of the resulting soil moisture profile is often not sufficient for hydrological or agricultural applications.