Recent studies suggest that the accurate representation of the low-level water vapor is a crucial for quantitative precipitation forecast. However, mesoscale observations of moisture usually are not available for most regions around the world. An Infra-Red Sounding (IRS) Mission on the Meteosat Third Generation (MTG) would provide high-resolution (in both space and time) temperature and water vapor information. Assimilating these observations into a mesoscale model is expected to improve skills in regional weather forecast. To evaluate such potentials, quantitative analyses of the added values of the IRS candidate mission for regional forecasts are performed by the means of Observing System Simulation Experiment (OSSE).

An OSSE of a series of convective storms occurred during 11 to 16 June 2002 has been conducted. A 5-day nature or “truth” run is generated with a high-resolution of 4-km using the Penn-State University/National Center for Atmospheric Research (NCAR) Mesoscale Model Version 5 (MM5). The conventional observations and MTG-IRS retrieved temperature and humidity profiles are simulated from the “truth”. These observations are assimilated using the NCAR Weather Research and Forecasting (WRF) model and variational data assimilation system (WRF-Var) in cycling mode. Numerical forecasts initialized from the analyses are then carried out.

To calibrate the OSSE setup, data assimilation experiments using real conventional observations are conducted. Assimilating real or simulated conventional observations give similar error statistics in analyses and forecasts.

The results of data assimilation and forecast experiments show that, on average, the MTG retrieved profiles have positive impact on the analysis and forecast. The analysis reduces the errors in not only the temperature and the humidity, but also in the horizontal wind fields.