Assessing the impact of passive microwave cloudy and precipitating observations onto the three‐dimensional ensemble‐variational kilometer‐scale AROME numerical weather prediction system

Abstract Passive microwave observations are highly sensitive to many variables (temperature, humidity, hydrometeors, etc.) that are useful for numerical weather prediction (NWP), especially in an all‐sky context. Several studies have demonstrated the positive impact of all‐sky passive microwave assimilation in cloudy and rainy areas on global NWP model forecasts, but the impact is less documented on regional‐scale NWP models. The aim of this study is to investigate the impact of the assimilation of passive microwave observations in cloudy and precipitating areas in the operational kilometre‐scale NWP model AROME of Météo‐France. In particular, this study takes advantage of the recent works performed at Météo‐France, allowing the direct initialization of hydrometeors in the three‐dimensional ensemble‐variational (3DEnVar) assimilation system of AROME, and the all‐sky assimilation method of European Center for Medium‐Range Weather Forecasts for microwave radiances. In this context, we examine the impact of the all‐sky assimilation of four passive microwave sensors (microwave humidity sounder, microwave humidity sounder‐2, GPM microwave imager, and advanced microwave scanning radiometer 2) in the 3DEnVar kilometre‐scale AROME NWP on forecasts up to 48 hr. To this end, cycled experiments were carried out over two 1‐month periods. For both periods, there is a reference experiment assimilating the four sensors in clear‐sky conditions only and an all‐sky experiment, assimilating them in all‐sky conditions. Statistical studies were carried out to assess the impact on precipitation, clouds, temperature, relative humidity, geopotential height, and wind forecasts in order to evaluate the benefits of the all‐sky assimilation of the four sensors. Significant improvements are demonstrated in 1‐hr forecasts, especially regarding rainfall. For longer range forecasts, the results indicate both positive and negative effects, although there is a general trend toward a slight significant improvement.