Abstract
Many research and societal applications such as surface solar irradiance assessment and forecasting require accurate short-term cloudiness forecasts at kilometre and hourly scales. Today limited-area numerical weather prediction models have the potential to provide such forecasts by simulating clouds at high spatial and temporal resolutions. However, the forecast performance during the first 12-24 h is strongly influenced by the accuracy of the cloud and thermodynamic analyses in the initial conditions. Geostationary meteorological satellites provide valuable observations that can be used in data assimilation for frequent cloud analysis determination. This paper provides an up-to-date review of the state of the art in cloud-related geostationary satellite data assimilation with limited-area models dedicated to improve cloudiness forecast performance. Research and operational studies have been reviewed by differentiating between satellite radiance and cloud property retrieval assimilation. This review gives insight into the best practices considering the large variety of limited-area models, data assimilation methods, satellite sensors and channels, cloud property retrieval products and various methodological challenges. Cloud analysis methods for regional models have become more sophisticated in recent years and are increasingly able to exploit observations from geostationary satellites. Important proofs of concept have been performed in this decade, paving the way for an optimal synergy of geostationary satellite data assimilation and convection-permitting limited-area model forecasts. At the same time, the increasing amount of channels of geostationary satellite instruments leads to more opportunities and challenges for data assimilation methods.