The challenge of modeling low-level cloudiness in the West African monsoon region
The southern parts of West Africa are frequently covered by an extensive deck of shallow, low (200–400m above ground) stratus or stratocumulus clouds during the summer monsoon season. Climatologically, the ultra-low stratus decks form after sunset along the Guinea coast, spread inland and reach about 10-11°N between 0900 and 1000 UTC of the following day. In the late morning the northern boundary gets fragmented due to the breakup of stratus decks into fair-weather cumuli. A thorough validation of weather and climate models with respect to this process is challenging due to a lack of local observations and a misrepresentation of ultra-low cloud decks in most satellite products. Nevertheless it could be shown that Coupled Model Intercomparison Project 3 (CMIP3) global climate models largely failed to represent these clouds, leading to errors in surface solar radiation of up to 90 W m−2. New results using CMIP5 and YOTC (Years of Tropical Convection) GASS (GEWEX Atmosphere System Study) global climate models show that the representation of low-level clo ... mehrudiness has not improved in the latest generation of global models. Large systematic errors are even found in much higher-resolution model simulations, e.g. using the COSMO model. Recent high-resolution simulations using the WRF model in contrast have shown promising results after a careful selection of boundary layer and surface schemes. They reveal that the night-time cloud formation is related to a subtle balance between ‘‘stratogenic’’ upward (downward) fluxes of latent (sensible) heat caused by shear-driven turbulence below the night-time low-level jet, cold advection, orographic lifting and radiative cooling on one hand, and ‘‘stratolytic’’ dry advection and latent heating on the other hand. Results from these and other new simulations are currently being used to better understand deficits in other models and to explore impacts on the larger-scale monsoon circulation and precipitation.