future projections and the development of adequate adaption measures. Low-level clouds over
southern West Africa (5°–10°N, 8°W–8°E) during July–September are an integral part of the WAM through
their effect on the surface energy balance and precipitation, but their representation in climate models has
received little attention. Here 30 (20) years of output from 18 (8) models participating in phase 5 of the
Coupled Model Intercomparison Project (Year of Tropical Convection) are used to identify cloud biases and
their causes. Compared to ERA-Interim reanalyses, many models show large biases in low-level cloudiness of
both signs and a tendency to too high elevation and too weak diurnal cycles. At the same time, these models
tend to have too strong low-level jets, the impact of which is unclear because of concomitant effects on
temperature and moisture advection as well as turbulent mixing. Part of the differences between the models
and ERA-Interim appear to be related to the different subgrid cloud schemes used. While nighttime tendencies
in temperature and humidity are broadly realistic in most models, daytime tendencies show large
problems with the vertical transport of heat and moisture. Many models simulate too low near-surface relative
humidities, leading to insufficient low cloud cover and abundant solar radiation, and thus a too large diurnal
cycle in temperature and relative humidity. In the future, targeted model sensitivity experiments will be
needed to test possible feedback mechanisms between low clouds, radiation, boundary layer dynamics,
precipitation, and the WAM circulation.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner