No abstract

No abstract

Surface moisture supply is known to be a factor which can be important in triggering convection. This paper aims to investigate its impact on short-range numerical forecasts of tropical rainfall.The ECMWF (European Centre for Medium Range Weather Forecasts) operational forecasting model was integrated for 5 days with realistic positive and negative anomalies of the initial surface moisture distribution in the West African Sahel. The large-scale flow is found to be relatively unaffected, but significant differences occur in the rainfall forecasts for about the first 4 days. Rainfall is greatest with a moister surface, except on days 3 and 4 in the south Sahel/Savanna region, for which a decrease of rainfall occurs. The opposite result is seen when an initially drier surface is introduced. The mechanisms involved are investigated by carrying out a detailed hydrological budget, and by studying the evolution of the atmospheric profiles of moisture and temperature.We conclude that an improved surface moisture analysis is likely to result in some improvement of shortrange rainfall forecasts in the Sahel, and presumably also in other tropical regions. In the coming years it should be possible to achieve this using satellite-derived soil wetness maps.

We present analytical and numerical calculations of the effective roughness length (ERL) over a flat surface with varying roughness elements, for use in large-scale models. It is shown that ERL is mostly determined by the roughest elements present inside the averaging domain and that, more surprisingly, the ERL increases as the first level of the numerical model gets closer to the surface and its altitude approaches the value of the largest local roughness length. This effect further increases the drag coefficient, in addition to the well-known increase due to the lowering of the first model level.

No abstract