A mesoscale simulation of land surface heterogeneity from HAPEX-Sahel

A mesoscale version of the UK Meteorological Office Unified Model has been used to study the evolution of the planetary boundary layer over heterogeneous terrain in the HAPEX-Sahel region. The model simulates a typical diurnal cycle during the dry down period of the intensive observation period on 8 October 1992. The model has a grid length of 10 km, with a domain covering a region of approximately 1000 km(2), centred on the HAPEX experiment. Satellite images are used to generate a mesoscale land cover map over the region. To initialise the soil moisture, a simple soil moisture accounting model was developed. This generates soil moisture patterns from daily rainfall estimates over the entire mesoscale domain. Where possible, the land surface scheme has been calibrated with measurements taken during the experiment. Moisture availability is the principle component in determining the partition of energy at the surface. The mean midday Bowen ratio of the square is 0.62. However, spatial contrasts in soil moisture availability lead to a range of midday sensible heat fluxes of between 100 W m(-2) and 280 W m(-2) The pattern of surface fluxes is closely related to the mesoscale variation of the seasonal rainfall. Comparisons of the model output with surface meteorology, surface fluxes, sonde and aircraft measurements are encouraging. Both model and climate observations show similar gradients at the surface across the square. Features of the aircraft measurements are well represented by the model. Anomalous increases in latent heat Bur with height measured by the aircraft are also found in the model. This may be related to high entrainment fluxes at the inversion separating contrasting air flows. Heat and moisture budget calculations show the importance of large scale flow in the evolution of the boundary layer. Periods of low level south-westerly flow for example produce considerably larger inputs of moisture into the planetary boundary layer than surface evaporation. In this study, the influence of surface flux variability on the boundary laver is limited. However, scaling surface fluxes up to larger scales must take some account of the considerable variation in soil moisture.