AVHRR-Derived Land Surface Conditions for Flux Simulations with a Mesoscale Model over the HAPEX-Sahel Study Area

Abstract

The description of land surface conditions at a spatial scale adapted to climate and meteorological models is at the core of major problems in environment studies. In this regard, the information routinely provided by remote sensing observations is fundamental. Presented herein is a technique to obtain regional and seasonal maps of key surface properties involved in mass and momentum flux exchanges at the atmospheric boundary layer. The investigated zone encompasses landscape units that are typical of the natural semiarid ecosystem forming the Hydrologic Atmospheric Pilot Experiment in the Sahel (HAPEX-Sahel) study area. The selected surface parameters?albedo, fractional vegetation cover, leaf area index, and absorbed photosynthetic radiation?are mapped from times series of Advanced Very High Resolution Radiometer (AVHRR) data. First, the inversion of a bidirectional reflectance model against AVHRR data provides a set of coefficients for composite time periods. Further, the latter serve to simulate spectral and directional vegetation indices, which are finally linked to the surface parameters. The resulting 1-km-resolution maps show the high degree of spatial heterogeneity of the Sahelian landscapes and the presence of a strong north?south vegetation gradient implied by the precipitation regime. The database of generated surface parameters is evaluated for use in the Nonhydrostatic Mesoscale (Meso-NH) model. In comparing modeled and measured heat and water fluxes, it is shown that the proposed AVHRR-derived maps of land surface conditions are more reliable than a landscape prescription based on a land-cover map and lookup tables with a global vegetation nomenclature. Based on these results, it is recommended that spatially distributed land surface products be considered in meteorological and climate models.