Pluriannual comparisons of satellite-based rainfall products over the Sahelian belt for seasonal vegetation modeling
Résumé
The Sahel corresponds to the transition from the dry arid desert to wet savannahs, where vegetation exhibits a well-marked seasonal cycle in response to the West African Monsoon. Precipitation data sets with high spatial and temporal resolutions are therefore relevant to investigate the dynamics of the Sahelian vegetation. Three satellite-based precipitation products (TRMM3B42, RFE2.0, and CMORPH) are compared and tested against kriged rain gauge measurements. The objective is to evaluate their capability to retrieve the main precipitation characteristics during the rainy season. Comparisons are performed over a 4 year period (2004-2007) at spatial resolutions of 0.25° × 0.25° or 0.5° × 0.5° by looking at sensitive criteria for vegetation: spatial distribution of the rainfall field, precipitation frequency, dry spell distribution, and precipitation amounts. Intercomparisons between satellite data sets are conducted over the Sahelian belt (10°N-20°N; 20°W-35°E) at a 1-10 day time scale, while comparisons with 10 day kriged rain gauge measurements are performed over a smaller area (10°N-17.5°N; 17.5°W-2.5°E). The precipitation spatial distributions are in good agreement between satellite products and with the kriged data. Considering the daily frequency, the satellite products show a high agreement between them (∼80%). The TRMM3B42 product exhibits the lowest number of rainy days, and RFE2.0 exhibits the highest. The CMORPH product overestimates rainfall amounts, while TRMM3B42 and RFE2.0 are both in good agreement with the kriged data. The impacts of these distinctive behaviors on simulated vegetation are investigated by comparisons with MODIS LAI, considering vegetation dynamics and amounts. The studied criteria of precipitation fields appear as a critical issue for Sahelian vegetation modeling.
Domaines
Géomorphologie
Origine : Fichiers éditeurs autorisés sur une archive ouverte