Abstract
In the sub-Saharan Sahel, energy and water cyclingat the land surface is pivotal for the regional climate,water resources and land productivity, yet it is still verypoorly documented. As a step towards a comprehensive climatological description of surface fluxes in this area, thisstudy provides estimates of long-term average annual budgetsand seasonal cycles for two main land use types of thecultivated Sahelian belt: rainfed millet crop and fallow bush.These estimates build on the combination of a 7-year fielddata set from two typical plots in southwestern Niger withdetailed physically based soil-plant-atmosphere modeling,yielding a continuous, comprehensive set of water and energyflux and storage variables over this multiyear period. Inthe present case in particular, blending field data with mechanistic modeling makes the best use of available data andknowledge for the construction of the multivariate time series.Rather than using the model only to gap-fill observationsinto a composite series, model-data integration is generalizedhomogeneously over time by generating the wholeseries with the entire data-constrained model simulation. Climatological averages of all water and energy variables, withassociated sampling uncertainty, are derived at annual to subseasonal scales from the time series produced. Similaritiesand differences in the two ecosystem behaviors are highlighted.Mean annual evapotranspiration is found to represent82-85% of rainfall for both systems, but with differentsoil evaporation/plant transpiration partitioning and differentseasonal distribution. The remainder consists entirelyof runoff for the fallow, whereas drainage and runoff standin a 40-60% proportion for the millet field. These resultsshould provide a robust reference for the surface energy- andwater-related studies needed in this region. Their significanceand the benefits they gain from the innovative data-modelintegration approach are thoroughly discussed. The modeldeveloped in this context has the potential for reliable simulations outside the reported conditions, including changingclimate and land cover.