Building a field- and model-based climatology of surface energy and water cycles for dominant land cover types in the cultivated Sahel. Annual budgets and seasonality

In the sub-Saharan Sahel, energy and water cycling at the land surface is pivotal for the regional climate, water resources and land productivity, yet it is still very poorly documented. As a step towards a comprehensive climatological description of surface fluxes in this area, this study provides estimates of long-term average annual budgets and seasonal cycles for two main land use types of the cultivated Sahelian belt: rainfed millet crop and fallow bush. These estimates build on the combination of a 7-year field data set from two typical plots in southwestern Niger with detailed physically based soil–plant–atmosphere modeling, yielding a continuous, comprehensive set of water and energy flux and storage variables over this multiyear period. In the present case in particular, blending field data with mechanistic modeling makes the best use of available data and knowledge for the construction of the multivariate time series. Rather than using the model only to gap-fill observations into a composite series, model–data integration is generalized homogeneously over time by generating the whole series 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. Similarities and differences in the two ecosystem behaviors are highlighted. Mean annual evapotranspiration is found to represent 82–85% of rainfall for both systems, but with different soil evaporation/plant transpiration partitioning and different seasonal distribution. The remainder consists entirely of runoff for the fallow, whereas drainage and runoff stand in a 40–60% proportion for the millet field. These results should provide a robust reference for the surface energy- and water-related studies needed in this region. Their significance and the benefits they gain from the innovative data–model integration approach are thoroughly discussed. The model developed in this context has the potential for reliable simulations outside the reported conditions, including changing climate and land cover.