Groundwater in Africa supports livelihoods and poverty alleviation 1,2 , maintains vital ecosystems, and strongly influences terrestrial water and energy budgets 3. However, hydrologic processes governing groundwater recharge sustaining this resource, and their sensitivity to climatic variability, are poorly constrained 4,5. Here we show, through analysis of multi-decadal groundwater hydrographs across sub-Saharan Africa, how aridity controls the predominant recharge processes whereas local hydrogeology influences the type and sensitivity of precipitation-recharge relationships. Some humid locations show approximately linear precipitation-recharge relationships with small rainfall intensity exceedance thresholds governing recharge; others show surprisingly small variation in recharge across a wide range of annual precipitation. As aridity increases, precipitation thresholds governing initiation of recharge increase, recharge becomes more episodic, and focussed recharge via losses from ephemeral overland flows becomes increasingly dominant. Extreme annual recharge is commonly associated with intense rainfall and flooding events, themselves often driven by largescale climate controls. Intense precipitation, even during lower precipitation years, produces substantial recharge in some dry subtropical locations, challenging the 'high certainty' consensus that drying climatic trends will decrease water resources in such regions 4. The likely resilience of groundwater in many areas revealed by improved understanding of precipitation-recharge

International audienceAmong the three sites distributed along the West African latitudinal gradient in the AMMA-CATCH observation system, the experimental setup in the Niamey area of south-west Niger samples the cultivated Sahel environment, for hydrological, vegetation and land surface processes. The objective is to investigate relationships between climate, land cover, and the water cycle, in a rapidly changing semiarid environment. This paper first presents the main characteristics of the area, where previous research, including the EPSAT and HAPEX-Sahel experiments, had evidenced a widespread decadal increase in water resources, concurrently with severe drought conditions. The specifics of AMMA-CATCH research and data acquisition at this site, over the long-term (not, vert, similar2001–2010) and enhanced (not, vert, similar2005–2008) observation periods, are introduced. Objectives and observation strategy are explained, and the main characteristics of instrument deployment are detailed. A very large number of parameters – covering rainfall, vegetation ecophysiology, phenology and production, surface fluxes of energy, water vapour and CO2, runoff and sediment, pond water, soil moisture, and groundwater – were monitored at local to meso scales in a nested structure of sites. The current state of knowledge is summarized, connecting processes and patterns of variation for rainfall, vegetation/land cover, and the terrestrial hydrologic cycle. The central role of land use and of its spectacular change in recent decades is highlighted. This paper provides substantial background information that sets the context for papers relating to the south-west Niger site in this AMMA-CATCH special issue

No abstract

No abstract

Abstract. 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, with associated 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.Published by Copernicus Publications on behalf of the European Geosciences Union.