Abstract:
This study evaluates the impact of climate change on water resources in a large, semi-arid
urban watershed located in the Niamey Republic of Niger, West Africa. The watershed was modeled
using the fully integrated surface–subsurface HydroGeoSphere model at a high spatial resolution.
Historical (1980–2005) and projected (2020–2050) climate scenarios, derived from the outputs of
three regional climate models (RCMs) under the regional climate projection (RCP) 4.5 scenario, were
statistically downscaled using the multiscale quantile mapping bias correction method. Results
show that the bias correction method is optimum at daily and monthly scales, and increased RCM
resolution does not improve the performance of the model. The three RCMs predicted increases of up
to 1.6% in annual rainfall and of 1.58 C for mean annual temperatures between the historical and
projected periods. The durations of the minimum environmental flow (MEF) conditions, required to
supply drinking and agricultural water, were found to be sensitive to changes in runo resulting
from climate change. MEF occurrences and durations are likely to be greater from 2020–2030, and
then they will be reduced for the 2030–2050 statistical periods. All three RCMs consistently project a
rise in groundwater table of more than 10 m in topographically high zones, where the groundwater
table is deep, and an increase of 2 m in the shallow groundwater table.
