Quantifying Uncertainties in Modeling Climate Change Impacts on Hydropower Production

Abstract

Climate change will have large impacts on water resources and its predictions are fraught with uncertainties in West Africa. With the current global drive for renewable energy due to climate change, there is a need for understanding the effects of hydro-climatic changes on water resources and hydropower generation. A hydrological model was used to model runoff inflow into the largest hydroelectric dam (Kainji) in the Niger Basin (West Africa) under present and future conditions. Inflow to the reservoir was simulated using hydro-climatic data from a set of dynamically downscaled 8 global climate models (GCM) with two emission scenarios from the CORDEX-Africa regional downscaling experiment, driven with CMIP5 data. Observed records of the Kainji Lake were used to develop a hydroelectricity production model to simulate future energy production for the reservoir. Results indicate an increase in inflow into the reservoir and concurrent increases in hydropower production for the majority of the GCM data under the two scenarios. This analysis helps planning hydropower schemes for sustainable hydropower production.

Influence of Parameter Sensitivity and Uncertainty on Projected Runoff in the Upper Niger Basin under a Changing Climate

“…This is in agreement with the study of Oyerinde et al [9] who disclosed similar high efficient runoff simulation with a similar version of the IHACRES -CMD model in the Niger basin. Clear differences in monthly simulation patterns in the different RR models were due to contrasting model structures (structural uncertainties).…”

“…Future PET was computed from extracted temperature with the Hamon's model. In line with previous studies [9,35,43], rainfall and temperature projections were bias corrected with quantile mapping [44] at monthly time step. Similar to other studies [9,35,45], future annual runoff was aggregated into two future time periods ("near future" (2010-2035) and "far future" (2036-2099)) and were compared to the historical period .…”

“…GPCP is available from 1997-2016 while MERRA is available from 1979-2010. PET was computed from MERRA temperature using the Hamon's model that was earlier reported to provide acceptable estimations of PET [9,34,35]. Catchment boundary of the Niger basin was obtained from Hydrosheds [36].…”

“…The same scenarios are therefore also the highest priority of the CORDEX simulations [40]. CORDEX data have been evaluated and used for hydrological studies in the region [9,35,41,42]. In this study, basin CORDEX projection data were extracted as described for the observations data.…”

“…More complex, physically-based, distributed models often require input data such as soil and land use data that are either missing or weakly reliable in West Africa [9]. However, there is uncertainty in the results of any modelling, of different types and from different sources [10].…”