Improving hydrological modelling across West Africa with the HYPE model

Floods pose an increasing challenge for societies in West Africa; causing loss of lives, damaged infrastructure, and food insecurity. Improving flood management is hence an important development priority for the region, which several initiatives aim to contribute to. Hydrological forecasting systems can help, in which a core component is hydrological models. This research focus on the refinement of the hydrological model HYPE across West Africa within FANFAR (www.fanfar.eu).

The FANFAR domain stretches across West Africa (from Cap Verde to Chad), and hence the research focussed spatial hydrology, specifically finding suitable representations of runoff and key hydrological features across the domain. Our starting point was the World-Wide HYPE model (Arheimer et al. 2020). We analysed its performance in the region and found that runoff response was generally too fast, and not sufficiently differentiated between sub-regions (median daily KGE was -0.09 for 151 gauges). We setup an experiment to try four different approaches to regionalize runoff-regulating processes, namely Hydrobelts, Köppen, Soil capacity index, and Hydroclimatic regions. We divided representation of runoff-generating processes in HYPE (infiltration excess, saturation excess, soil storage, subsurface runoff) into regions linked to each regionalization scheme. We found that the Soil capacity index by Wang-Erlandsson et al. (2016) provided the most useful basis for regionalisation, giving the best and most balanced representation of discharge across the domain. However, we also noted that the added value of regionalisation compared to domain-wide calibration was rather limited, possibly due limited data availability. Another key adaptation was to calibrate regulation parameters of dams and floodplain dynamics of the Inland Niger Delta. All in all, this resulted in a much-improved West Africa HYPE model (for 151 daily streamflow gauges the median KGE was 0.36, reaching up to 0.9 at specific locations). This model was put into an operational forecasting chain within FANFAR, where West African hydrological services and disaster managers used it to disseminate alerts and reduce flood impacts.

Here we summarize this hydrological experiment, and how spatial hydrology, big and open data, participatory research, and capacity development can help reaching development priorities.

References

Arheimer et al. 2020, https://doi.org/10.5194/hess-24-535-2020

Wang-Erlandsson, et al. 2016, https://doi.org/10.5194/hess-20-1459-2016