Modeling the Hydrological Dynamic of the Breeding Water Bodies in Barkedji’s Zone

• Published in: Journal of water resource and protection 2014-06, Vol.6 (8), p.741-755
• Main Authors: Bop, Mamadou, Amadou, Angelina, Seidou, Ousmane, Kébé, Cheikh Mouhamed Fadel, Ndione, Jacques André, Sambou, Soussou, Sanda, Ibrah Seidou
• Format: Article
• Language: English
• ISSN: 1945-3094; 1945-3108
• DOI: 10.4236/jwarp.2014.68071

Temporary water bodies' dynamics play an important role in the epidemiological chain-borne diseases such as Rift Valley fever as they are the main breeding habitats for mosquitoes. During the rainy season, hundreds of these temporary water bodies appear and grow in the Ferlo region (Senegal). The purpose of this research is to generate historical and future time series water levels and areas at three temporary ponds located in the environment and health observatory of Barkedji. A simple lumped hydrological model was developed for that purpose. It describes each pond watershed as three interconnected reservoirs: canopy, surface storage and soil storage and uses a linear relation to describe infiltration, percolation and baseflow (out of the soil reservoir). Given the depth of the water table in the region, percolation out of the soil surface is considered lost. Evapotraspiration was calculated using the Penman equation and withdraws water from the canopy and surface water reservoirs. Excess runoff from the soil storage is turned into runoff using a triangular unit hydrograph. The calibration was done using two years of hydrological and climatic data collected during the 2011 and 2012 rainy seasons. The calibration was successful and water level in the two ponds was simulated with a Root Mean Square Error (RMSE) of 11.2 to 15 cm. Because of the short duration of the observation, no validation could be done. Given the excellent agreement of the simulated and observed water levels during the calibration phase, the modeling exercise was considered to be successful. The developed models were used to generate historical time series of pond areas and correlate these to mosquitoes' infestation in the region. Future time series of pond areas were also generated using downscaled outputs of three regional climate models from the AMMA ENSEMBLES experiment. The generated pond levels and areas are being used to assess the evolution of the disease in the next 40 years.