Water resources modeling in a context of climate change in Cote d'Ivoire

Abstract

During the period 1961-2000, Côte d’Ivoire, has been challenged by irregular rainfall distribution leading to frequent extreme events (drought and floods), resulting in agriculture losses. This study is sought to analyze spatial and temporal variability of rainfall time series by using a statistical model based on R software and their packages and its impact in rainfall/runoff modeling. The methodological approach consisted to analyze the annual rainfall variability from the twenty-seven meteorological stations by the use of statistical test (Mann Kendall and Pettitt), to characterize the daily rainfall extreme events and the meteorological drought by the use of climate’s indices based on ETCCDI and the Standardized Precipitation Index (SPI) with RClimDex and RDIT software’s. Indeed, in order to know the best suitable period for the calibration and validation, to analyze the transferability of the parameters and to prevent natural hazards, a hydrological model GR2M at monthly scale (Mouelhi 2006 version) was performed at five stations (Semien, Soubré, Mbasso, Nzienoa and Bada).

It was found that the North and South portion of Côte d’Ivoire present a significant decreasing trend. Pettitt test showed a break in the time series around 1969-1977 and 1977-1982. The result of SPI-6 and SPI-12 indicated that most of frequent drought in term of severity, duration and frequency was observed after 1980. An alteration of dry and wet conditions was showed after1990. Based on the results of the climates indices, the annual total precipitation (PRCPTOT) is highly decreasing while the Consecutive Wet Days (CWD) and the monthly maximum 1-day precipitation (R*1day) showed an increasing trend. The results obtain shows that the GR2M performs well in Semien as the average Nash criterion is greater than 0.75.