Evaluation of Drought Indices in the Niger Basin, West Africa
Okpara J.N. and Tarhule, A.

The Niger River Basin (NRB) in West Africa is drought-prone. This study evaluates and compares the performance of three drought indices in the Upper Niger subwatershed, two of which drought managers in the region are already used to, the Standardized Rainfall Anomaly Index (SAI) and Bhalme and Mooley Drought Index (BMDI). The third one is the Standardized Precipitation Index (SPI). A time series of the three indices were derived using 52 years growing season monthly station rainfall (April-October), regionalized into areal rainfall. The calculated statistical relationships of the indices provides diagnostics for their performance evaluation based on six decision criteria, whose weightings were determined using pairwise comparison of the Analytic Hierarchy Process (AHP) approach. Two-parameter gamma distribution is the best fit and most suitable for transformation of rainfall distribution in the region. SPI requiring equiprobablity transformation of the data, satisfied the normality assumption, whereas it was violated by others. The three drought indices showed similar temporal trends in all the time scales, with the historical extreme climatic anomalies in the basin well captured. Results further showed that SPI, which is more robust and sensitive to dryness, identified 42 and 17 moderate and extreme drought events respectively, against 35 and 7 captured by the SAI and BMDI that are less robust. In this paper, we find that SPI ranked first among other meteorological drought index in the Niger River basin, having the highest priority weight of 0.6123, with the inconsistency in the pairwise comparison with the tolerable limit (i.e. CR < 0.1).

Full Text: PDF     DOI: 10.15640/jges.v3n2a1