Abstract

The influence of climate change on the runoff of Niger River is investigated for the period 1055-2010 using Gaussian Low Pass Filler, Spearman rank correlation co-efficient and multiple regression analysis. Niger River runoff is characterized by marked temporal variability which must have obliterated any trend in the long-term. An increasing trend which is insignificant (0.0226) during the long-term period (1955-2010) but significant (0.6762) in the recent period (1981-2010) has been detected. There is high positive correlation between annual rainfall (0.653) annual heavy rainfall (0.658) with annual runoff while (he correlation between annual temperature and annual runoff is significant but negative (-0.325). The multiple regression analysis shows that the annual rainfall, annual heavy rainfall and annual temperature account for 54.4% of its variance and are good predictors of runoff. The runoff of Niger River has been observed to be highly sensitive to changes in rainfall as a small increase or decrease in rainfall resulted in marked changes in runoff particularly in dry climate. The projected changes in runoff in the sub-Sahara Africa are consistent with the observed increase in the Niger River runoff.

Keywords: Climate; Runoff; Regression; Rainfall; Variance