Trend analysis and ARIMA modelling of recent groundwater levels in the White Volta River basin of Ghana

Abstract

Groundwater based irrigation in sub-Saharan African is believed to hold the key for economic growth and poverty reduction. However, the associated consequences of the groundwater levels decline are not clearly understood due to lack of dedicated monitoring and scientific studies. In this study, we present, the first direct attempt to use non parametric tool to study trends in rainfall, temperature and groundwater levels from 2005 to 2014 using Mann-Kendall test, Sen's slope estimator and ARIMA models in the Upper East Region of Ghana. The results showed a seasonal variation with no trends in rainfall and temperature. The mean groundwater levels (depth below the ground surface) ranged from 1.24 to 5.46m with Gowrie, Bawku and Kabingo showing a narrow variation in groundwater levels as compared to Datoku and Bongo. On the seasonal scale, both the Mann-Kendall and Sen's slope estimator have shown rising seasonal trends in all the wells except Kabingo which showed a declining seasonal trend of 0.312–0.097m/year. However, on the annual scale, neither significant positive nor negative trends were observed in the groundwater levels in all the monitoring wells. Forecast for rainfall and groundwater levels using ARIMA models indicates that there will be a significant decline in rainfall at a rate of 4.779mm/year by 2020. The groundwater levels, however, showed a relatively stable trend to slight increase in most of the wells varying in magnitude across the study area and ranging from 0.000 to 0.006m/yr. The forecasted result showed an average rate of decline of 1.008m/yr at Kabingo where the groundwater level are expected to decline to about 12m by 2020. Implications for groundwater utilization in the area and future research needs have been highlighted and discussed.