Analyses of freshwater stress with a couple ground and surface water model in the Pra Basin, Ghana

Abstract

The optimal management of water resources requires that the collected hydrogeological, meteorological, and spatial data be simulated and analyzed with appropriate models. In this study, a catchment-scale distributed hydrological modeling approach is applied to simulate water stress for the years 2000 and 2050 in a data scarce Pra Basin, Ghana. The model is divided into three parts: The first computes surface and groundwater availability as well as shallow and deep groundwater residence times by using POLFLOW model; the second extends the POLFLOW model with water demand (Domestic, Industrial and Agricultural) model; and the third part involves modeling water stress indices—from the ratio of water demand to water availability—for every part of the basin. On water availability, the model estimated long-term annual Pra river discharge at the outflow point of the basin, Deboase, to be 198 m3/s as against long-term average measurement of 197 m3/s. Moreover, the relationship between simulated discharge and measured discharge at 9 substations in the basin scored Nash–Sutcliffe model efficiency coefficient of 0.98, which indicates that the model estimation is in agreement with the long-term measured discharge. The estimated total water demand significantly increases from 959,049,096 m3/year in 2000 to 3,749,559,019 m3/year in 2050 (p < 0.05). The number of districts experiencing water stress significantly increases (p = 0.00044) from 8 in 2000 to 21 out of 35 by the year 2050. This study will among other things help the stakeholders in water resources management to identify and manage water stress areas in the basin.