Floodplain wetland-river flow synergy in the White Volta River basin, Ghana

2007

English

The Upper East Region of Ghana is characterized by an unreliable mono-modal rainfall pattern making rain-fed agriculture a risky business. Therefore, farmers along the White Volta River cultivate on the floodplain to make use of residual moisture. This has prompted the Government of Ghana through the Ministry of Food and Agriculture to encourage dry season floodplain wetland cultivation, using pumped river water as a source of irrigation. The increasing use of floodplains for dry season cultivation has placed pressures on land allocation practices and floodplain wetlands resources. To ensure good management and sustainable water resource use, the role of floodplainwetlands in regulating stream flow needs to be recognized, understood and taken into account when modeling hydrological processes within the basin. This study examines floodplain wetland-river flow synergy within the White Volta River basin. The methods applied in this study involve generating a floodplain wetland probability map and selection of floodplain wetlands, the use of HYDRUS-1D to model floodplain hydrodynamics, the result served as an input into the MODFLOW model to simulate interaction between floodplain wetland and the White Volta River; and applying isotopic tracers δ18O and δ2H to derive a water balance. The mapping process used a combination of geographic information system, remote sensing and statistical techniques. Logistic regression, a statistical technique used within the GIS platform, identified distance, texture, log transformation of ETM- band4 and evapotranspiration as the parameters having the greatest predictive power in wetland mapping, at 95 per cent confidence level. The map generated enabled the selection of Pwalugu and Tindama wetland sites as suitable for detailed hydrological analysis. Applied to the Pwalugu floodplain wetland as a test site, the HYDRUS-1D model indicated that the infiltration contribution to sub-surface storage over 16 months was 444 mm. The period of highest contribution occurred between July and September. In addition, the estimated vertical gradient indicates a low upwelling of sub-surface water in areas close to the river. The results of the isotope analysis of δ18O and δ2H showed that the trajectory of the tropical continental and tropical maritime air masses influenced isotopic composition of the rainfall over the Pwalugu and Tindama wetlands. Using the Rayleigh equation, evaporative fractions from Pwalugu wetland and Tindama were estimated to be 53.25% and 16.79%. However, to verify surface and sub-surface water interaction for the Pwalugu wetland, within August and September, the isotope signatures showed a similar ratio and plot around the local meteoric water line indicating some form of interaction. To determine wetland-river flow interaction, HYDRUS-1D bottom flux was used as groundwater recharge, an input into PW-WIN (MODFLOW). The PW-WIN (MODFLOW) simulation showed a systematic variation in hydraulic head of the wetland to changes in rainfall pattern, the observed interaction between floodplain wetlands and the White Volta River was bidirectional in terms of horizontal direction. Sensitivity analysis was performed using the dimensionless and dimension scaled methods, and model outputs were found to be highly sensitive to the parameters such as horizontal hydraulic conductivity, specific storage and specific yield. The study assisted to understand the relationship between recurrent spatial and temporal patterns of water table response within the floodplain and their controlling factors. Additionally, the study showed that a combination of methods such as tracers and hydrological models can be successful used to understand the dynamics of floodplain wetlands in White Volta River basin.

Nyarko, B. K.