The current study was carried out on a small rural catchment in South Africa (263 km2), where according to the locals, a change in hydrological regime has occurred during the past decade. A small wetland area occupies the bottom of the catchment, and this wetland has been drastically reduced by small scale farming activities. The purpose of the study was twofold; on the one hand to determine if any hydrological change occurred, and on the other hand to relate it to the land use change, which is considered the cause of the change by conservationist groups.
A comparative analysis of the land use of the catchment during the past decade was carried out using satellite images. The hydrological behavior of the basin during the last two decades was then studied using the conceptual rainfall-runoff model GR4J. Indeed the available climate and hydrological data was very poor and only a scarce-parameters lumped model could be used.
According to Nash coefficient of efficiency, the model was able to simulate the observed runoff of the catchment with acceptable accuracy. The method showed that under similar rainfall conditions, the basin was generating more runoff in recent years than in the early 1990s. This observation matched an observed reduction in the wetland area. To check if this change in land use could be the cause of the hydrological change, rapid estimations were made of the water transfer processes in the different land use units. It appears that, in particular in the wetlands area, the processes related to land use change only account for a small proportion of the observed increase in runoff at the catchment scale.
Under conditions of inaccurate hydrometeorological information, the approach used is appropriate to study the changes in the water balance at the catchment scale. We suggest that future applications of this method should in addition include an uncertainty assessment. Moreover it may be relevant to analyse if this method performs better than other parsimonious empirical models.