Toward an integrated watershed zoning framework based on the spatio-temporal variability of land-cover and climate: Application in the Volta river basin

Abstract This article details a watershed regionalization approach which uses the concept of entropy in combination with the k-means clusters analysis. The regionalization approach aims to subdivide a watershed into relatively uniform zones based on the spatial variability of the local climate and land-covers. A case study is presented to illustrate the approach and outline the environmental implications of the outcomes. Especially, the study reports an application in the Volta river basin which is a transnational watershed, shared by six different countries in West Africa. Over years, the transboundary status of the Volta watershed seems to have exacerbated its environmental challenges, because the environmental policies in the six countries do not necessarily complement. Subsequently, it is desirable to envision unified scientific tools to support the management platform of the Volta basin. To date, the literature on the Volta has virtually neglected this aspect. Hence, this case study is timely as it intends to create a unified zoning system for the Volta river basin. In the study, formulations of entropy theory and k-means clustering were jointly applied to 16-years gridded time-series of monthly leaf area index, precipitation, and temperature across the Volta basin. Based on a clustering optimization criterion, a total of five zones were identified then the related land-cover and climatic patterns were comparatively analyzed. Significant environmental contrasts were diagnosed then specificities were pinpointed for each zone. A comparison of the new zones with an existing macro-scale ecoregion shows similarities which sustain the capacity of the regionalization approach to capturing meaningful biophysical signals. Hence, the zoning technique may be valued for further applications in environmental management.