Detection of historical landscape changes in Lake Victoria Basin, Kenya, using remote sensing multi-spectral indices

Abstract

The Kenyan Lake Victoria basin has been experiencing land degradation and overexploitation as seen in various landscape changes. Quantifying and understanding the nature of these changes is essential for management decision-making. This study sought to apply the Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), and Normalized Difference Built-up Index (NDBI) - Enhanced Normalized Difference Built-up Index (NDBI-ENDBI) indices to assess the changes in vegetation, water bodies, and build-up area, respectively, in the basin between 1978 and 2018. Landsat imageries for 1978, 1988, 1998, 2008, and 2018 with thematic-oriented bands were acquired, pre-processed, and subjected to spectral-based classification techniques in ArcGIS software to compute and generate NDVI, NDWI, and NDBI-ENDBI maps. According to the results, NDVI indicated an intense decreasing trend in the density of vegetated landscapes while NDVI, NDWI, and NDBI all depicted fluctuations in moisture content (in Lake Victoria and its wetlands). The NDBI-ENDBI showed an intense increasing trend in the built-up area expansion. Based on these changes, the study concluded that urban development has been rapidly progressing within the Lake Victoria basin over the years. Intensive and strategic planning that strikes a balance between economic development and environmental conservation is recommended to realize the sustainability of this important watershed.

Keywords

• Lake Victoria basin
• Remote sensing
• Change Detection
• Multi-spectral indices