Quantifying the magnitude and trend of land use/land cover changes in gilgel gibe catchment, southwest Ethiopia

Abstract

Analysis of land use/land cover changes is vital to hydrologists, planners and watershed management decision makers to design appropriate response strategies to tackle the undesirable effects of the changes. This study was aimed at analyzing the land use/cover changes of Gilgel Gibe Catchment in 1991 – 2021. LULC data of 1991–2021 were derived from multispectral Landsat images. Data were also gathered using field observations and key informant interview. Data of LULC classes (1991–2021) were generated using supervised classification with maximum likelihood algorithm of ENVI 5.1 and ArcGIS 10.5. Change detection analysis and accuracy assessment were done where accuracy levels all the study periods were > 85%, and the overall Kappa statistics of the periods were > 0.89. Cultivated land and built-up area of the catchment are increasing with increasing magnitude of change; whereas, while forest cover and water body of the catchment are shrinking with declining magnitudes of change, grazing land and shrubland covers are declining with increasing magnitude of change in the catchment. The net increase in degraded land is a reflection of the increasing degradation of natural resources in the catchment. Rapid population growth and the subsequent raising demand for farmland and forest and shrub (e.g. fuel-wood and construction) products, decline yield, unemployment and lack of alternative income source, and open access and limited conservation of resources are the main causes for the dramatic shrinkages of forest, grazing, water body and shrubland resources. Thus, concerned bodies should take rehabilitation measures to restore degraded lands, improve production and yield of farmland by increasing improved farm-inputs and irrigation, and create employment and alternative income sources for the youth, women and the poor so as to ensure sustainable rural livelihoods and to curb the impacts on forest, shrubland and other resources.