Geospatial assessment of land use and land cover change in Guinea Savannah Region, Nigeria

Abstract

Land use/land cover (LULC) changes are critical determinants of water resource sustainability and food security in savannah landscapes. Understanding LULC change dynamics is crucial for sustainable land management and policy-making. This study evaluated the historical and future LULC change dynamics in the Kampe Omi Dam Basin (KODB), Nigeria. Historical LULC were analysed for 2002, 2014, and 2024 using Landsat imagery processed on Google Earth Engine and classified with a random forest algorithm, achieving high accuracies (overall accuracy: 96.67-98.39%; kappa: 0.95-0.97). Future LULC patterns were projected for 2034, 2044, and 2054 using a multilayer perceptron and Markov chain modelling framework. Cropland expanded by approximately 320 km² between 2002 and 2024 and is projected to occupy more than 82% of the basin by 2054, primarily at the expense of vegetation and transient bare land. Transition probability analysis identified strong conversion pathways from vegetation and bare land to cropland, whereas water bodies exhibited high, long-term stability. Distance to settlements and proximity to streams appear to be the most influential drivers of land conversion, underscoring the dominance of accessibility and water availability over topographic constraints. Future projections indicate continued agricultural intensification under business-as-usual conditions, raising concerns about ecosystem resilience, soil conservation, and watershed sustainability. These findings highlight the need for integrated watershed management that protects the remaining vegetation, guides agricultural expansion, and safeguards water resources through sustainable land-use planning. This study provides spatially explicit evidence to support watershed management and climate adaptation in the KODB and comparable savannah catchments across sub-Saharan Africa.