TITLE:
Land Cover Change Assessment, Prediction Modelling, and Ecosystem Service Implications in Lubigi Wetland, Kampala, Uganda
AUTHORS:
Rose Malot Waswa, Paul Thomas Obade, Steven Njuguna
KEYWORDS:
Land Change Modeler, Land Use Land Cover, Markov Chain, Multi-Layer Perceptron, Wetland degradation, Remote sensing, Lubigi Wetland, Kampala, Uganda
JOURNAL NAME:
Journal of Geographic Information System,
Vol.17 No.5,
September
30,
2025
ABSTRACT: Urbanization and agricultural expansion often result in the degradation of natural wetlands, leading to significant ecological impacts. This study explored the dynamics of land cover changes and their implications on ecosystem services in Lubigi Wetland, Kampala, Uganda. The objectives were to assess the dynamics of land cover changes between 1999 and 2022, predict future land cover scenarios for 2030, and identify the ecosystem services under threat due to wetland degradation. Sentinel-2 and Landsat imagery were processed using ArcGIS and the Land Change Modeler (LCM) to assess land cover transitions. Multi-Layer Perceptron (MLP) neural network and Markov Chain analysis were employed to model future land cover scenarios. Land cover transitions were quantitatively calculated in terms of gains, losses, and net change. Regression analysis was used to identify key drivers of wetland degradation. The ecosystem services evaluated were categorized into regulating, provisioning, cultural, and supporting services. The findings reveal significant land cover changes in Lubigi Wetland from 1999 to 2022, with built-up areas increasing by 1140.60% from 10.74 km2 in 1999 to 122.5 km2 in 2022, while wetland areas decreased by 49.5%, from 453.54 km2 to 228.96 km2. By 2030, built-up areas are predicted to expand further to 279 km2, and wetland areas are projected to shrink to 200 km2. The loss of wetland area has had a profound impact on ecosystem services, particularly in flood regulation, water quality, and biodiversity support, all of which have been significantly degraded due to increasing anthropogenic pressures. The results suggest that population growth and urban expansion are major drivers of wetland degradation, and urgent conservation measures are necessary to mitigate further damage. The MLP neural network demonstrated an accuracy rate of 87.64%, supporting its use in land cover change modelling. The study concludes that without intervention, the continued expansion of anthropogenic land cover will exacerbate the loss of critical ecosystem services, further threatening the ecological stability of Lubigi Wetland.