Modeling spatiotemporal land use/land cover dynamics by coupling multilayer perceptron neural network and cellular automata markov chain algorithms in the Wabe river catchment, Omo Gibe River Basin, Ethiopia

Abstract

Land Use/Land Cover (LULC) change has been a substantial environmental concern, hindering sustainable development over the past few decades. To that end, comprehending the past and future patterns of LULC change is vital for conserving and sustainably managing land resources. This study aimed to analyze the spatiotemporal patterns of landscape dynamics from 1986 to 2022 and predict situations for 2041 and 2058, considering a business-as-usual (BAU) scenario in the Wabe River Catchment. The historical land use image classification employed a supervised technique using maximum likelihood algorithms in ERDAS Imagine, and identified six major land cover classes. For future projections of LULC changes in 2041 and 2058, multilayer perceptron neural network and cellular automata-Markov chain algorithms were utilized, incorporating various driving factors and independent spatial datasets. The findings revealed significant and ongoing LULC dynamics in the catchment, with persistent trends expected. Notably, woodland, built-up areas, and agriculture experienced substantial net increases by 0.24%, 1.96%, and 17.22% respectively, while grassland, forest, and agroforestry land faced notable decreases of 4.65%, 3.58%, and 11.20% respectively from 1986 to 2022. If the current rate of change continues, built-up and agricultural lands will expand by 1.28% and 5.07%, while forest and agroforestry land will decline by 2.69% and 3.63% respectively by 2058. However, woodland and grassland cover will exhibit divergent patterns, with a projected decrease of 0.57% in woodland and an anticipated increase of 0.54% in grassland cover. Overall, the observed changes indicated a shift towards intensive agriculture, built-up area expansion, and potentially adverse environmental consequences such as soil degradation, biodiversity loss, and ecosystem decline. To mitigate these consequences and promote sustainable development, immediate action is necessary, including environmentally friendly conservation approaches, sustainable land management practices, habitat protection, and reforestation efforts, ensuring the long-term resilience and viability of the catchment’s ecosystems.

Keywords

• land use/land cover
• dynamics
• multilayer perceptron
• cellular automata
• markov chain
• Wabe River catchment