Mapping, quantifying and predicting future urban green infrastructure dynamics: The case study of Dire Dawa, Ethiopia

Urban green infrastructure is a nature-based remedy for urban environmental, social and economic problems. This study mapped and quantified urban green infrastructure (UGI) dynamics in Dire Dawa city and projected future changes for 2055. Landsat and Sentinel-2 data, obtained from the USGS and ESA, respectively, were accessed via the Google Earth Engine platform. A random forest model was employed to classify the images, achieving kappa coefficients of 97%, 88%, 89%, and 90% for the years 1995, 2005, 2016, and 2025, respectively. To quantify and predict UGI dynamics, this study utilized the CA-Markov model and the multi-layer perceptron-artificial neural network (MLP-ANN) method. The results revealed a significant decline in UGI area: reductions of 454.96 ha (1995–2005), 884.22 ha (2005–2016), 269.40 ha (2016–2025), and a total loss of 1608.58 ha (1995–2025). Over the 30-year period (1995–2025), Dire Dawa’s UGI decreased by 23%, while built-up areas increased by 31.9%, and bare land decreased by 8.4%. Spatially, UGI fragmentation was most pronounced in the northern, southern, and central parts of the city. The findings suggest that existing urban planning, management, and land-use policies in Dire Dawa have failed to mitigate UGI loss. Therefore, analyzing UGI dynamics is crucial for informing future urban planning, management strategies, and policymaking to address fragmentation and promote sustainable urban development.