Geospatial Evidence for Sustainable Urban Development

Zambia michael.mutale@cbu.ac.zm Rapid urbanisation in developing countries poses significant threats to land cover integrity and ecological sustainability. This study assesses Land Use and Land Cover (LULC) dynamics in Kitwe District, Zambia, over a twenty-year period (2004–2024) using multi-temporal Landsat imagery processed through Google Earth Engine (GEE). Supervised Random Forest classification was applied to five benchmark years, achieving overall accuracies of 86.8%–93.2% (Kappa: 0.80–0.91). Results reveal a 60.8% decline in dense vegetation (36,380 ha to 14,261 ha) and a 174.4% expansion of built-up areas (5,746 ha to 15,768 ha), driven by urban growth and artisanal mining encroachment. Post-classification change detection in ArcGIS Pro identified woodland and dense vegetation cover fragmentation and encroachment into protected areas as dominant processes. These findings provide quantitative geospatial evidence to guide urban planners and policymakers in sustainable land management in Kitwe and comparable mining towns in sub-Saharan Africa. Assessment of energy-efficient features for building designs in the