Spatial modelling of land degradation and restoration dynamics in mining-affected landscapes in Ghana

Land degradation remains a critical environmental challenge in Ghana, particularly in landscapes dominated by artisanal and small-scale mining (ASM). This study assessed land degradation and restoration trajectories in two ecologically distinct mining districts; Prestea Huni Valley (Rainforest zone) and Talensi (Guinea Savannah zone), by operationalizing the UNCCD Land Degradation Neutrality (LDN) framework. Multi-temporal Sentinel-2 imagery (2015, 2018, 2021, and 2024) was classified using a Random Forest algorithm in Google Earth Engine, achieving classification accuracies exceeding 85%. The analysis quantified mining’s proportional contribution to District-level degradation and recovery dynamics, disaggregated by small-scale and large-scale mining (SSM and LSM). Results show that mining’s share of total degraded land increased from 8.8% to 19.1% in Prestea and reached 9.3% in Talensi by 2024. SSM accounted for over 90% of mining-driven degradation in the forest zone, while LSM became increasingly influential in the savannah landscape during the final period (2021–2024). Across both districts, recovery from mined lands remained marginal (<7% in Prestea; <3% in Talensi), indicating that restoration gains lag behind cumulative losses. Importantly, the study reveals that ecological context and mining typology jointly shape degradation–recovery balances, with forest systems exhibiting higher turnover but savannah systems showing greater vulnerability and limited regenerative capacity. By integrating mining-type disaggregation within the LDN framework, this study advances spatial diagnostics of degradation neutrality in extractive landscapes. The findings provide evidence for differentiated land-use policy responses that prioritize preventive stabilization in fragile savannah ecosystems and strengthened regulation of dispersed artisanal mining in forest zones.