From mineral exploration to mine closure: a systematic review of multispectral and hyperspectral remote sensing for soil monitoring in mining landscapes

Soils in mining landscapes are key indicators of environmental degradation and recovery, yet their monitoring remains challenging due to spatial heterogeneity and complexity. Remote sensing is advancing in capturing hundreds of contiguous bands, particularly across the visible and shortwave infrared (400–2500 nm). This review systematically analyses 101 peer-reviewed studies applying multispectral and hyperspectral remote sensing for soil assessment across the whole mining lifecycle. Imaging spectroscopy is applied to estimate soil mineral composition, acidity and contamination by potentially toxic elements, particularly when calibrated and validated with field and laboratory data. However, these properties are often not detected directly but inferred from their association with diagnostic features of spectrally active constituents. Hyperspectral sensors provide detailed spectral products, while multispectral missions supply essential temporal depth for disturbance-recovery analyses. Despite major progress, the literature remains biased: concentrated in China, dominated by coal mines, mostly focused on operational and post-mining environmental assessments. Early baseline and multi-temporal hyperspectral analyses are almost absent. Furthermore, time series analysis with currently operational hyperspectral satellite missions (PRISMA, EnMAP) is yet unexploited. Challenges include inconsistent data preprocessing and limited model transferability across sites and sensors. Operational progress will depend on integrated multi-scale approaches and stronger collaboration between research and the mining industry to turn existing case studies into systematic monitoring practices.