From Dust Mapping to Vegetation Stress: Multi-Sensor Satellite Insights from Mining Landscapes

Foliar dust (FD) deposition is a major but often overlooked environmental stressor in mining regions, affecting both vegetation health and satellite-based ecological assessments. This study explores the potential of multi-source optical satellite datasets, including Landsat-8, Landsat-9, Sentinel-2B, and PlanetScope, for estimating FD and understanding its impacts on vegetation physiological processes in a tropical coal mining region of eastern India. In situ FD observations collected from mining-affected vegetation were integrated with spectral bands and radiometric indices to develop FD estimation models. The results show clear changes in vegetation spectral behaviour under increasing dust deposition. A noticeable saturation effect in spectral reflectance was observed beyond ~50 g/m² of FD accumulation. Across all satellite sensors, near-infrared (NIR) and shortwave infrared (SWIR) bands demonstrated strong sensitivity to FD, while radiometric indices such as the Global Environmental Monitoring Index (GEMI) and Non-Linear Index (NLI) performed better for FD prediction. This study further reveals that FD significantly affects vegetation functioning. Gross Primary Productivity (GPP) decreased by approximately 2–3 gC for every additional gram of FD per square meter, while vegetation temperature increased by ~0.04 K with increasing dust deposition. Water use efficiency and evapotranspiration also show consistent declining trends under higher dust loads. In addition, FD concentration decreased with increasing distance from roads and mining areas, indicating the strong influence of mining activities and vehicular movement on dust deposition patterns. The findings highlight the usefulness of multi-sensor satellite observations for monitoring FD and demonstrate how mining-induced dust can alter vegetation functioning and ecosystem processes. This study provides new insights into the hidden ecological impacts of dust pollution and their implications for remote sensing-based vegetation analysis in mining landscapes.