Assessment of Spatio-Temporal Dynamics of Drought Stress Anomalies Using Hyperspectral Imagery Fusion

Abstract. Drought stress generates considerable ecological and physiological effects, endangering vegetation health, forest resistance, and the management of water resources. In this study, the combination of hyperspectral (AVIRIS Classic, 2013–2018) and multispectral (Sentinel-2, 2019–2025) imagery was utilized to evaluate the spatio-temporal dynamics of drought in California's ecologically fragile Sierra Nevada region. In order to overcome the spectral limitation of multispectral data, a solid methodology based on an Analytic Hierarchy Process (AHP) framework was established. This method algorithmically weighted and combined eight spectral indices according to their biophysical correspondence with drought, with greatest weight on Moisture Stress Index (MSI), then Normalized Difference Drought Index (NDDI) and Normalised Burn Ratio (NBR), in order to produce a composite drought severity rating. This rating was classified with a Random Forest model with high accuracy with an overall accuracy of 86.53% and balanced accuracy of 86.37%. Our analysis between 2013 and 2025 indicates varied but strengthening trends of drought with an increased escalation in the spatial magnitude and severity of exceptional drought (D4) conditions, especially for southern areas since 2019. The 2026 forecast shows an increased deterioration in drought conditions. This paper shows how multi-sensor integration, with the help of a decision-theoretic weighing scheme is a very powerful, scalable, and transferable paradigm of reliable drought monitoring, offering valuable information to anticipatory resource management and mitigation plans in mountain-like drought-prone areas around the world.