Multi-Sensor Satellite Indicators for Early Detection of Agricultural Drought Using Cross-Regional Validation

Abstract

Agricultural drought poses a critical threat to global food security, especially in regions where climate variability has intensified in recent decades. The emergence of multi-sensor satellite systems has provided unprecedented opportunities for detecting drought onset earlier and with greater accuracy. This study develops an integrated framework that combines vegetation indices, land surface temperature products, evapotranspiration estimates, and soil moisture measurements derived from multiple satellite platforms. Cross-regional validation is conducted across heterogeneous agricultural landscapes to assess the robustness of the indicators and to ensure that the model performs consistently across diverse climatic and land-use conditions. The research utilizes long-term satellite archives and ground-based observations to construct a harmonized dataset for evaluating temporal fluctuations in drought severity. The analysis demonstrates that multi-sensor fusion significantly enhances sensitivity to early-stage drought signals, outperforming single-sensor indicators in both spatial coherence and temporal responsiveness. The model’s validity is tested through correlation analysis, error metrics, and inter-regional comparison, revealing strong agreement between satellite-derived drought indicators and ground-based measurements. The results highlight the importance of integrating datasets such as NDVI, EVI, LST, ET, and passive microwave soil moisture to improve accuracy during pre-drought transition phases. Furthermore, the cross-regional assessment reveals that although drought progression differs among climatic regions, early-warning signals can be captured reliably using unified satellite-driven indicators. This study contributes to improving early drought monitoring systems and provides a replicable methodological foundation for agricultural planning and climate-risk management.