A Hierarchical Robust Combined Index for Agricultural Drought Detection and Monitoring Using Earth Observation Big Data and Google Earth Engine: Application to a Case Study in Southern Italy

Agricultural drought poses a significant threat to crop productivity and food security. This study provides a detailed review of the approaches used to construct Combined Drought Indices (CDIs) in the existing literature and introduces the Hierarchical Robust Combined Drought Index (HRCDI) designed to improve agricultural drought assessment on a municipal scale, leveraging Earth Observation (EO) data freely available in Google Earth Engine. The HRCDI integrates meteorological and remote sensing indicators - including the Standardized Precipitation Evapotranspiration Index (SPEI), Soil Moisture (SM), Land Surface Temperature (LST) and Normalized Difference Vegetation Index (NDVI) - through a fuzzy logic-based decision framework. The methodology reflects the typical progression of drought impacts, from climatic drivers to biophysical responses, allowing for dynamic and temporally coherent detection and monitoring of drought severity. Applied to the Province of Foggia (southern Italy) from 2017 to 2022, the index provided monthly, municipality-level classifications that successfully captured the temporal evolution of drought events, particularly during the drought-affected years of 2017 and 2022, when yield losses of 5% (2017) and 22% (2022) were registered for the durum wheat, the main cultivated crop of the Province of Foggia. Key advantages of the approach include its scalability, the use of a robust z-score, and its operational relevance for national and local institutions in supporting drought compensation schemes. The results demonstrate the potential of the index as a robust and adaptable tool for early warning and monitoring of agricultural drought, and agricultural risk management under changing climate conditions.