Upscaling drought resilience by coupling soil data and UAV-multispectral imageries

Monitoring crop responses to drought is crucial for understanding the progressive impact of drought on food production and identifying management practices that can enhance agricultural resilience. This study combined drone-based multispectral data (MDd) with laboratory determination over multiple pilot farms to identify the main soil physical and chemical parameters correlated with a crop health index (SVI- Standardized Vegetation Index), which compares the Normalized Difference Vegetation Index (NDVI) at the observed time to historical (NDVI at similar dates in previous years) values. Significant relationships were found between MDd and selected soil properties for different crops. Differences found at the plot scale were primarily related to texture, organic carbon and total nitrogen content, resulting in heterogeneous responses to droughts. The performance of the proposed indicators was further validated for the same crops by extending the findings across similar climatic regions in Europe, using satellite-based multispectral data (MDs) and field-based soil data from LUCAS (Land Use/Cover Area frame statistical Survey Soil) for 2018, as well as MDs and digital soil data from SoilGrids 2.0 for 2022. Varying drought magnitudes were also considered. The method effectively identified drought-prone areas and distinguished crop health across varying drought intensities, making it a valuable tool for drought monitoring and agricultural planning.