Enhancement of drought monitoring by means of soil sampling and drone-based multispectral sensing

Monitoring crop physiological responses to drought is crucial to understand progressive impacts on food production and identify resilient and sustainable irrigation practices. Although many climate research experiments provide valuable data, long-term measurements of soil properties and plant characteristics at the field scale are not always affordable. We combined drone-based multispectral remote sensing with measurements of soil properties over multiple pilot farms, where soil sampling was performed for each plot during the drone survey. Our goal was to determine if drone-based indices capture drought stress responses of different crops (maize and sugar beet) and whether responses are affected by soil physical and chemical characteristics (e.g., texture, density, porosity, moisture, pH, electrical conductivity, organic carbon and total nitrogen contents, availability of micro and macro nutrients). 
Significant relationships were found between vegetation indices and soil features for different crop types. Differences found at the field scale were related mostly to organic carbon content and resulted in heterogeneous responses to irrigation practices. Our spatial variability analysis pointed out an overall homogeneous response for areas submitted to severe and moderate drought having similar soil properties, independently of the crop type. More investigation is needed to address the possible effect of local practices (e.g., fertilization, amendment, tillage) at the field scale. The feasibility of carrying out systematic drone flights coincidentally or close to-ground campaigns will reveal the consistency of the observed spatial patterns in the long run.