Mapping of wheat lodging susceptibility with synthetic aperture radar data

Crop lodging reduces yield quantity and grain quality of cereal crops. Understanding seasonal variation in crop lodging susceptibility enables lodging risk assessments and predictions of associated crop yield losses. We demonstrate a novel remote sensing-based approach, using sparse field observations and widely available synthetic aperture radar (SAR) satellite imagery, to map a safety factor against root lodging (SFA) in wheat. SFA quantifies the ability of the rooting system to support the self-weight moment of the whole plant and can be used as an indicator of in-season root lodging susceptibility. SAR satellite images, from Sentinel-1 and RADARSAT-2, were acquired synchronously with field measurements in Jolanda di Savoia, Ferrara, Italy during the 2018 winter wheat growing season. The field data included measurements from non-lodged (healthy) wheat such as plant height, height at the centre of gravity, self-weight moment of the whole plant, soil anchorage strength, SFA and those from lodged wheat such as crop angle of inclination, lodged area and the point of plant failure (stem or root). Field measurements confirmed that SFA decreased progressively through the season and was consistent with the observed lodging. Strong and significant correlations through the season were observed between SFA and SAR satellite image metrics. The validated regression models showed a strong relationship between field-measured SFA and the metrics from RADARSAT-2 (R2CV = 0.84, RMSECV = 0.54) and Sentinel-1 data (R2CV = 0.73, RMSECV = 0.59). However, it is possible that the relationship between SFA and SAR parameters is confounded by the variations in crop biomass, which is highly correlated with SFA (r = −0.71, p