Characterization of the hyperspectral reflectance response following hailstorm damage in winter wheat (Triticum aestivum L.) and maize (Zea mays L.)

Climate is a significant determinant of agricultural productivity. Extreme weather events such as hailstorms and strong winds cause significant losses in crops yields, which subsequently worsens farmers' financial situation. Such harms are connected to the stage of crop development when the following occurs. Hail damage to winter wheat and maize can be reduced by both direct (spike) and indirect (leaves, stem) damages. E.g., hail inhibits grain production by shredding off leaf blades, which decreases leaf area and, in turn, photosynthetic activity, causing an indirect production loss. Moreover, strong wind associated with convective storms can lodge the crops causing massive yield reduction. Assessing these damages to crops is of pivotal importance for farmers, insurance, and decision makers. Traditional damage assessment is labor intensive and spatially limited. Therefore, innovative techniques of proximal sensing approaches based on hyperspectral responses could offer new insights into plant responses to such stresses that can upscaled to remote sensing as an aiding tool for more precise and spatially detailed damage assessments. The experiment was carried out at the University of Padova's experimental farm "Lucio Toniolo" located in Legnaro, Agripolis campus (Northern Italy), during the 2022 cropping season. Two fields were grown with winter wheat (Triticum aestivum L.) and maize (Zea mays L.). Each crop was damaged with simulated hail and strong winds using specifically designed prototypes. Three replicas received hail treatments at three different damage intensities: 20%, 50%, and 80%, while wind treatments were considered as lodged or not lodged. Samples of leaves, stems, spikes (wheat) and leaves (maize) were collected after the damage interventions to assess the hyperspectral response using a spectrometer (Fieldspec 4). The findings indicated that plant organs were impacted by hail and strong winds, and that healthy vegetation could be discriminated from that that had been harmed by analysis of their reflectance spectra and mostly of their first derivatives. Damage intensity discrimination was less clear, probably due to its canopy-related effects, less visible at the leaf level.