Remote sensing efforts made as part of European initiatives via POnTE, XF-ACTORS and the JRC, as well as through regional programs, have focused, among others, on the development of algorithms for the early detection of Xylella fastidiosa (Xf)-induced symptoms. Airborne campaigns carried out between 2016 and 2019 collected high-resolution hyperspectral and thermal images from infected areas in the Apulia region (Italy), in the province of Alicante and on the island of Mallorca (Spain). The

more »

... e sensing imagery collections were performed alongside field surveys and laboratory analyses to assess the presence of Xf, and the severity and incidence of disease in olive and almond trees. Radiative transfer models and machine learning algorithms were used to quantify spectral plant traits for each individual infected tree, assessing their importance as pre visual indicators of Xf-induced stress. These studies conducted across species have demonstrated that specific spectral plant traits successfully revealed Xf induced symptoms at early stages, i.e., before visual symptoms appear. The results show that spectral plant traits contribute differently to symptom detection across host species (olive vs. almond), and that abiotic-induced stress affects the performance of the algorithms used for detecting infected trees. Together, the different European initiatives studying the use of remote sensing to support the monitoring of landscapes for Xylella fastidiosa detection lead us to conclude that the early detection of Xf-induced symptoms is feasible when high-resolution hyperspectral imagery and physically-based plant trait retrievals are used, obtaining accuracies exceeding 92% (kappa>0.8). These results are essential to enable the implementation of effective control and management of plant diseases using airborne- droneand satellite-based remote sensing technologies. Moreover, these large-scale hyperspectral and thermal imaging methods greatly contribute to the future operational monitoring of infected areas at large scal [...]