Evaluating the potential of high-resolution hyperspectral UAV imagery for grapevine viral disease detection in Australian vineyards

•Developed a rapid and cost-effective method for detecting grapevine viral diseases.•Achieved prediction accuracies of up to 98%.•Unique spectral regions and detection times were identified.•Demonstrated robustness of prediction model at various locations and seasons. Grapevine (Vitis spp.) viral diseases cause substantial productivity and economic losses to the viticulture industry. Existing disease detection methods are both costly and labour-intensive, prompting a need within the industry for rapid and cost-effective detection methods. The present study evaluates the feasibility of unmanned aerial vehicle (UAV)-based hyperspectral sensing in the visible and near-infrared (VNIR) spectral bands to detect two economically significant viral diseases – Grapevine Leafroll Disease (GLD) and Shiraz Disease (SD) – in four popular winegrape cultivars in Australian vineyards. The partial least squares discriminant analysis (PLS-DA) and Receiver Operating Characteristics Curve (ROC) were used to discriminate diseased and healthy pixels and predict the disease for individual vines. The model predictions for red- and white-berried grapevine cultivars achieved an accuracy of 98% and 75%, respectively. For each viral disease, unique spectral regions and optimal detection times during the growing season were identified. Our work demonstrates the value of high-resolution hyperspectral remote sensing for the detection of viral disease symptoms in vineyards.