#桃园

Abstract Purpose Capturing rapid changes in water status is key to optimizing deficit irrigation in Mediterranean orchards, but thermal remote sensing is constrained by the availability of high-spatial-resolution data. This study assesses the ability of visible, near-infrared, and shortwave infrared (VNIR/SWIR) indices to detect short-term water stress in peach orchards. Methods An experiment was conducted in two commercial orchards in south-eastern Spain, where mild water stress was induced by withholding irrigation for four days. High-resolution hyperspectral and thermal imagery were acquired concurrently with stem water potential measurements (ψ stem ). Structural, pigment-related, and water-sensitive indices were evaluated at several spatial resolutions, from high (20–50 cm) to medium (30 m), to analyze the effects of pixel size on stress detection. The Crop Water Stress Index (CWSI), derived from thermal imagery, served as a reference indicator. Results Those optical indices based on SWIR reflectance at 1240 nm, the Normalized Difference Water Index (NDWI₁₂₄₀) and the Simple Ratio Water Index (SRWI), showed the strongest sensitivity to ψ stem variability (R 2 = 0.63, p < 0.001), comparable to the thermal CWSI (R 2 = 0.67, p < 0.001). The analyzed SWIR indices proved more robust under spatial aggregation than other classical VNIR indices, preserving sensitivity when soil/shadow effects were included. Radiative transfer simulations (PROSPECT/3D FLIGHT model) confirmed the NDWI₁₂₄₀ response to variations in leaf water content and leaf angle under short-term stress. Conclusion The results demonstrate the potential of SWIR-based indices, especially in the 1200–1300 nm range, to detect early water stress in Mediterranean woody crops and support their applicability for large-scale monitoring with current and upcoming satellite sensors, though further validation using operational satellite data is required.