Variable Rate Irrigation Scheduling Based on High-Resolution Short-Wave Infrared Sensing and Internet-of-Things. AVF Project #2023-2718.
Summary of Final Report:
Amidst escalating drought conditions in the San Joaquin Valley, grapevine growers face an urgent need for effective water management strategies. This research contributes valuable insights into sustainable viticulture practices, particularly for optimizing water use in vineyards while addressing the physiological responses of grapevines to varying irrigation intensities and timings.
The study was conducted in a commercial Vitis vinifera L. cv. Cabernet-Sauvignon vineyard in California’s West San Joaquin Valley. The vineyard was equipped with a variable rate automated irrigation system administering 12 distinct irrigation regimes across four randomized replicates, creating 48 experimental zones. Over the 2022 and 2023 seasons, irrigation schedules included sustained deficit irrigation (SDI) at 40%, 60%, 80%, and 100% of ETc and regulated deficit irrigation (RDI) applied at varying intensities pre- and post-veraison (e.g., 100/40, 80/60, 60/100). Water use was meticulously tracked with flow meters to ensure precision.
Remote sensing technology was integrated into the study to assess plant water status and model physiological responses. A hyperspectral NIR/SWIR camera (900–1700 nm) mounted on a UAV captured spectral data throughout the growing season, focusing on pure grapevine pixels. The sensor’s high water absorption in the NIR/SWIR range facilitated the estimation of stem water potential (Ψstem), a key indicator of water stress, using machine learning models. Hyperspectral images were processed into bottom-of-atmosphere reflectance via standard ground panels and the Quick Atmospheric Correction Method (QUAC). The best regression models achieved an R² of 0.54 and RMSE of 0.11 MPa, while the top classification model reached 74% accuracy, demonstrating the potential of remote sensing for spatial and temporal water stress mapping.
Field measurements included Ψstem and gas exchange from June to harvest, grape composition from veraison to harvest, and yield components at harvest. Results revealed that while 40% SDI significantly underperformed compared to 100% ETc, some RDI strategies and SDI regimes with moderate reductions in ETc delivered comparable or superior outcomes while conserving water.
This research advances irrigation management by integrating remote sensing and field measurements, offering a nuanced understanding of how water regimes impact plant water status, gas exchange, berry composition, and yield. These findings provide practical, data-driven solutions for grapevine growers navigating water scarcity, enabling more resilient and sustainable vineyard practices in the face of climate change.
