S12 - Session O2 - Spatial decision support systems for precision horticultural water management

Authors: Alon Ben-Gal *, Aviva Peeters, Noa Ohana-Levi, Idan Bahat, Livia Katz, Hagai Yasuor, Alex Barski, Osher Burkis, Yafit Cohen

Advances in variable rate drip irrigation technology combined with large volume data acquisition over both space and time offer unique opportunities for precision water management of horticultural crops. Three major opportunities in the realm of precision, spatially relevant, decision-making involve 1) predicting variables of interest for management, 2) delineation of management zones, and 3) optimization of sampling/sensing strategies. We investigate methods for decision support in VRI systems in drip irrigated orchards and vineyards. We are developing agricultural spatial decision support systems (AgSDSS) to a) provide applicable, user-friendly spatio-temporal models and tools to predict variables needed for precision water management decision making and b) offer a decision support system for distributing sample locations/sensors that most accurately represent a field spatially. The AgSDSSs consider all available data for a given field. In our cases this includes ECa maps, elevation based maps and indices, historical yield maps, historical and current maps using spectral and thermal imaging and their related plant and plant-water-status indices. These are evaluated using multivariate spatial statistical analysis including machine learning to account for spatial location and spatial processes. Model output includes recognition and weighting of response variables that can serve as indicators of pre-determined dependent variables useful for management decision-making, delineation of fields into management zones using the most influential and accessible data, and determination of optimized number and location for