Field Crop Irrigation - Multi-Objective Optimization and Sensitivity to Weather Forecast Accuracy

Accurate irrigation and fertigation of field crops are crucial for maximizing crop yield while avoiding overuse of water and fertilizer. Weather forecasts can predict potential evapotranspiration (ET₀)but are still far from perfect. We used a case study of sprinkler irrigated spring potatoes in Coastal Israel as a test case in order to define a minimal accuracy level of ET₀ predictions for irrigation planning. The working stages of simulation-optimization-sensitivity analysis are described in the workflow. By modeling crop irrigation based on varying forecasted ET₀ relative bias ranges as well as crop and soil parameters we were able to rank the parameters by contribution to crop-model output variance. Our main findings are (Fig. 4): ET₀ prediction accuracy dominates the crop model parameters when ET₀ relative bias (δ_ET0) range <5%. The soil_n parameter dominates model output when δ_ET0 range <5% for all objective functions but transpiration (RMSE_Ta). The max. root depth is dominating transpiration output (RMSE_Ta) when δ_ET0 range <5%. This procedure of optimization and sensitivity analysis can be extended to a wide range of case studies and help define what is an adequate weather forecast accuracy suitable to base crop irrigation upon.