UC Davis

Almonds are a major crop in the state of California, in which 90% of all the world’s almonds are produced. Widespread drought and strict groundwater regulations pose significant challenges to growers throughout the state. Irrigation regimes based on observed crop water status can help to optimize water use efficiency, but consistent and accurate measurement of water status can prove challenging. In almonds, crop water status is best represented by midday stem water potential, which despite its accuracy is impractical for growers to measure on a regular basis. This study aimed to use machine learning models to predict stem water potential in an almond orchard based on canopy spectral reflectance values, soil moisture, and daily evapotranspiration. Both artificial neural network and random forest models were trained and were used to produce high resolution spatial maps of stem water potential covering the entire orchard area. Additionally, for each model type one model was trained to predict raw stem water potential values, while another was trained to predict baseline-adjusted values. Together, all models resulted in an average coefficient of correlation of R2=0.73 and an average root mean squared error (RMSE) of 2.5 bar. Prediction accuracy decreased significantly when models were expanded to spatial maps (R 2=0.33, RMSE=3.31 [avg]). These results indicate that both artificial neural network and random forest frameworks can be used effectively to predict and map stem water potential, but that both approaches are unable to fully account for the spatial variability observed throughout the orchard. Random forest models predicting raw stem water potential produced the most accurate maps. Overall, the most accurate maps were those produced by the random forest modeling predicting raw stem water potential values (R2=0.47, RMSE=2.71 bar).