Lead Researcher: Pasha Fayzul
Industry Partners: WiseConn, PowWow, AgH2O, Dynamax,
Jain Irrigation, and Irrometer 

Numerical techniques offer various tools that can be used to observe the relationships between crop water requirements and the variables that contribute to the plant intakes. These variables range from meteorological to field level physical conditions. Field conditions, specifically soil type and soil moisture level in the root zone, are important to consider for quantifying crop water requirements. If total soil moisture (TSM) curve over a period of time can be constructed, the curve can be used to calculate the rate of change of moisture level in the root zone. This rate of change of soil moisture in the root zone offers the basis to calculate the plant water intake (PWI) if the water percolation rate is known or estimated. The calculated PWI can then be correlated with the meteorological variables including evapotranspiration (ET), air and soil temperatures, solar radiation, wind speed, and relative humidity to observe the relationships. Multilinear (MLR) and nonlinear (NLR) regression models have been developed in this study to observe the correlation between PWI and the weather variables.

A brute force method was applied to identify the number of combinations to observe the lag and cumulative effects of weather variabilities. A total of 1,211,040 combinations for MLR and NLR models were run to observe different impacts on PWI. Results show that some of the weather variables are highly correlated with the PWI at lower lag days. As the lag day increases, the correlation decreases; and after a certain number of lag days, the correlation does not change. Results also show that the correlation increases with the increase of number of independent variables in a combination to predict the PWI. The approach was applied to an olive field at the University Agricultural Lab (UAL) located at California State University, Fresno, identifying the significant water savings potential with this method.

Project Background:
Relying on evapotranspiration to determine the amount of irrigation to apply is a common practice but can result in over or under irrigating. Identifying which other parameters are most important for contribute to estimating crop water requirement is useful to better predict crop water requirements and apply water more accurately in real time. This team used AI to developed a model using AI to successfully capture and better understand the general pattern and dynamics of the crop water requirement.