Click on “Download PDF” for the PDF version or on the title for the HTML version.


If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

Prediction of daily reference evapotranspiration using machine learning algorithms across the central United States

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan www.asabe.org

Citation:  2021 ASABE Annual International Virtual Meeting  2100572.(doi:10.13031/aim.202100572)
Authors:   Sunil K Bhandari, Vahid Rahmani
Keywords:   extreme learning machine, irrigation, random forest, reference evapotranspiration.

Abstract. Accurate prediction of reference evapotranspiration (ET0) plays a significant role in irrigation water management and helps to increase the agricultural production. Our study evaluates the power of random forest (RF) and extreme learning machine (ELM) for predicting daily ET0. Four different input combinations were acquired using minimum and maximum temperature (Tmin and Tmax), relative humidity (Hr), net solar radiation (Rn), and wind speed (U2) from 39 weather stations from 1979-2014 for warm season (May through September) across the central United States. Results revealed that the RF model showed better performance when used with all or limited input variables of Tmax, Tmin, Rn, U2, and Hr. Coefficient of determination (R2) ranged between 0.769 and 0.997 and when limited or all parameters were used, respectively. Considering the prediction accuracy, the RF model can be recommended over ELM. These models can be utilized to predict ET0 in other parts of the world with limited temporal or spatial climate parameters.

(Download PDF)    (Export to EndNotes)