Utilizing digital image processing and two‑source energy balance model for the estimation of evapotranspiration of dry edible beans in western Nebraska

Authors

Wei‑zhen Liang, University of Nebraska-LincolnFollow
Isabella Possignolo, University of Nebraska-Lincoln
Xin Qiao, University of Nebraska-LincolnFollow
Kendall DeJonge, Water Management and Systems Research Unit, USDA-ARSFollow
Suat Irkmak, University of Nebraska-LincolnFollow
Derek M. Heeren, University of Nebraska-LincolnFollow
Daran Rudnick, University of Nebraska-LincolnFollow

ORCID IDs

Xin Qiao http://orcid.org/0000-0003-3163-6231

Date of this Version

3-18-2021

Citation

Irrigation Science, 2021. https://doi.org/10.1007/s00271-021-00721-7

Comments

U.S. government work

Abstract

Having an accurate yet simple method to estimate crop evapotranspiration ( ET_C) is a vital component of reliable irrigation scheduling. In this study, two versions of the two-source energy balance (TSEB) model: the TSEB model with the Priestley–Taylor equation (TSEB-PT) and the Penman–Monteith equation (TSEB-PM), were used to estimate ET_C of dry edible beans in western Nebraska. Compared with previous studies, this study is unique in that a Visual Basic software—Crop Canopy Image Analyzer (CCIA) was developed to process digitally captured RGB canopy images to obtain necessary canopy cover (CC) parameters for the TSEB models such as CC percentage and leaf shape factor (leaf area divided by its perimeter). Software-estimated CC percentage was closely correlated with commercial sensor-derived CC percentage with an R² of 0.96. Additionally, estimated leaf shape factor was closely correlated with measured leaf shape factor with R² of 0.99. Both TSEB-PT and TSEB-PM models estimated ET_C well for fully irrigated dry edible beans with a root-mean-square error (RMSE) that ranged from 0.95 to 1.63 mm day⁻¹ in 2018, and 0.75 to 1.35 mm day⁻¹ in 2019, as compared to ET_C estimated from FAO56. Furthermore, ET_C from TSEB-PT and TSEB-PM were compared with a soil water balance-derived ET_C and the RMSE ranged from 2.03 to 9.65 mm in an approximate 1-week period under four irrigation treatments ranging from dry land to fully irrigated. The proposed methods in this study, by integrating digital image processing with TSEB models, have great potential to be automated and used in field-scale operations for various irrigation management scenarios of many crops.