(2025) Evaluation of evapotranspiration models integrating convolutional neural network-predicted leaf area 
for Pak Choi (Brassica campestris ssp. chinensis) in greenhouse environments Journal: Scientia Horticulturae

Journal: Scientia Horticulturae

Author: Young-Bae Choi, In-bok Lee

Abstract

This study evaluates how predicted leaf area index (LAI) affects evapotranspiration (ET) model performance and 
uncertainty in greenhouse Pak Choi cultivation. Five ET models (Penman-Monteith, Stanghellini, Fynn, Shin, and Baille) were 
compared using both measured and Convolutional Neural Network-Predicted LAI data. Greenhouse environment experiments 
from June to August 2021 provided validation data under controlled conditions. LAI was estimated using image analysis 
with high accuracy (R2  = 0.9986, RMSE = 0.0547 m2⋅m− 2). Sensitivity analysis revealed that ET models were most responsive 
to radiation and LAI variations, with lower sensitivity to air temperature and relative humidity. Among physical models, 
the Fynn model demonstrated superior perfor- mance based on ET prediction accuracy (R2 > 0.87), while the Shin model 
excelled among simplified approaches (R2  > 0.92). Uncertainty propagation analysis revealed that the Stanghellini model 
exhibited the highest sensitivity to LAI estimation errors (12.55 W⋅m− 2 error when LAI error  = 1.0 m2⋅m− 2), whereas 
the Pen- man–Monteith model showed minimal sensitivity. Model performance remained consistent when using predicted 
versus measured LAI (R2 > 0.99 for all models), indicating the robustness of image-based LAI estimation for ET modelling. 
This research provides quantitative insights into model selection and uncertainty assessment for precision irrigation management 
in protected cultivation systems, with particular applicability to leafy vegetable crops under controlled greenhouse conditions.

Keywords: Evapotranspiration modelling Leaf area index Protected cultivation Uncertainty analysis Sensitivity 
                  analysis Pak choi Greenhouse irrigation

Download Link: https://www.sciencedirect.com/science/article/pii/S0304423825003887