article · Journal of Hydrology Regional Studies
Accurate estimation of actual evapotranspiration, or crop water use, is critical for planning and designing effective irrigation schedules. However, precise prediction is difficult because of the non-linear dynamics of this process. Research focused on Chengdu, Wuhan, Chongqing, and Kunming in China evaluated five machine learning models enhanced with the RReliefF feature selection algorithm to predict monthly actual evapotranspiration across diverse agroclimatic conditions. The evaluated approaches included support vector machines, ensemble bagged and boosted trees, robust linear regression, and Matern 5/2 Gaussian process regression. Among these techniques, the Matern 5/2 Gaussian process regression model delivered the highest accuracy across both training and testing phases, achieving an R squared of 0.982 and the lowest error values. In contrast, robust linear regression performed least effectively. The findings indicate that the optimal Gaussian process model offers a dependable method for long-term evapotranspiration forecasting in these humid environments.
Understanding exact crop water use is essential for sustainable agriculture and efficient water management, particularly under variable climate conditions. By demonstrating that advanced machine learning models can accurately calculate actual evapotranspiration over long periods, this work provides agricultural planners and hydrologists with a reliable tool to improve irrigation efficiency, conserve water resources, and support food security.
The research presents an applied analytical model that could be integrated into precision agriculture platforms, irrigation management software, and hydrological planning tools. Potential users include agricultural planners, irrigation district managers, and agronomic software developers seeking to optimise water delivery. The technology appears to be applied and tested on historical regional datasets, indicating that commercial deployment would require integration into decision support software and validation across other target climates.
AI-generated from the published abstract. Always read the original work before citing.
Chengdu, Wuhan, Chongqing, and Kunming regions in China. Accurate estimation of crop water use or actual evapotranspiration (AET) remains a key obstacle in the effective design of irrigation schedules, plans, and design. This is due to the non-linear nature of this phenomenon. To address this issue and guarantee more accurate ET predictions, this study attempts the following: i) to assess the performance of five machine learning (ML) models optimized by the RReliefF algorithm in estimating actual ET values for each month in four Chinese provinces under various agroclimatic conditions; and ii) to select the optimal model based on statistical metrics while minimizing discrepancies between the estimated and actual ET values. AET was estimated using support vector machine (SVM), ensemble bagged and boosted trees, robust linear regression (RLR), and Matern 5/2 Gaussian process regression (M-GPR) models. The M-GPR model outperformed the other models and generated the best values for all statistical measures for training and testing stages: R 2 (0.979, 0.982), RMSE (5.56, 5.09), MAE (3.29,3.16). In comparison, the RLR model exhibited the lowest training and testing performances metrics. The results of this study demonstrate the capacity of the M-GPR model to accurately predict long-term AET values. This model is best suited for further research on AET prediction at the stations under investigation, which could improve irrigation and boost agricultural productivity. • Machine learning models were optimized using the RReliefF algorithm. • RReliefF was used to apply the rank importance of AET predictor technique. • M-GPR model outperformed all other models with best statistical metrics. • M-GPR model can accurately predict long-term actual evapotranspiration.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.1016/j.ejrh.2024.102043
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
