Shi et al. (2026) A theoretical framework for critical canopy temperature and its application to improve remote-sensing-based estimation of evapotranspiration

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Environmental Research Letters
• Year: 2026
• Date: 2026-01-01
• Authors: Zhe Shi, Wei Zhao, Yujiu Xiong, Xu Lian, Jianing Fang, Shijie Jiang, Chunhua Yan, Alexander J. Winkler, Guo Yu Qiu
• DOI: 10.1088/1748-9326/ae33d3

Research Groups

Not specified in the provided abstract.

Short Summary

This study introduces 'critical canopy temperature (Tcc)' to define the dry boundary in surface energy balance (SEB) models and develops a physics-constrained machine learning (ML) model to predict Tcc, significantly improving evapotranspiration estimation when integrated into the SEBAL model.

Objective

• To define and predict 'critical canopy temperature (Tcc)', corresponding to the dry boundary where latent heat (LE) equals 0, using a physics-constrained machine learning model.
• To integrate the predicted Tcc into the Surface Energy Balance Algorithm for Land (SEBAL) model to improve the accuracy of latent heat (LE) estimation, particularly under extreme conditions and at regional scales.

Study Configuration

• Spatial Scale: Point-scale (103 eddy-covariance stations) to regional scales (pixel-level estimation).
• Temporal Scale: Not explicitly defined, but implied continuous measurements from eddy-covariance stations are used for model training and validation.

Methodology and Data

• Models used: Physics-constrained machine learning (ML) model (hybrid model), Surface Energy Balance Algorithm for Land (SEBAL) model.
• Data sources: Meteorological measurements from 103 eddy-covariance (EC) stations, remote-sensing data.

Main Results

• The hybrid model effectively captures canopy temperature anomalies during stomatal closure and demonstrates better generalization than pure ML approaches for LE estimation, especially under extreme conditions.
• Incorporating Tcc into the SEBAL model significantly improves its performance, reducing the root mean square error (RMSE) for LE from 119.33 W m⁻² to 81.71 W m⁻² (a 31.52% reduction) against EC observations.
• The hybrid model enables pixel-level estimation of Tcc at regional scales, addressing the long-standing challenge of dry-boundary underrepresentation in SEB models.

Contributions

• Introduction of 'critical canopy temperature (Tcc)' as a novel, physically-defined parameter for the dry boundary in SEB models.
• Development of a physics-constrained machine learning model that conserves the surface energy balance equation for robust and accurate Tcc prediction.
• Significant improvement in the accuracy and generalization of remote-sensing-based evapotranspiration estimation (specifically with SEBAL) by replacing conventional dry-boundary selection schemes with Tcc.
• Provides a robust framework for predicting theoretical dry-boundary temperatures, supporting improved monitoring of vegetation physiological status and enhancing the overall accuracy of SEB models.

Funding

Not specified in the provided abstract.

Citation

@article{Shi2026theoretical,
  author = {Shi, Zhe and Zhao, Wei and Xiong, Yujiu and Lian, Xu and Fang, Jianing and Jiang, Shijie and Yan, Chunhua and Winkler, Alexander J. and Qiu, Guo Yu},
  title = {A theoretical framework for critical canopy temperature and its application to improve remote-sensing-based estimation of evapotranspiration},
  journal = {Environmental Research Letters},
  year = {2026},
  doi = {10.1088/1748-9326/ae33d3},
  url = {https://doi.org/10.1088/1748-9326/ae33d3}
}