Wang et al. (2025) Multi-scale evaluation of six fused evapotranspiration products over mainland China: Accuracy, consistency and uncertainty

Identification

• Journal: Journal of Hydrology
• Year: 2025
• Date: 2025-10-08
• Authors: Dayang Wang, Dagang Wang, Shaobo Liu, Zequn Lin, Ya Huang, Xiaohang Ma, Mengjiao Wu, Ma Ying, Jialu Zhu, Bai-Lian Li
• DOI: 10.1016/j.jhydrol.2025.134371

Research Groups

• College of Water Resources and Modern Agriculture, Nanyang Normal University, Nanyang, China
• School of Geography and Planning, Sun Yat-sen University, Guangzhou, China
• School of Freshwater Science, University of Wisconsin-Milwaukee, WI, USA
• State Key Laboratory of Hydrology-Water Resources and Hydraulic Engineering, Hohai University, Nanjing, China
• International Center for Ecology and Sustainability, University of California, Riverside, CA, USA

Short Summary

This study comprehensively evaluates the accuracy, consistency, and uncertainty of six fused evapotranspiration (ET) products over mainland China from 1982 to 2017, revealing that AutoML performs best at the site scale while REA exhibits the lowest uncertainty.

Objective

• To evaluate the accuracy, consistency, and uncertainty of six fused evapotranspiration (ET) products (AutoML, CAMELE, DOLCE, GLASS, REA, and SGAE) over mainland China using site-based eddy covariance observations and terrestrial water balance estimates.

Study Configuration

• Spatial Scale: Mainland China, including six major river basins, four climate zones, and four land cover types.
• Temporal Scale: 1982 to 2017 (36 years).

Methodology and Data

• Models used: AutoML, CAMELE, DOLCE, GLASS, REA, and SGAE (six fused ET products).
• Data sources: Site-based eddy covariance (EC) observations and terrestrial water balance-based ET estimates (ETwb).

Main Results

• AutoML product demonstrated the lowest root mean square error (RMSE), highest correlation coefficient (Corr), Kling-Gupta efficiency (KGE), and index of agreement (IOA) at the site scale relative to in-situ ET observations.
• DOLCE showed the strongest correlation with terrestrial water balance-based ET estimates (ETwb).
• Multi-year average ET across mainland China ranged from 363 mm/yr to 525 mm/yr, exhibiting a southeast-to-northwest gradient.
• A significant increasing trend in ET was observed, ranging from 0.45 mm/yr² to 1.30 mm/yr².
• CAMELE and DOLCE displayed the highest spatial consistency among the products.
• CAMELE and REA showed the highest temporal consistency.
• REA exhibited the lowest uncertainty and relative uncertainty across six major river basins, four climate zones, and four land cover types.
• SGAE showed the highest uncertainty and relative uncertainty.

Contributions

• First comprehensive multi-scale evaluation of six fused ET products (AutoML, CAMELE, DOLCE, GLASS, REA, SGAE) over mainland China.
• Provides quantitative insights into the accuracy, consistency, and uncertainty of these products across various spatial and temporal scales, climate zones, and land cover types.
• Offers a valuable reference for the selection and application of fused ET products to support sustainable water resource management.

Funding

• Not specified in the provided text.

Citation

@article{Wang2025Multiscale,
  author = {Wang, Dayang and Wang, Dayang and Wang, Dagang and Wang, Dagang and Liu, Shaobo and Lin, Zequn and Huang, Ya and Ma, Xiaohang and Wu, Mengjiao and Ying, Ma and Zhu, Jialu and Li, Bai-Lian},
  title = {Multi-scale evaluation of six fused evapotranspiration products over mainland China: Accuracy, consistency and uncertainty},
  journal = {Journal of Hydrology},
  year = {2025},
  doi = {10.1016/j.jhydrol.2025.134371},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134371}
}