Wang et al. (2026) Optimizing evapotranspiration products by merging multi-source datasets with three no-prior-knowledge methods over China

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-09-14
• Authors: Dayang Wang, Dagang Wang, Xiaogang Shi, Shaobo Liu, Mingfei Ji, Ya Huang, Zequn Lin, Kaihao Long, B. Larry Li, Haoyu Wang
• DOI: 10.1016/j.jhydrol.2026.136398

Research Groups

• College of Water Resources and Modern Agriculture, Nanyang Normal University, China
• School of Geography and Planning, Sun Yat-sen University, Guangzhou, China
• Department of Civil Engineering, University of New Brunswick, Fredericton, Canada
• School of Social and Environmental Sustainability, University of Glasgow, Dumfries, UK
• School of Hydrology and Water Resources, Nanjing University of Information Science and Technology, Nanjing, China
• School of Freshwater Science, University of Wisconsin-Milwaukee, WI, USA
• Ecological Complexity and Modelling Laboratory, Department of Botany and Plant Sciences, University of California, Riverside, CA, USA

Short Summary

This study proposes a data fusion framework for optimizing evapotranspiration (ET) products by merging four ET datasets using three fusion methods without prior knowledge. The results show that merged ET datasets generally demonstrate better performances than the original ET datasets.

Objective

• Investigate the feasibility of merging multi-source ET datasets to improve ET estimation accuracy

Study Configuration

• Spatial Scale: China-wide, with site and basin-scale evaluations
• Temporal Scale: Not specified

Methodology and Data

• Models used: Ensemble mean (EM), Bayesian-based three-cornered hat (BTCH) and extended triple collocation (ETC)
• Data sources: ERA5-Land, GLDAS, GLEAM, SiTH, and other ET datasets

Main Results

• Merged ET datasets generally demonstrate better performances than the original ET datasets in most sites and basins
• Lower mean bias error (MBE) and relative root mean square error (RRMSE), and higher coefficient of determination (R2) and Kling-Gupta efficiency (KGE)
• BTCH and ETC exhibit comparable ET fusion performances and are both superior to the EM method

Contributions

• This study provides valuable insights into the ET data fusion without prior knowledge, which can be applied to improve ET estimation accuracy in various regions.

Funding

• Not specified

Citation

@article{Wang2026Optimizing,
  author = {Wang, Dayang and Wang, Dagang and Shi, Xiaogang and Liu, Shaobo and Ji, Mingfei and Huang, Ya and Lin, Zequn and Long, Kaihao and Li, B. Larry and Wang, Haoyu},
  title = {Optimizing evapotranspiration products by merging multi-source datasets with three no-prior-knowledge methods over China},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.136398},
  url = {https://doi.org/10.1016/j.jhydrol.2026.136398}
}