Sun et al. (2026) Assessing and optimizing high-resolution global river streamflow estimates with triple collocation analysis

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-02-15
• Authors: Mingze Sun, Natthachet Tangdamrongsub, Yu Sun, Jianzhi Dong, Edwin Sutanudjaja, Mikhail Smilovic
• DOI: 10.1016/j.jhydrol.2026.135122

Research Groups

• Key Lab of Spatial Data Mining and Information Sharing of Ministry of Education, Fuzhou University, Fujian, China
• National Engineering Research Center of Geospatial Information Technology, Fuzhou University, Fujian, China
• Water Engineering and Management, Faculty of Civil and Environmental Engineering, Asian Institute of Technology, Pathum Thani, Thailand
• Institute of Surface-Earth System Science, Tianjin University, Tianjin, China
• Department of Physical Geography, Utrecht University, Utrecht, the Netherlands
• Water Security Research Group, International Institute for Applied Systems Analysis (IIASA), Laxenburg 2031, Austria
• ETH Zürich, Chair of Hydrology and Water Resources Management, 8093 Zurich, Switzerland

Short Summary

This study evaluates three global hydrological models (CWatM, PCR-GLOBWB, H08) at high and low spatial resolutions and optimizes streamflow estimates using Triple Collocation and simple averaging. It finds that high-resolution CWatM and Triple Collocation-based data fusion significantly improve global streamflow simulation accuracy, offering practical guidance for hydrological assessments.

Objective

• To assess the performance of three global hydrological models (Community Water Model (CWatM), PCRaster Global Water Balance (PCR-GLOBWB), and H08) at both high (5 arcmin) and low (30 arcmin) spatial resolutions.
• To enhance streamflow simulations by employing two data fusion methods—Triple Collocation (TC) and simple averaging—to integrate outputs from the three models.

Study Configuration

• Spatial Scale: Global, with model evaluations at 5 arcmin and 30 arcmin spatial resolutions.
• Temporal Scale: Not explicitly defined in the provided text, but implies a continuous or long-term assessment of streamflow.

Methodology and Data

• Models used:
  • Global Hydrological Models: Community Water Model (CWatM), PCRaster Global Water Balance (PCR-GLOBWB), H08.
  • Data Fusion Methods: Triple Collocation (TC), Simple Averaging.
• Data sources:
  • Observed discharge data from 1,707 Global Runoff Data Centre (GRDC) stations.
  • Observed discharge data from 62 stations in Thailand.

Main Results

• CWatM, particularly at high resolution, consistently outperforms PCR-GLOBWB and H08.
• All models demonstrate stronger performance in Europe compared to other regions.
• Both Triple Collocation (TC) and simple averaging methods improve simulation accuracy compared to individual models.
• Triple Collocation (TC) offers the highest overall performance, especially when using H08 as the reference system.
• High-resolution models generally yield more accurate streamflow estimates in most regions.

Contributions

• Provides a systematic evaluation of three global hydrological models (CWatM, PCR-GLOBWB, H08) at varying spatial resolutions.
• Demonstrates the effectiveness of multi-model integration techniques (Triple Collocation and simple averaging) for enhancing global streamflow simulations.
• Highlights the superior performance of Triple Collocation, especially with H08 as a reference, and the benefits of high-resolution modeling.
• Offers practical guidance for model selection and data fusion strategies to improve hydrological assessments in ungauged or poorly monitored regions.

Funding

The provided paper text does not contain a funding section or explicit mention of funding projects/programs.

Citation

@article{Sun2026Assessing,
  author = {Sun, Mingze and Tangdamrongsub, Natthachet and Sun, Yu and Dong, Jianzhi and Sutanudjaja, Edwin and Smilovic, Mikhail},
  title = {Assessing and optimizing high-resolution global river streamflow estimates with triple collocation analysis},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.135122},
  url = {https://doi.org/10.1016/j.jhydrol.2026.135122}
}