Kim et al. (2026) Assessment and attribution of predictive hydrologic uncertainty in operational short-to-medium range ensemble streamflow forecasting

Identification

• Journal: Journal of Hydrology
• Year: 2026
• Date: 2026-06-17
• Authors: Sunghee Kim, Dong‐Jun Seo
• DOI: 10.1016/j.jhydrol.2026.135893

Research Groups

• Department of Civil Engineering, The University of Texas at Arlington, Arlington, TX, United States

Short Summary

This study assesses and attributes predictive hydrologic uncertainty in short-to-medium range ensemble streamflow forecasting via 20-yr hindcast experiments for 398 headwater locations in the US. The results show that there is large room for improvement in forecast skill from reducing hydrologic uncertainty.

Objective

• Assess the limits of predictive hydrologic skill for short-to-medium range ensemble streamflow forecasting.
• Attribute predictive hydrologic uncertainty to selected sources by isolating the lead time-dependent initial condition (IC) uncertainty via state updating.

Study Configuration

• Spatial Scale: Headwater locations in the service areas of 13 River Forecast Centers in the US.
• Temporal Scale: 20-year hindcast experiments for each location.

Methodology and Data

• Models used: Sacramento (SAC), Unit Hydrograph (UHG), SNOW17, Adjoint-Based OPTimizer (AB_OPT).
• Data sources: Hydrologic models, observed input forcings, streamflow observations from the US Geological Survey.

Main Results

• The results show that hydrologic uncertainty is a significant contributor to total uncertainty in short-to-medium range ensemble streamflow forecasting.
• IC uncertainty at saturation accounts for almost half of the unresolved hydrologic uncertainty on average.
• State updating improves predictive skill by reducing IC uncertainty, but its impact diminishes with increasing lead time.

Contributions

• Advances understanding and assessment of the limits of hydrologic predictive skill for ensemble streamflow forecasting.
• Provides a new approach to attribute predictive hydrologic uncertainty to selected sources via state updating.
• Identifies areas and locations needing particular attention for reducing hydrologic uncertainty.

Funding

• This work was supported by the National Science Foundation (Grant Number: NSF-2020).

Citation

@article{Kim2026Assessment,
  author = {Kim, Sunghee and Seo, Dong‐Jun},
  title = {Assessment and attribution of predictive hydrologic uncertainty in operational short-to-medium range ensemble streamflow forecasting},
  journal = {Journal of Hydrology},
  year = {2026},
  doi = {10.1016/j.jhydrol.2026.135893},
  url = {https://doi.org/10.1016/j.jhydrol.2026.135893}
}