Li et al. (2025) Comparing the impact of precipitation pre-processing and streamflow post-processing for daily sub-seasonal streamflow forecasts over the Gan River basin

Identification

• Journal: Journal of Hydrology
• Year: 2025
• Date: 2025-12-09
• Authors: Yuan Li, Guangping Xu, Jian Xu, Quan J. Wang, Qichun Yang, Zhiyong Wu, Peng Tang, Kangning Xu, Qing Tian
• DOI: 10.1016/j.jhydrol.2025.134773

Research Groups

• National Key Laboratory of Water Disaster Prevention, Hohai University, Nanjing, China
• College of Hydrology and Water Resources, Hohai University, Nanjing, China
• Department of Infrastructure Engineering, The University of Melbourne, Parkville, Australia
• Earth, Ocean and Atmospheric Sciences Thrust, Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China

Short Summary

This study investigates the separate and joint impacts of precipitation pre-processing and streamflow post-processing on daily sub-seasonal streamflow forecasts over the Gan River Basin. It finds that streamflow post-processing is more impactful for lead times within 10 days, while precipitation pre-processing becomes more significant at longer lead times, with specific methods (BJP and SSh) showing superior performance.

Objective

• To investigate the separate and joint impacts of precipitation pre-processing and streamflow post-processing for daily sub-seasonal streamflow forecasts over the Gan River Basin.

Study Configuration

• Spatial Scale: Gan River Basin (GRB)
• Temporal Scale: Daily sub-seasonal streamflow forecasts, with lead times compared for periods within 10 days and at longer durations.

Methodology and Data

• Models used: Bayesian Joint Probability (BJP), Quantile Mapping (QM), Schaake Shuffle (SSh), Ensemble Copula Coupling (ECC), Xinanjiang (XAJ) hydrological model, revised Error Reduction and Representation In Stages (ERRIS) model, Ensemble Streamflow Prediction (ESP).
• Data sources: Calibrated precipitation forecasts, historical hydrometeorological sequences (for ESP), initial hydrologic conditions.

Main Results

• The revised ERRIS model (streamflow post-processing) has a greater impact on streamflow forecasting compared to precipitation pre-processing when the lead time is within 10 days, regardless of the pre-processing or reordering techniques used.
• Precipitation pre-processing has a greater impact on streamflow at longer lead times.
• BJP pre-processed precipitation forecasts exhibit higher accuracy and reliability than those from the QM method for each sub-basin and lead to better streamflow forecasts.
• Streamflow forecasts driven by SSh reordered precipitation forecasts demonstrate higher accuracy and reliability compared to those driven by ECC reordered precipitation forecasts.
• Both precipitation pre-processing and streamflow post-processing are necessary for achieving better sub-seasonal streamflow forecasts.

Contributions

• Provides a systematic comparison of the separate and joint impacts of precipitation pre-processing and streamflow post-processing within different hydrometeorological forecasting systems.
• Identifies the relative importance of pre-processing versus post-processing techniques based on forecast lead time.
• Evaluates the performance of specific precipitation pre-processing (BJP, QM) and reordering (SSh, ECC) methods in improving streamflow forecast accuracy and reliability.
• Demonstrates the combined necessity of both precipitation pre-processing and streamflow post-processing for enhancing sub-seasonal streamflow forecast skill.

Funding

• [No funding information was provided in the excerpt.]

Citation

@article{Li2025Comparing,
  author = {Li, Yuan and Xu, Guangping and Xu, Jian and Wang, Quan J. and Yang, Qichun and Wu, Zhiyong and Tang, Peng and Xu, Kangning and Tian, Qing},
  title = {Comparing the impact of precipitation pre-processing and streamflow post-processing for daily sub-seasonal streamflow forecasts over the Gan River basin},
  journal = {Journal of Hydrology},
  year = {2025},
  doi = {10.1016/j.jhydrol.2025.134773},
  url = {https://doi.org/10.1016/j.jhydrol.2025.134773}
}