Zhao et al. (2026) Daily Streamflow Prediction Using Multi-State Transition SB-ARIMA-MS-GARCH Model

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Water
• Year: 2026
• Date: 2026-01-16
• Authors: Jin Zhao, Jianhui Shang, Qun Ye, Huimin Wang, G. X. Zhang, Feng Yao, Weiwei Shou
• DOI: 10.3390/w18020241

Research Groups

Not explicitly mentioned in the provided text. The study focuses on five hydrological stations in the middle reaches of the Yellow River.

Short Summary

This study develops multi-state Markov-switching GARCH models incorporating structural breaks to improve daily streamflow prediction accuracy, demonstrating that accounting for these changes significantly enhances the characterization of volatility and forecast performance in the Yellow River basin.

Objective

• To develop and evaluate a multi-state Markov-switching GARCH (MS-GARCH) modeling framework that accounts for structural breaks (SBs) in daily streamflow time series to improve prediction accuracy.

Study Configuration

• Spatial Scale: Five hydrological stations in the middle reaches of the Yellow River, China.
• Temporal Scale: Daily streamflow data.

Methodology and Data

• Models used: Autoregressive Integrated Moving Average (ARIMA), Generalized Autoregressive Conditional Heteroskedasticity (GARCH), GJR-GARCH (gjrGARCH), Markov-switching (MS), SB-ARIMA-MS-GARCH, SB-ARIMA-MS-gjrGARCH, ARIMA-GARCH, ARIMA-gjrGARCH.
• Data sources: Daily streamflow observations from five hydrological stations.

Main Results

• Daily streamflow series exhibit structural breaks, with the number and timing of breakpoints varying among stations.
• Standard GARCH and gjrGARCH models show limited ability to capture runoff volatility clustering.
• MS-GARCH and MS-gjrGARCH models effectively characterize volatility features within individual states.
• The multi-state switching structure substantially improves daily streamflow prediction accuracy compared with single-state volatility models, increasing the coefficient of determination (R²) by approximately 5.8% and the Nash-Sutcliffe efficiency (NSE) by approximately 36.3%.

Contributions

• Introduces a robust new modeling framework (SB-ARIMA-MS-GARCH/gjrGARCH) for streamflow prediction that explicitly accounts for structural breaks and multi-state volatility.
• Addresses a critical limitation in previous GARCH studies by incorporating changes in series structure, leading to more accurate identification of volatility forms.
• Provides more reliable evidence for water resource management and flood risk mitigation in changing hydrological environments like the Yellow River basin.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Zhao2026Daily,
  author = {Zhao, Jin and Shang, Jianhui and Ye, Qun and Wang, Huimin and Zhang, G. X. and Yao, Feng and Shou, Weiwei},
  title = {Daily Streamflow Prediction Using Multi-State Transition SB-ARIMA-MS-GARCH Model},
  journal = {Water},
  year = {2026},
  doi = {10.3390/w18020241},
  url = {https://doi.org/10.3390/w18020241}
}