Bayati et al. (2026) Enhancing large-scale basin rainfall-runoff modelling through the integration of flow routing: a case study in Iran’s Karun-4 Basin

Identification

• Journal: Acta Geophysica
• Year: 2026
• Date: 2026-04-06
• Authors: Samira Bayati, Khodayar Abdollahi, Afshin Honarbakhsh, Mohammad-Ali Nasr Esfahani
• DOI: 10.1007/s11600-026-01825-4

Research Groups

• Department of Natural Engineering, Faculty of Natural Resources and Earth Science, Shahrekord University, Shahrekord, Iran
• Department of Water Engineering, Faculty of Agriculture, Shahrekord University, Shahrekord, Iran

Short Summary

This study developed and evaluated a daily lumped water balance model for Iran's large-scale Karun-4 Basin, demonstrating that integrating flow routing significantly enhances streamflow simulation accuracy, particularly for peak flows, in data-scarce mountainous regions.

Objective

• To explore the role of considering/ignoring the flow routing component in daily rainfall-runoff simulations for a large-scale basin.
• To investigate the impact of integrating the linear Muskingum routing method on improving daily peak flow simulations.
• To provide insights into the importance of flow routing in enhancing the predictive capability of lumped models under data-scarce conditions.

Study Configuration

• Spatial Scale: Karun-4 basin, southwest of Iran, covering an area of 14,553 square kilometers (km²), with elevations ranging from 785 meters (m) to 4148 m above sea level.
• Temporal Scale: Daily data spanning from 2000 to 2020. The calibration period was 2000–2013, and the validation period was 2014–2020.

Methodology and Data

• Models used:
  • A newly developed daily lumped water balance model, incorporating a snow bucket, catchment bucket, TopModel for runoff generation, USAE method for snowmelt, probability distribution method for actual evapotranspiration (AET), and an empirical equation for interception.
  • The Exp-Hydro model (a Python-coded lumped hydrological model) as a benchmark, featuring a two-reservoir framework and thermal degree-day concept for snowmelt.
  • The linear Muskingum method for daily flow routing.
  • The Recursive Digital Filter (RDF) method (via the Web-based Hydrograph Analysis Tool - WHAT system) for base-flow separation.
• Data sources:
  • Daily temperature, potential evapotranspiration, rainfall, and discharge data.
  • Data obtained from the Iranian Meteorological Organization and the Chaharmahal-va-Bakhtiari’s Regional Water Authority.
  • Utilized 37 rain gauge stations, 8 evaporation measurement stations, 18 temperature measurement stations, and 4 hydrometric stations.
  • Missing data were reconstructed using normal ratio and non-linear regression methods.
  • Spatial averages for rainfall, temperature, and potential evaporation were calculated using the Thiessen polygons method.

Main Results

• The developed model, with Nash–Sutcliffe Efficiency (NSE) of 0.71 and Kling-Gupta Efficiency (KGE) of 0.75, significantly outperformed the Exp-Hydro model (NSE = 0.37, KGE = 0.69) during the calibration period.
• The incorporation of flow routing substantially enhanced the performance of the developed model in flow simulation, particularly improving the simulation of peak discharges.
• With flow routing, the percentage difference between observed and simulated runoff volume for the developed model decreased from 36% (calibration) and 27% (validation) to just 6% for both periods.
• The Muskingum method for daily flow routing consistently demonstrated "very good" performance (NSE > 0.75, KGE > 0.75, R² > 0.75) across all stations and hydrological periods (dry, wet, and entire study period).
• The developed model consistently delivered good results across both dry and wet periods, whereas the Exp-Hydro model struggled significantly in the dry season (NSE = 0.49).
• Both models showed a tendency to underestimate flood flows, but the developed model with routing showed better alignment with observed data.

Contributions

• Demonstrates the critical importance of integrating daily flow routing into lumped water balance models to significantly enhance streamflow simulation accuracy, especially for peak flows, in large-scale, data-scarce basins.
• Introduces a robust, newly developed lumped hydrological model that consistently outperforms an established benchmark (Exp-Hydro) in a challenging mountainous basin.
• Provides valuable insights for rainfall-runoff simulation in data-scarce regions, advocating for the adoption of lumped models with effective flow routing over data-intensive distributed models.
• Confirms the effectiveness and suitability of the linear Muskingum method for daily flow routing in the context of large, complex river systems.
• Connects improved hydrological simulation capabilities to broader applications in environmentally sustainable water and infrastructure management.

Funding

This research received no external funding.

Citation

@article{Bayati2026Enhancing,
  author = {Bayati, Samira and Abdollahi, Khodayar and Honarbakhsh, Afshin and Esfahani, Mohammad-Ali Nasr},
  title = {Enhancing large-scale basin rainfall-runoff modelling through the integration of flow routing: a case study in Iran’s Karun-4 Basin},
  journal = {Acta Geophysica},
  year = {2026},
  doi = {10.1007/s11600-026-01825-4},
  url = {https://doi.org/10.1007/s11600-026-01825-4}
}