Mkhonta et al. (2026) Towards improved flood prediction: a review of deterministic hydrologic-hydraulic model coupling

Identification

• Journal: Natural Hazards
• Year: 2026
• Date: 2026-03-30
• Authors: Glen Mkhonta, Nicholas Byaruhanga, Daniel Kibirige
• DOI: 10.1007/s11069-026-07992-3

Research Groups

• Centre for Water Resources Research, School of Agriculture and Science, University of KwaZulu-Natal, South Africa
• Environmental and Geographical Science, University of Cape Town, South Africa
• Water Research Centre, Smart Places Cluster, Council for Scientific and Industrial Research (CSIR), South Africa

Short Summary

This scoping review synthesizes 94 peer-reviewed studies from 1994 to 2024 to trace the evolution of deterministic hydrologic–hydraulic model coupling for flood forecasting, highlighting model prevalence, performance, and regional disparities, particularly in data-scarce regions like Africa. It finds that HEC-HMS and HEC-RAS are the most widely used models, consistently achieving high predictive accuracy, and emphasizes the need for context-specific solutions and improved data infrastructure in underrepresented areas.

Objective

• To examine the most popular, feasible, and accurate hydrologic and hydraulic models in the context of flood prediction and forecasting.
• To provide an overview of the chronological evolution of integrated or coupled hydrologic-hydraulic models in the context of flood prediction and forecasting.

Study Configuration

• Spatial Scale: Global review of studies, with specific focus and examples from high-income nations (North America, Europe, Asia) and data-scarce regions (Africa, particularly Southern Africa).
• Temporal Scale: Review of peer-reviewed studies published between 1994 and 2024.

Methodology and Data

• Models used: HEC-HMS, HEC-RAS, SWAT, CREST, ANUGA, TOPMODEL, CARIMA, MARINE, IMECH-1D, XSRAIN, OMEGA, HEC-2, WRF-Hydro, SWMM, MIKE SHE, MIKE 11, SUPERFLEX, LISFLOOD-FP, Xinjiang-based hydrological model, HiPIMS, Flood2D-GPU, SCS-CN, VIC, Xinanjiang (XAJ).
• Data sources: Scoping literature review (SLR) of 94 peer-reviewed journal articles, conference papers, review articles, and books, identified using the PRISMA framework from Scopus, Web of Science, and Google Scholar databases. Bibliometric analysis was conducted using VOS viewer.

Main Results

• HEC-HMS is the most widely used hydrologic model (33% of studies), and HEC-RAS dominates hydraulic applications (45% of studies).
• The HEC-HMS + HEC-RAS coupling was the most frequent (23 studies) and consistently achieved strong predictive performance (coefficients of determination above 0.98 and inundation mapping accuracies of 80–94%).
• Other effective couplings, such as SWAT–HEC-RAS and CREST–ANUGA, demonstrated context-specific advantages but were less frequently applied.
• Regional disparities are pronounced, with 78% of studies concentrated in high-income nations (North America, Europe, Asia) and fewer than 5% from Africa, reflecting structural barriers like sparse hydrometric networks and limited data/capacity.
• The evolution of coupling techniques progressed from loose coupling (pre-2000s) to sequential/tight coupling (2000-2010), and then to fully coupled and hybrid (data-driven) models (post-2010).
• Deterministic model coupling is a practical pathway for enhancing flood early warning systems in data-scarce regions, provided calibration, validation, and local constraints are systematically addressed.

Contributions

• This review provides a comprehensive synthesis of the evolution of deterministic hydrologic-hydraulic model coupling for flood forecasting over three decades (1994-2024).
• It quantifies the prevalence and performance of specific models and coupling pairs, identifying HEC-HMS and HEC-RAS as dominant due to their technical reliability, open-source nature, ease of use, and institutional support.
• It critically highlights significant geographical inequalities in research contributions, particularly the underrepresentation of African contexts, and discusses the structural barriers hindering model adoption and validation in data-scarce regions.
• The study offers practical recommendations for governments and researchers in resource-limited environments, advocating for the strategic implementation of open-source models, leveraging remote sensing and crowdsourced data, and exploring hybrid modeling frameworks to improve flood early warning systems.

Funding

• Water Research Commission of South Africa in conjunction with uMngeni-uThukela Water in KwaZulu-Natal, South Africa, grant number C2023-2024-01256.

Citation

@article{Mkhonta2026Towards,
  author = {Mkhonta, Glen and Byaruhanga, Nicholas and Kibirige, Daniel},
  title = {Towards improved flood prediction: a review of deterministic hydrologic-hydraulic model coupling},
  journal = {Natural Hazards},
  year = {2026},
  doi = {10.1007/s11069-026-07992-3},
  url = {https://doi.org/10.1007/s11069-026-07992-3}
}