Wang et al. (2026) Advancing Physical Realism in Hydrological Modelling: Selection and Integration — A Review and Synthesis

Identification

• Journal: Water Resources Management
• Year: 2026
• Date: 2026-03-01
• Authors: Kunyang Wang, Shin‐ichi Onodera, Mitsuyo Saito, Yu War Nang
• DOI: 10.1007/s11269-026-04534-1

Research Groups

• Graduate School of Advanced Science and Engineering, Hiroshima University, Higashi-Hiroshima, Japan

Short Summary

This review synthesizes 30 hydrological models and 186 peer-reviewed studies to propose a decision-oriented framework for model selection and integration, aiming to enhance physical realism, address data scarcity, and improve long-term and sub-daily simulations for resilient water resources management.

Objective

• To propose a decision-oriented model-selection framework that maps model capabilities to fit-for-purpose hydro-environmental research and management tasks.
• To explore the trends, prospects, and challenges of hydrological models in future work.
• To consider the utility and availability of integrated models and other advanced technologies.

Study Configuration

• Spatial Scale: Field, urban, peri-urban, catchment, regional, and global scales.
• Temporal Scale: Sub-hourly (e.g., 6-minute, 15-minute), hourly, daily, and long-term (decades to millennia) dynamics.

Methodology and Data

• Models used: Soil and Water Assessment Tool (SWAT), MODFLOW, Hydrologic Engineering Center's Hydrologic Modeling System (HEC-HMS), MIKE SHE, Topography based Hydrological Model (TOPMODEL), Variable Infiltration Capacity (VIC), Precipitation Runoff Modeling System (PRMS), Identification of unit Hydrographs And Component flows from Rainfall, Evaporation and Streamflow (IHACRES), Water Evaluation And Planning System (WEAP), LISFLOOD, Agricultural Policy/Environmental extender (APEX), CropWaT, Gridded Surface Subsurface Hydrologic Analysis (GSSHA), Weather Research and Forecasting-Hydro modeling system (WRF-Hydro), CE-QUAL-W2, SIMHYD, River Basin Simulation (RIBASIM), Distributed time-variant gain model (DTVGM), WaterGAP/Water GAP 2/WGHM, Hydrological Predictions for the Environment (HYPE), Community Land Model (CLM), Urban Runoff Branching Structure (URBS), Stormwater Management Model (SWMM), Peri-Urban Model for landscape Management (PUMMA), Catchment-based Macro-scale Floodplain (Cama-Flood), Soil–Water-Atmosphere-Plant (SWAP), Sacramento Soil Moisture Accounting (SAC-SMA), Hydrologiska Byråns Vattenbalansavdelning (HBV), FEFLOW, GR4.
• Data sources: Systematic literature search of the Web of Science Core Collection (PRISMA 2020 guidelines), synthesis of 186 peer-reviewed journal articles, multi-source satellite datasets, deep learning, real-time satellite observations, field observations, UAV-derived canopy information, natural archives (lake sediments, tree rings, ice cores).

Main Results

• Thirty hydrological models were classified by structural and functional characteristics (distributed, semi-distributed, lumped; physically based, process-based, conceptual) and mapped to specific hydro-environmental research and management tasks.
• Twenty-four model integration modes were synthesized, categorized into surface–subsurface, functional extending, scale bridging, and process-to-decision integration, highlighting the benefits of combining complementary model capabilities.
• Strategic directions for advancing hydrological modeling were outlined, emphasizing long-term dynamics for understanding centennial feedback and planetary boundaries, and sub-daily dynamics for forecasting compound extreme events.
• The concept of "Hydro-Limit Science (HLS)" was proposed to quantify thresholds for hydrological extremes under climate stress and integrate with planetary boundaries.
• A "process-guided transfer learning" approach was advocated for data-scarce and ungauged regions, combining physics-based models with machine learning and remote sensing.
• Key challenges in achieving physical realism were identified, including the need for finer spatial and temporal resolution, fusion of multisource data via physics-guided AI, calibration of more physical processes, improved parameterization, integrated modeling frameworks, and enhanced geological representation.
• Three criteria for improving and assessing physical realism were proposed: accuracy/performance (beyond Nash–Sutcliffe efficiency, emphasizing PBIAS across flow regimes), data/evidence (targeted input subdivision, calibration of key processes), and consistency/applicability (physically interpretable parameters, transferability across scales).

Contributions

• Provides a decision-oriented framework and map linking hydrological model structures and process representations to diverse planning, operation, protection, and risk-management tasks.
• Synthesizes practical strategies for model integration, including 24 distinct modes, based on interface characteristics and input–output constraints.
• Proposes a physical-realism checklist and criteria (finer resolution, pre-parameterization, multi-process checks) to guide model development and evaluation.
• Emphasizes the critical role of long-term modeling for slow feedback and sub-daily calibration for extreme events in the context of planetary boundaries.
• Outlines pathways for ungauged and data-scarce basins through physics-guided machine learning integrated with multi-source remote sensing.
• Advances adaptive water resources management by offering an integrated cognitive and methodological framework for model choice, integration, and evaluation.

Funding

• Grant-in-Aid for Fund for the Promotion of Joint International Research [Fostering Joint International Research (B)] of the Japan Society for the Promotion of Science (Project No. 21KK0192, PI: Shin-ichi Onodera)
• Grant-in-Aid for Scientific Research (A) of the Japan Society for the Promotion of Science (Project No. 18H04151, PI: Shin-ichi Onodera)
• Asia–Pacific Network for Global Change Research (APN; Grant No. CRRP2019-09MY-Onodera, funder ID: https://doi.org/10.13039/100005536)
• Open Access funding provided by Hiroshima University.

Citation

@article{Wang2026Advancing,
  author = {Wang, Kunyang and Onodera, Shin‐ichi and Saito, Mitsuyo and Nang, Yu War},
  title = {Advancing Physical Realism in Hydrological Modelling: Selection and Integration — A Review and Synthesis},
  journal = {Water Resources Management},
  year = {2026},
  doi = {10.1007/s11269-026-04534-1},
  url = {https://doi.org/10.1007/s11269-026-04534-1}
}