Feng et al. (2025) Disentangling atmospheric, hydrological, and coupling uncertainties in compound flood modeling within a coupled Earth system model

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2025
• Date: 2025-09-26
• Authors: Dongyu Feng, Zeli Tan, Darren Engwirda, Jonathan D. Wolfe, Donghui Xu, Chang Liao, Gautam Bisht, James J. Benedict, Tian Zhou, Mithun Deb, Hong‐Yi Li, L. Ruby Leung
• DOI: 10.5194/nhess-25-3619-2025

Research Groups

• Atmospheric, Climate, and Earth Sciences Division, Pacific Northwest National Laboratory, Richland, WA, USA
• T-3 Fluid Dynamics and Solid Mechanics Group, Los Alamos National Laboratory, Los Alamos, NM, USA
• Marine and Coastal Research Laboratory, Pacific Northwest National Laboratory, Sequim, WA, USA
• Department of Civil and Environmental Engineering, University of Houston, Houston, TX, USA

Short Summary

This study leverages the Energy Exascale Earth System Model (E3SM) with multi-component coupling to disentangle atmospheric, hydrological, and coupling uncertainties in compound riverine and coastal flood modeling. It demonstrates the critical role of two-way river-ocean coupling and antecedent soil moisture conditions in amplifying flood impacts, advocating for a broader definition of compound flooding.

Objective

• To evaluate compound uncertainties arising from two-way land–river–ocean coupling within the Energy Exascale Earth System Model (E3SM) and track the cascading meteorological and hydrological uncertainties through ensemble simulations.
• To assess the roles of hydrological drivers, such as infiltration and soil moisture, in the generation of compound flooding and their susceptibility to often overlooked drivers like antecedent soil moisture conditions.

Study Configuration

• Spatial Scale: Global (E3SM), with variable-resolution meshes refined to approximately 25 km for the atmosphere over the North Atlantic Ocean and eastern North America, 3 km for land/river within Mid-Atlantic watersheds, and 250 m for the ocean along the US East Coast. The study focuses on the Delaware River basin and estuary.
• Temporal Scale: Hurricane Irene event (August 2011) for ensemble simulations; historical soil moisture trends from 2005–2011 for antecedent soil moisture conditions; daily and event-accumulated impacts analyzed.

Methodology and Data

• Models used:
  • Energy Exascale Earth System Model (E3SMv2) coastal configuration.
  • E3SM Atmospheric Model (EAM) (3D).
  • E3SM Land Model (ELM) (1D).
  • Model for Scale Adaptive River Transport (MOSART) (1D).
  • Model for Prediction Across Scales ocean model (MPAS-O) (2D barotropic).
  • Artificial Neural Networks (ANNs) for quantifying hydrological driver importance.
  • Structural Equation Modeling (SEM) for path analysis of uncertainty propagation.
• Data sources:
  • ECMWF Reanalysis v5 (ERA5) for EAM initialization.
  • Global Soil Wetness Projects version 3 (GSWPv3) for ELM/MOSART spin-up.
  • 90 m HydroSHEDS digital elevation model (DEM) for MOSART bathymetry.
  • 450 m GEBCO dataset for MPAS-O bathymetry.
  • USGS streamflow measurements for MOSART validation.
  • NOAA tidal gauges for MPAS-O water level assessment.
  • 250 m resolution satellite imagery (Tellman et al., 2021) for inundation extent benchmarking.

Main Results

• The E3SM coastal configuration, with two-way river–ocean coupling, significantly improves the simulation of compound riverine and coastal inundation, notably increasing the hit rate and success index by 2-fold and doubling the predicted flooded area compared to riverine-only simulations.
• Two-way land–river coupling leads to a decrease in peak discharge (10–50 m³/s) and riverine flooded area due to floodplain infiltration, while two-way river–ocean coupling increases local streamflow and inundation near the river outlet by accurately representing backwater effects.
• Meteorological uncertainty, originating from atmospheric simulations, amplifies as it cascades through the multi-component system, resulting in approximately 2-fold higher variability in riverine flood parameters (river discharge and inundation area) compared to precipitation.
• Antecedent soil moisture conditions (AMC) and runoff generation parameters (fover, fdrain) are crucial hydrological drivers, with a saturated AMC scenario potentially increasing the flooded area by more than 2-fold (∼2.4) compared to the actual Hurricane Irene event.
• The relative importance of hydrological drivers evolves dynamically during a flood event, with surface runoff dominating during peak flow and soil moisture playing a significant buffering role post-peak.

Contributions

• First comprehensive assessment of atmospheric, hydrological, and coupling uncertainties in compound flood modeling within a fully coupled Earth System Model (E3SM).
• Demonstrates the critical importance of two-way land–river and river–ocean coupling for accurate compound flood simulations in coastal regions.
• Quantifies the cascading amplification of meteorological uncertainty through an integrated Earth system framework.
• Highlights the significant, often underappreciated, role of hydrological drivers, particularly antecedent soil moisture conditions, in modulating tropical cyclone-induced compound flood severity.
• Proposes a broader definition of compound flooding that encompasses the simultaneous occurrence of intense precipitation, storm surge, and high antecedent soil moisture conditions during tropical cyclones.
• Introduces a two-stage Artificial Neural Network (ANN) approach with permutation importance to quantify the relative importance and temporal evolution of hydrological drivers.

Funding

• Earth System Model Development program areas of the US Department of Energy, Office of Science, Office of Biological and Environmental Research, as part of the multi-program, collaborative Integrated Coastal Modeling (ICoM) project (grant no. KP1703110/75415).
• Earth and Environmental Systems Sciences Division of the US Department of Energy, Office of Science, Office of Biological and Environmental Research, as part of the project “A strategic partnership between the College of Engineering at University of Houston and Pacific Northwest National Lab” (grant no. DE-SC0023295).

Citation

@article{Feng2025Disentangling,
  author = {Feng, Dongyu and Tan, Zeli and Engwirda, Darren and Wolfe, Jonathan D. and Xu, Donghui and Liao, Chang and Bisht, Gautam and Benedict, James J. and Zhou, Tian and Deb, Mithun and Li, Hong‐Yi and Leung, L. Ruby},
  title = {Disentangling atmospheric, hydrological, and coupling uncertainties in compound flood modeling within a coupled Earth system model},
  journal = {Natural hazards and earth system sciences},
  year = {2025},
  doi = {10.5194/nhess-25-3619-2025},
  url = {https://doi.org/10.5194/nhess-25-3619-2025}
}