Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

A Comparative Study of Physics‐Informed and Data‐Driven Neural Networks for Compound Flood Simulation at River‐Ocean Interfaces: A Case Study of Hurricane Irene

Journal Article · · Journal of Geophysical Research: Machine Learning and Computation
DOI:https://doi.org/10.1029/2025JH000758· OSTI ID:2997685

Simulating compound flooding (CF) at the river-ocean interface within large-scale Earth System Models (ESMs) presents significant challenges due to complex interactions between river discharge, storm surge, and tides. This study assesses the comparative advantages of physics-informed and data-driven machine learning (ML) approaches for enhancing local ESM performance. We systematically compare data-driven neural network models (i.e., CNNs, U-Net, Long Short-Term Memory (LSTM), Gated Recurrent Unit), and physics-informed neural network (PINN) models, including vanilla PINN and a finite-difference-based PINN (FD-PINN). Specifically, FD-PINN is introduced to enhance computational efficiency, accelerating vanilla PINNs by ∼6.5 times while improving accuracy. To enhance data-driven model training, a new data-generation approach is developed to sample historical fluvial and coastal flood events, which ensures a robust data set for extreme event prediction. The models are evaluated using a realistic one-dimensional river domain extracted from an ESM's river mesh and the Hurricane Irene event as an independent test case. Results show that FD-PINN achieves accurate predictions with significantly reduced computational costs relative to vanilla PINNs. Among data-driven models, the best overall performance is achieved by a CNN-LSTM hybrid, which balances accuracy and efficiency. While a fully connected CNN (CNN-FC) provides the best accuracy, it incurs high computational cost. Architectures lacking strong temporal modeling tend to underperform on unseen events. These findings highlight the importance of sequence-aware designs for robust generalization. This study reveals the trade-offs between physics-informed and data-driven models and proposes an adaptive hybrid framework for integrating ML into ESMs to enhance local flood simulations.

Research Organization:
Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
Sponsoring Organization:
USDOE Office of Science (SC), Biological and Environmental Research (BER)
Grant/Contract Number:
AC05-76RL01830
OSTI ID:
2997685
Report Number(s):
PNNL-SA--211143
Journal Information:
Journal of Geophysical Research: Machine Learning and Computation, Journal Name: Journal of Geophysical Research: Machine Learning and Computation Journal Issue: 4 Vol. 2; ISSN 2993-5210
Publisher:
American Geophysical Union (AGU)Copyright Statement
Country of Publication:
United States
Language:
English

References (88)

Prolongation of SMAP to Spatiotemporally Seamless Coverage of Continental U.S. Using a Deep Learning Neural Network journal November 2017
Compound simulation of fluvial floods and storm surges in a global coupled river-coast flood model: Model development and its application to 2007 Cyclone Sidr in Bangladesh: A GLOBAL COUPLED RIVER-COAST FLOOD MODEL journal August 2017
Deep Learning for the Earth Sciences book August 2021
Applications of finite difference‐based physics‐informed neural networks to steady incompressible isothermal and thermal flows journal June 2023
Enhanced physics‐informed neural networks for efficient modelling of hydrodynamics in river networks journal April 2024
Transfer learning‐based physics‐informed neural networks for magnetostatic field simulation with domain variations
  • Lippert, Jonathan Rainer; von Tresckow, Moritz; De Gersem, Herbert
  • International Journal of Numerical Modelling: Electronic Networks, Devices and Fields, Vol. 37, Issue 4 https://doi.org/10.1002/jnm.3264
journal July 2024
Physics-informed neural networks for multiphysics data assimilation with application to subsurface transport journal July 2020
Spatio-temporal attention-based hidden physics-informed neural network for remaining useful life prediction journal January 2025
Solving groundwater flow equation using physics-informed neural networks journal September 2023
Flood spatial prediction modeling using a hybrid of meta-optimization and support vector regression modeling journal April 2021
Neural-Integrated Meshfree (NIM) Method: A differentiable programming-based hybrid solver for computational mechanics journal July 2024
Frequency-adaptive multi-scale deep neural networks journal March 2025
Challenges in Continental River Dynamics journal December 2013
A comprehensive review of compound inundation models in low-gradient coastal watersheds journal September 2019
Automatic procedure for selecting flood events and identifying flood characteristics from daily streamflow data journal November 2021
Improved flood susceptibility mapping using a best first decision tree integrated with ensemble learning techniques journal May 2021
Inverse flow prediction using ensemble PINNs and uncertainty quantification journal July 2024
Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations journal February 2019
NSFnets (Navier-Stokes flow nets): Physics-informed neural networks for the incompressible Navier-Stokes equations journal February 2021
A physics-informed convolutional neural network for the simulation and prediction of two-phase Darcy flows in heterogeneous porous media journal March 2023
Multi-head physics-informed neural networks for learning functional priors and uncertainty quantification journal June 2025
Forecasting of compound ocean-fluvial floods using machine learning journal July 2024
Generation and selection of training events for surrogate flood inundation models journal January 2025
Characterizing water surface elevation under different flow conditions for the upcoming SWOT mission journal June 2018
Flood susceptibility mapping using convolutional neural network frameworks journal March 2020
A deep convolutional neural network model for rapid prediction of fluvial flood inundation journal November 2020
Using machine learning models, remote sensing, and GIS to investigate the effects of changing climates and land uses on flood probability journal April 2021
XGBoost-based method for flash flood risk assessment journal July 2021
Can artificial intelligence and data-driven machine learning models match or even replace process-driven hydrologic models for streamflow simulation?: A case study of four watersheds with different hydro-climatic regions across the CONUS journal July 2021
Urban flood modeling using deep-learning approaches in Seoul, South Korea journal October 2021
Combined statistical and hydrodynamic modelling of compound flooding in coastal areas - Methodology and application journal May 2023
1-D coupled surface flow and transport equations revisited via the physics-informed neural network approach journal October 2023
Flood forecasting based on radar precipitation nowcasting using U-net and its improved models journal March 2024
Physics-informed neural network for diffusive wave model journal June 2024
Physics-informed neural networks for solving flow problems modeled by the 2D Shallow Water Equations without labeled data journal June 2024
Efficient river hydrodynamics modelling in realistic river systems using a Fourier neural operator-based network journal June 2024
Exploring the compound nature of coastal flooding by tropical cyclones: A machine learning framework journal December 2024
Use of machine learning algorithms to assess flood susceptibility in the coastal area of Bangladesh journal April 2023
Gaussian process metamodeling of functional-input code for coastal flood hazard assessment journal June 2020
A comparative assessment of decision trees algorithms for flash flood susceptibility modeling at Haraz watershed, northern Iran journal June 2018
Modeling flood susceptibility using data-driven approaches of naïve Bayes tree, alternating decision tree, and random forest methods journal January 2020
From local to regional compound flood mapping with deep learning and data fusion techniques journal August 2021
Graph neural network-based surrogate modelling for real-time hydraulic prediction of urban drainage networks journal October 2024
Data-Driven Science and Engineering book January 2022
The DOE E3SM Coupled Model Version 1: Overview and Evaluation at Standard Resolution journal July 2019
Emulation of the Saint Venant Equations Enables Rapid and Accurate Predictions of Infiltration and Overland Flow Velocity on Spatially Heterogeneous Surfaces journal August 2019
A Rainfall‐Runoff Model With LSTM‐Based Sequence‐to‐Sequence Learning journal January 2020
Physics‐Informed Neural Network Method for Forward and Backward Advection‐Dispersion Equations journal July 2021
Development of Land‐River Two‐Way Hydrologic Coupling for Floodplain Inundation in the Energy Exascale Earth System Model journal August 2022
Integrating Recurrent Neural Networks With Data Assimilation for Scalable Data‐Driven State Estimation journal March 2022
Storm Surge Modeling as an Application of Local Time‐Stepping in MPAS‐Ocean journal January 2023
Recent Advances and New Frontiers in Riverine and Coastal Flood Modeling journal June 2023
Multiphysics‐Informed Neural Networks for Coupled Soil Hydrothermal Modeling journal January 2023
Differentiable, Learnable, Regionalized Process‐Based Models With Multiphysical Outputs can Approach State‐Of‐The‐Art Hydrologic Prediction Accuracy journal October 2022
Physics‐Informed Neural Networks of the Saint‐Venant Equations for Downscaling a Large‐Scale River Model journal February 2023
The DOE E3SM Model Version 2: Overview of the Physical Model and Initial Model Evaluation journal December 2022
Closing in on Hydrologic Predictive Accuracy: Combining the Strengths of High‐Fidelity and Physics‐Agnostic Models journal September 2023
Simulation of Compound Flooding Using River‐Ocean Two‐Way Coupled E3SM Ensemble on Variable‐Resolution Meshes journal June 2024
Physics‐Informed Neural Networks for the Augmented System of Shallow Water Equations With Topography journal October 2024
Impacts of Sea‐Level Rise on Coastal Groundwater Table Simulated by an Earth System Model With a Land‐Ocean Coupling Scheme journal August 2024
Recommendations for Comprehensive and Independent Evaluation of Machine Learning‐Based Earth System Models journal March 2025
Modeling Non‐Stationary Wind‐Induced Fluid Motions With Physics‐Informed Neural Networks for the Shallow Water Equations in a Polar Coordinate System journal January 2025
Uncertainties in Simulating Flooding During Hurricane Harvey Using 2D Shallow Water Equations journal January 2025
Improving Streamflow Prediction Using Multiple Hydrological Models and Machine Learning Methods journal January 2025
Increasing risk of compound flooding from storm surge and rainfall for major US cities journal July 2015
Combined 1D and 2D Hydraulic Modeling within HEC-RAS conference May 2015
An application of local linear radial basis function neural network for flood prediction journal January 2019
Water Level Prediction Model Based on GRU and CNN journal January 2020
Comparison of Deep Learning Techniques for River Streamflow Forecasting journal January 2021
Adaptive-Sampling Physics-Informed Neural Network for Viscoacoustic Wavefield Simulation journal January 2024
Learning in Sinusoidal Spaces With Physics-Informed Neural Networks journal March 2024
Higher probability of compound flooding from precipitation and storm surge in Europe under anthropogenic climate change journal September 2019
Integrating Scientific Knowledge with Machine Learning for Engineering and Environmental Systems journal November 2022
Flood Susceptibility Mapping Using Image‐Based 2D‐CNN Deep Learning: Overview and Case Study Application Using Multiparametric Spatial Data in Data‐Scarce Urban Environments journal January 2023
Regularization Theory and Neural Networks Architectures journal March 1995
The Model Coupling Toolkit: A New Fortran90 Toolkit for Building Multiphysics Parallel Coupled Models journal August 2005
On Pixel-Wise Explanations for Non-Linear Classifier Decisions by Layer-Wise Relevance Propagation journal July 2015
A comprehensive review of deep learning applications in hydrology and water resources journal August 2020
Discharge measurements at gaging stations report January 2010
Forecasting Flood Inundation in U.S. Flood-Prone Regions Through a Data-Driven Approach (FIER): Using VIIRS Water Fractions and the National Water Model journal November 2024
Artificial Neural Network and Multiple Linear Regression for Flood Prediction in Mohawk River, New York journal August 2018
Application of Long Short-Term Memory (LSTM) Neural Network for Flood Forecasting journal July 2019
A Physics-Informed, Machine Learning Emulator of a 2D Surface Water Model: What Temporal Networks and Simulation-Based Inference Can Help Us Learn about Hydrologic Processes journal December 2021
A Deep U-Net-ConvLSTM Framework with Hydrodynamic Model for Basin-Scale Hydrodynamic Prediction journal February 2024
Hybrid Finite Difference with the Physics-informed Neural Network for solving PDE in complex geometries preprint January 2022
Solving Oscillator Ordinary Differential Equations in the Time Domain with High Performance via Soft-constrained Physics-informed Neural Network with Small Data preprint January 2024
Physics-Informed Neural Networks for Solving the Two-Dimensional Shallow Water Equations with Terrain Topography and Rainfall Source Terms preprint January 2025
Data and Code for 'A Comparative Study of Physics-Informed and Data-Driven Neural Networks for Compound Flood Simulation at River-Ocean Interfaces: A Case Study of Hurricane Irene' dataset January 2025

Similar Records

Use of Physics to Improve Solar Forecast: Part II, Machine Learning and Model Interpretability
Journal Article · Thu Sep 01 00:00:00 EDT 2022 · Solar Energy · OSTI ID:1890143

Use of physics to improve solar forecast: Part II, machine learning and model interpretability
Journal Article · Thu Sep 01 00:00:00 EDT 2022 · Solar Energy · OSTI ID:1887823

Techno-economic implications and cost of forecasting errors in solar PV power production using optimized deep learning models
Journal Article · Thu Mar 27 00:00:00 EDT 2025 · Energy · OSTI ID:2587373