Using Neural Networks to Predict Hurricane Storm Surge and to Assess the Sensitivity of Surge to Storm Characteristics

Abstract

Hurricane storm surge represents a significant threat to coastal communities around the world. Here, we use artificial neural network (ANN) models to predict storm surge levels using hurricane characteristics along the US Gulf and East Coasts. The ANN models are trained with storm surge levels from a hydrodynamic model and physical characteristics of synthetic hurricanes which are downscaled from National Centers for Environmental Prediction (NCEP) reanalysis using a statistical‐deterministic hurricane model. The ANN models are able to accurately predict storm surge levels with root‐mean‐square errors (RMSE) below 0.2 m and correlation coefficients > 0.85. The ANN models trained with the NCEP data also show satisfactory accuracy (RMSE below 0.7 m; correlation > 0.7) in predicting storm surge levels for hurricanes downscaled from future climate projections. Once trained, we use the ANN models to assess the sensitivity of storm surge levels to variations in hurricane characteristics and local geophysical features. Progressively stronger maximum wind speeds and larger outer radius sizes independently increase storm surge levels at all locations along the US East and Gulf Coasts. The response of storm surge levels to changes in hurricane translation speed, however, is found to be sensitive to coastal configuration, with increases in hurricane translation speed amplifying (reducing) storm surge levels in open ocean (semi‐enclosed) regions.

Increasing Flood Hazard Posed by Tropical Cyclone Rapid Intensification in a Changing Climate

“…This may be a result of the rain rates levels being more directly influenced by changes in storm intensity change, as in the TCRM. Additionally, storm surge is also significantly influenced by other factors unrelated to intensity, including the shape of the coastline, depth of the continental shelf and storm translation speed (Lockwood, Lin, et al., 2022).…”

“…We note that in this modeling set‐up with the outer size of the TCs derived from an empirical lognormal distribution (Chavas et al., 2015), wind intensity has an inverse correlation with Rmax. However, storm track including storm forward speed, can affect also storm surge levels, but its effect are complex depending on coastal geometry (Lockwood, Lin, et al., 2022). We have not examined if the track features of RI storms are different from those of non‐RI storms.…”

Model of Storm Surge Maximum Water Level Increase in a Coastal Area Using Ensemble Machine Learning and Explicable Algorithm

“…Previous research indicates that wind speed, air pressure, wind direction, and proximity are crucial for predicting storm surges using machine learning techniques (Ian et al, 2022;Tiggeloven et al, 2021). Moreover, to provide a more holistic understanding of storm surges, additional studies have integrated morphological characteristics, oceanic conditions, and temporal patterns into their analyses (Huang, 2022;Li et al, 2023;Lockwood et al, 2022;Rajabi-Kiasari et al, 2023). These aspects help to elucidate the dynamics of storm surges, including wind force, oceanic movements, and landform interactions.…”

A Comprehensive Review of Assessing Storm Surge Disasters: From Traditional Statistical Methods to Artificial Intelligence-Based Techniques

“…Zhang and Jiang employed four optimization algorithms (including genetic algorithm, particle swarm optimization, beetle antennae search, and beetle swarm optimization) to optimize Back Propagation Neural Networks, proposing four optimized BPNN for predicting storm surge disaster risk [35]. Lockwood et al found that ANN models trained on synthetic datasets could predict storm surge levels along the eastern United States and the Gulf of Mexico, providing scientific support for relevant departments to take preventative and disaster mitigation measures [36].…”