Tropical Cyclone-Driven Storm Surge and Wave Database for the US Coastlines

This dataset accompanies the paper "Tropical cyclone-driven storm surge and wave database for the US North Atlantic and Gulf coastlines" submitted to Nature Scientific Data. It contains storm surge and wave modeling results from coupled ADCIRC+SWAN simulations covering major tropical cyclones impacting the U.S. coastline from 1981–2021. The database provides spatially distributed time series of water surface elevation, significant wave height, and peak wave period at coastal locations vulnerable to hurricane impacts. Data Organization: Organized by decade; four netCDF files (e.g., AdcSwan_StormSurge_Data_1981_1990.nc)CF-compliant netCDF-4 format with compressionEach file contains multiple storm events with node-based extraction points Key Variables: Water surface elevation (m)Significant wave height (m)Peak wave period (s)Node coordinates (longitude, latitude) Time Representation:Time is provided in ISO-8601 format (time_str: YYYY-MM-DDTHH:MM:SS) for each storm. Spatial Coverage:Extraction points cover hurricane-vulnerable regions including the Mid-Atlantic, West Florida, Louisiana, and Texas coasts. File Structure:Each decade archive contains 3D arrays dimensioned as (node, time, subdir), where subdir indexes individual storm events. Storm identification is provided via subdir_names. All variables include CF-compliant metadata (long_name, units, standard_name). Intended Use:This dataset supports coastal hazard assessment, risk analysis, and validation of storm surge models. For detailed methodology, validation, and use cases, please refer to the accompanying Nature Scientific Data paper. Citation:Deb et al. (2025). Tropical cyclone-driven storm surge and wave database for the US North Atlantic and Gulf coastlines. Nature Scientific Data (submitted). Dataset: 10.5281/zenodo.17905131 Quick example: access with Python/xarray We provide a spatial visualization tool, extract_storm_spatial_map.py, that generates publication-quality maps showing the maximum water surface elevation across the entire computational domain for individual storm events. Users specify a storm identifier (e.g., bal122005 for Hurricane Katrina) and optionally a NetCDF file path. The script then produces a 300-DPI georeferenced map of peak storm surge values (see Figure bal122005_max_zeta_map.png above). We also provide an example script (extract_node_timeseries.py) for opening a decade file and extracting peak values of the variables at the node nearest a target lon/lat. A companion plotting script (extract_timeseries_all_files.py) extracts the time series at the same user-selected location and generates a three-panel figure showing the temporal evolution of water surface elevation, significant wave height, and peak wave period across all storm events in a given decade file.