Hurricane Open Dataset

Historical Destructive Power Scores and full intensity, size, and damage metadata for major tropical cyclones — from Bhola (1970) through the current Atlantic season. Free to download, free to redistribute, CC BY 4.0.

Downloads

CSV Historical Hurricane Database ~200 storms · 14 columns · < 5 KB Download CSV
JSON Historical Hurricane Database ~200 storms · structured array · < 10 KB Download JSON

What's in the file

Each row is one named tropical cyclone. The columns are intentionally compact and human-readable; this is a curated reference dataset, not a full track archive (use IBTrACS for full 6-hourly tracks).

ColumnDescriptionExample
basinOriginating ocean basinAtlantic, West Pacific, North Indian, …
nameStorm name as assigned by the WMO regional centerKatrina, Haiyan, Mangkhut
yearYear of peak intensity2005
damage_billionsDamage in inflation-adjusted USD (B)125.0
peak_wind_mphPeak 1-min sustained wind175
central_pressure_mbMinimum central pressure902
rmw_nmRadius of maximum winds at peak14
r34_nmTropical-storm-force wind radius (R34)60
duration_daysDays from genesis to dissipation4.5
duration_factorDPS duration component (0–1.5)1.2
rapid_intensification_24h_mphPeak 24-hour intensification50
landfallsComma-separated landfall regionsLouisiana, Mississippi
notesFree-text storm description"Unprecedented storm surge in New Orleans, …"
validation_targetWhether this storm is in the DPS calibration setTrue

How to cite

If you use this dataset in academic, journalistic, or commercial work, please cite it using the permanent Zenodo DOI:

doi.org/10.5281/zenodo.20149123

APA

Reaves, R. (2026). StormDPS Historical Hurricane Database (Version 1.0.0) [Data set].
Zenodo. https://doi.org/10.5281/zenodo.20149123

BibTeX

@dataset{reaves_2026_stormdps,
  author    = {Reaves, Ryan},
  title     = {StormDPS Historical Hurricane Database},
  year      = 2026,
  month     = may,
  publisher = {Zenodo},
  version   = {1.0.0},
  doi       = {10.5281/zenodo.20149123},
  url       = {https://doi.org/10.5281/zenodo.20149123}
}

Quick start (Python)

import pandas as pd

df = pd.read_csv("https://stormdps.com/historical_storms_db.csv")
print(df.sort_values("damage_billions", ascending=False).head(10))

Live API access

For real-time DPS values on active and historic storms, the StormDPS REST API exposes:

Public, free, no key required. Rate-limited per IP. See the methodology page for component definitions and the live tracker for interactive exploration.

License: CC BY 4.0 You may share, adapt, and redistribute this dataset — including for commercial purposes — provided you cite StormDPS as the source. The underlying observational records (NHC, JTWC, IBTrACS) are in the public domain.