A Number Doesn't Tell You Everything
Hazards & Forecasts
Category Is Not the Whole Story
The Saffir-Simpson scale is the single most-quoted number in every hurricane story -- and it measures exactly one thing: wind. Everything else that makes a hurricane dangerous is a completely separate question.
Wind is what the category number actually measures -- sustained wind speed, nothing more. It's the easiest hazard to put a clean number on, which is probably why it became the public-facing scale, but it was never meant to describe the whole storm.
Storm surge is seawater pushed inland by the storm's wind and low pressure, and it's historically been the deadliest part of a hurricane in the United States -- not the wind. Surge depends on the storm's size, its forward speed, the angle it approaches the coast, and the shape of the seafloor just offshore, which means two storms with the same category can produce wildly different surge. You'll see exactly that in the case comparisons below.
Waves ride on top of surge and add their own battering force, especially along exposed coastlines and barrier islands.
Rainfall flooding has almost nothing to do with category and almost everything to do with forward speed -- a slow-moving Category 1 can dump far more rain on one spot than a fast-moving Category 4 ever does.
Tornadoes frequently spin up in a hurricane's outer rainbands, often far from the eyewall and sometimes with little warning, in a part of the storm most people have already stopped worrying about.
A "minor" category number is never a reason to stop paying attention to the rest of this list.
Surge Case Comparisons
Here's the proof that category and surge are two different questions: real Florida storms, similarly rated, with very different water.
Every entry here comes straight from an NHC Tropical Cyclone Report or official preliminary summary -- no estimates, no modeling, just what was actually measured and published for that storm. Charley (2004) is the case worth remembering: a Category 4 at landfall, but only about 7 feet of surge, because it was small and moving fast. Compare that to storms with a full category lower that pushed far more water inland, and the point makes itself.
This is landfall-zone specific, not a full coastline survey, and it only covers storms someone has gone back and sourced a real figure for. Storms without a verified number simply aren't shown, on purpose, rather than filled in with a guess. Surge datums differ by era and by source -- treat comparisons across very different time periods with some caution, the same way Pressure Lens asks you to when comparing storms across its era splits.
Cone Decoder
You've seen the cone on every hurricane graphic NOAA puts out. Here's the part almost nobody explains clearly enough: being outside the cone does not mean you're safe. In fact, for most storms, it's normal -- not the exception -- for dangerous conditions to reach well beyond the cone's edges.
The cone shows the forecast track's typical error -- a zone built from how far off NHC's forecasts have historically been at each lead time, centered on where the storm's center is expected to go. It is not the size of the storm. It is not an impact zone. It has never been either of those things, even though it gets treated that way constantly.
A hurricane's wind field, rain bands, and storm surge routinely extend far outside the cone -- sometimes hundreds of miles beyond it. Tropical-storm-force wind alone often reaches well past the cone's edge on both sides, and heavy rainfall and isolated tornadoes can occur even farther out, in bands that have nothing to do with where the center ends up. Standing outside the cone tells you almost nothing about whether you'll feel real effects. It only tells you that the center probably won't pass directly over you.
This is exactly why watches, warnings, and local emergency guidance are built around a much wider area than the cone, and why those are the products to actually act on -- not the cone shape itself. If you're near the cone at all, plan as if you're in it. That's not overcaution; historically, it's closer to the norm than the exception.
Track uncertainty and intensity uncertainty are two separate forecasts running side by side -- a track can be pinned down with real confidence while the intensity forecast for that same storm is still wide open, or the reverse. Neither one shrinks how far outward real effects can reach.
How Do We Know What Old Hurricanes Did?
A hurricane in 1875 and a hurricane in 2024 were tracked by completely different means -- and that difference shapes every chart on this entire site, not just this page.
Before satellites, the only way to know a hurricane existed at all was if a ship happened to sail through it, or it happened to pass near a coastal station with someone there to record it. That means storms that stayed out at sea and never crossed a shipping lane or a coastline could go completely unrecorded -- an older storm being "smaller" or "shorter-lived" in the record sometimes reflects the observing system, not the actual storm.
Aircraft reconnaissance -- flying directly into a storm to measure it -- started in the 1940s and dramatically improved both position and intensity estimates. Radar added detail close to the coast starting mid-century. Full-time satellite coverage, arriving in the 1960s and steadily improving since, is the reason the modern record is far more complete and far more confident than anything before it -- the chart above shows exactly that climb, from barely any pressure readings in the 1850s to complete coverage by 1980.
Where the record has real gaps, NOAA periodically runs a reanalysis -- going back through old ship logs, station records, and other historical sources with modern methods to correct and fill in the best-track data. That's why this site treats even "final" historical data as provisional until an official reanalysis replaces it, and never treats a lack of live monitoring as a reason to call the historical record itself stale.
This is exactly why Track Map's era/confidence coloring and Zone Lab's insufficient-sample hatching exist -- not as decoration, but as an honest map of where the historical record is strong and where it's thin. See the methodology page for the full picture.