Every tropical cyclone in NOAA's HURDAT2 record, with wind speeds every six hours. Pick a way to measure severity, choose the years, and see what the record says.
Hover a year for details; click one to show only that year on the globe.
Each year10-year averageTrend
Read before drawing conclusions
The early record misses storms. Before aircraft reconnaissance (1944) and especially before satellites (mid-1960s), weak or short-lived storms far out at sea often went unrecorded, and peak winds were estimated from ships and coastal stations. Counts before 1966 are probably too low, which can make any long-run "increase" look bigger than it is.
The Atlantic has natural busy and quiet spells. Ocean temperature patterns swing over decades: roughly 1970–1994 was quiet and 1995 onward busy. Comparing those two stretches can make a long-term trend look stronger than it is, so try different year ranges.
What the science says. The IPCC's 2021 assessment found it likely that the share of tropical cyclones reaching Category 3–5 has increased worldwide over the last four decades, while trends in the total number of storms are not clear. The "share that became major" measure is the closest match to that question here.
How these numbers are made. Counts use each storm's peak sustained wind (1-minute, in knots; 1 kt ≈ 1.15 mph). ACE adds up the square of the wind at each six-hourly fix (00, 06, 12, 18 UTC) while the storm was a tropical storm or hurricane at 35 kt or more, divided by 10,000. Trend lines are ordinary least-squares fits over the years selected. Data: NOAA National Hurricane Center HURDAT2 (file hurdat2-1851-2025-092326), 1851–2025. Older storms are periodically reanalyzed, so values can change between releases.