Storm Extremes by Decade

The counts use the IBTrACS best track archive, named storms from the 1950 season to 2027. A storm is counted in the decade of its season and in the basin where its track first appears, so a storm that moves into another basin is counted once. A storm reaches hurricane strength at a peak wind of 64 kt (119 km/h, 74 mph), major strength at 96 kt (178 km/h, 110 mph) and category 5 at 137 kt (254 km/h, 158 mph), in every basin whatever the local name. 137 kt is where the Saffir-Simpson scale puts category 5. Most agencies report wind in steps of 5 kt, so a storm at 135 kt is category 4 here and one at 140 kt is category 5.

Wind estimates before the satellite era, about 1970, are unreliable: they came from ships, aircraft, land stations and early satellite pictures, and storms away from shipping lanes were missed or under-measured. Counts of weaker storms are incomplete before the satellite era. The strength statistics therefore start with the 1980 season, when satellite coverage and the Dvorak technique (strength estimated from cloud patterns) were in routine use in every basin. The earlier decades appear only as counts of storms, in the last section, and are not comparable with later ones. Methods kept changing after 1980, so part of any trend in these counts is a trend in observation, and the newest seasons are provisional.

Agencies average wind over different periods. The United States agencies (National Hurricane Center, Central Pacific Hurricane Center and Joint Typhoon Warning Center) use the highest 1 minute average. The Japan Meteorological Agency, the Australian Bureau of Meteorology, Météo-France and the Fiji Meteorological Service use 10 minute averages, and the India Meteorological Department uses 3 minute averages. For the same storm a 10 minute wind is lower than a 1 minute wind, by roughly a tenth. For each point this site takes the wind in the archive's WMO fields (the agency responsible for the basin) where there is one and the United States value otherwise, so winds of both kinds are mixed in some basins and storms, and a change of agency inside a storm can show up as a jump. Counts at 64, 96 and 137 kt are not strictly comparable between basins.

Only named storms are included. Naming practice differs by basin and has changed over time: the North Indian basin has named cyclones only since 2004, so its earlier decades hold few or no storms by construction. A share is shown only where a decade and basin has at least 10 storms with a known peak wind, and otherwise reads n/a. Values outside 10 to 200 kt for wind or 850 to 1,020 hPa for pressure are treated as errors.

For season by season totals and accumulated cyclone energy see storm seasons; every storm has its own page under storms.

Storms in each decade whose peak wind reached the threshold, by the basin where they formed. The last row is the world, and a decade still running is marked with its last season.

Storms at 96 kt or more as a percentage of the named storms with a known peak wind. Where naming is selective or new, the denominator is small or biased toward stronger storms, so read the basins with few storms (n/a means fewer than 10) as a count, not a rate.

A storm went through rapid intensification when its wind rose by at least 30 kt between two six hourly points exactly 24 hours apart (00, 06, 12 and 18 UTC) in its best track. All four six hourly steps have to be present and the wind at both ends known; a missing wind in between does not matter, a gap in the track does. A storm is counted once however many times it intensified. Winds are reported in steps of 5 kt and a change of agency inside a storm can create a jump, so small differences between basins are not meaningful.

Major storms with at least one computable 24 hour wind change are the base. n/a means fewer than 5 such storms.

The 15 lowest minimum central pressures in the archive. Pressure measurement quality varies by era and agency. A central pressure is measured directly only when an aircraft, ship or land station samples the centre: aircraft reconnaissance flew routinely in the North Atlantic and, until 1987, in the western North Pacific. Elsewhere and later most values are estimates from satellite pictures or from a wind and pressure relationship, and agencies use different methods, so pressures from different basins and eras are not equally certain. The agencies column lists the sources the archive used for the storm's wind and pressure.

For each decade, the storm with the highest peak wind (ties go to the lower pressure) and the storm with the lowest central pressure. Decades before 1980 are shown for completeness and rest on less reliable estimates.

Each cell counts the named storms that had at least one recorded position inside it during the decade, at any stage including weak and post-tropical ones. A 5 degree cell is about 555 km (345 miles) tall. A fast storm can pass between two six hourly positions without a position in a cell, so cells crossed quickly may read slightly low. All maps use the same colour scale, so they can be compared, and they show full decades only. Counts of weaker storms are incomplete before the satellite era, which is why the maps start in the 1980 season.

Named storms with a known peak wind in the decades before the strength statistics start. Wind estimates before about 1970 are unreliable and counts of weaker storms are incomplete before the satellite era, so these numbers must not be set against the later decades.