Japan’s headline typhoon statistic, the number that makes landfall in an average year, is 3.0. It cannot include Okinawa. The Meteorological Agency defines landfall as the centre of a typhoon reaching the coastline of Hokkaido, Honshu, Shikoku or Kyushu, and Okinawa appears in none of those four. A storm that crosses Naha at full strength is counted as an approach and never as a landfall, and the approach normal for Okinawa and Amami is the higher of the two figures in the agency’s split between the mainland and Okinawa with Amami.
Landfall is defined on four islands, and Okinawa is not one of them
The definition sits in a footnote under the agency’s table of typhoon normals, and it is exact: landfall means the case in which the centre of the typhoon reaches the coastline of Hokkaido, Honshu, Shikoku or Kyushu. Four names, no others, and no clause covering the smaller island chains.
That is a counting rule rather than a claim about where typhoons go, and the agency does not present it as anything else. But it changes what the landfall row can mean. The normal of 3.0 landfalls a year describes four islands, while the formation normal of 25.1 describes the whole northwest Pacific basin and the South China Sea, and the approach normal of 11.7 describes proximity to any Japanese weather station. Three rows of the same table are counted over three different geographies.
Approach is defined as coming within 300 km of a weather station
The footnote defining approach is as concrete: an approach means the case in which the centre of the typhoon comes within 300 km of any domestic meteorological observatory. Not within 300 km of the coast, and not within 300 km of a city; within 300 km of a station. Formation, the first of the three footnoted definitions, is counted on Coordinated Universal Time rather than on Japan time.
All three normals are averages over the thirty years from 1991 to 2020. The agency states that a blank month means no typhoon fell into that category at all during the thirty-year statistical period, which is a different statement from a printed 0.0.
| Month | Formations | Approaches | Landfalls |
|---|---|---|---|
| January | 0.3 | Not printed | Not printed |
| February | 0.3 | Not printed | Not printed |
| March | 0.3 | Not printed | Not printed |
| April | 0.6 | 0.2 | Not printed |
| May | 1.0 | 0.7 | 0.0 |
| June | 1.7 | 0.8 | 0.2 |
| July | 3.7 | 2.1 | 0.6 |
| August | 5.7 | 3.3 | 0.9 |
| September | 5.0 | 3.3 | 1.0 |
| October | 3.4 | 1.7 | 0.3 |
| November | 2.2 | 0.5 | Not printed |
| December | 1.0 | 0.1 | Not printed |
| Year | 25.1 | 11.7 | 3.0 |
Formations carry a figure in all twelve months, approaches in nine and landfalls in six. The peak month is not the same for all three: formations peak in August at 5.7, approaches tie between August and September at 3.3 each, and landfalls peak in September at 1.0. The agency prints two warnings about adding the columns up, and both are needed. Approaches can straddle two months, so the monthly figures need not sum to the annual one, and rounding means the same is true of the other rows. Our own addition gives 25.2 for formations against the printed 25.1, 12.7 for approaches against the printed 11.7, and 3.0 for landfalls, which matches.
The two regional approach figures add up to more than the national one
The agency splits approaches a second way, into the mainland and Okinawa with Amami, and defines each the same way as before. The mainland figure counts a typhoon that comes within 300 km of a station in Honshu, Hokkaido, Kyushu or Shikoku, and the normal is 5.8 a year. The Okinawa and Amami figure counts a typhoon that comes within 300 km of a station in the Okinawa region or the Amami region, and the normal is 7.9 a year.
Those two add to 13.7, which is 2.0 more than the national approach normal of 11.7. Both figures are the agency’s; the addition and the subtraction are ours. The page prints no note reconciling the two regional rows with the national row, and the definitions it does print allow one typhoon to satisfy both of them on the same pass. What can be said from the two numbers alone is that the stretch of Japan excluded from the landfall definition has the larger approach normal of the two.
The strength scale has three named grades and the size scale has two
The agency expresses a typhoon’s rough force with two separate labels, both built on ten-minute mean wind speeds. Strength is set by the maximum wind speed and has three grades. Size is set by the radius of the area where winds of 15 m/s or more blow or may blow, and has two.
| Label | Grade | Range as published |
|---|---|---|
| Strength | Strong | 33 m/s, 64 knots, or more and under 44 m/s, 85 knots |
| Strength | Very strong | 44 m/s, 85 knots, or more and under 54 m/s, 105 knots |
| Strength | Violent | 54 m/s, 105 knots, or more |
| Size | Large | Radius of 500 km or more and under 800 km |
| Size | Very large | Radius of 800 km or more |
Two radii matter and they are not the same thing. The area used for the size label is the one with winds of 15 m/s or more. The separate area where winds of 25 m/s or more blow or may blow is the storm-wind area, which the agency draws as a circle on its typhoon information and describes as the zone where such winds could arrive at any moment.
A typhoon can be announced with neither a size word nor a strength word
Because both scales start above zero, both labels can be absent. The agency states that it does not express size when the radius of the 15 m/s area is under 500 km, and does not express strength when the maximum wind speed is under 33 m/s. A storm whose 15 m/s radius is under 500 km and whose maximum wind speed is under 33 m/s is therefore a typhoon with no adjective attached to it at all.
The reverse reading is the useful one, and the agency spells it out. If the announcement says a strong typhoon and nothing more, that tells the reader three things at once: the radius of the 15 m/s area is under 500 km, the maximum wind speed is between 33 and 43 m/s, and the storm has a storm-wind area. The labels combine in the order size then strength, as in a large and strong typhoon.
Numbering restarts on 1 January, and a typhoon that re-forms keeps its old number
The agency numbers typhoons by order of formation from 1 January each year, so the first to form in the calendar year is No. 1. The second rule is the one that breaks the sequence: if a typhoon weakens into a tropical depression and then develops back into a typhoon, it is given the same number again rather than the next one.
The underlying definition is narrower than the word suggests. A tropical cyclone counts as a typhoon only if it is in the northwest Pacific, meaning north of the equator and west of 180 degrees east, or in the South China Sea, and only if the maximum wind speed within the low-pressure area, as a ten-minute mean, is roughly 17 m/s, 34 knots, or force 8 on the wind scale, or above. The same page notes that a typhoon draws its energy from heat released as water vapour from the warm sea surface condenses, that it would die out in two or three days if that supply stopped, and that it loses its character and becomes an extratropical cyclone when cold air flows in aloft.
There are 140 Asian names, and Japan supplied ten of them from constellations
Asian names replaced the English personal names the United States used to assign. The Typhoon Committee, an intergovernmental body for typhoon disaster management in the northwest Pacific and South China Sea with 14 member countries and territories including Japan, decided from 2000 to give typhoons in that region a common set of names drawn from the region itself. The first, given to typhoon No. 1 of 2000, was Damrey, proposed by Cambodia, meaning elephant.
The list runs to 140 names, used in order of formation before returning to Damrey. We counted the published table: it has 140 numbered rows, and each of the 14 members has contributed exactly ten. Because the members repeat in a fixed order, Japan’s ten names fall at every fourteenth position, at 5, 19, 33, 47, 61, 75, 89, 103, 117 and 131. They are Koinu, Tomo, Hebi, Kajiki, Koto, Kujira, Koguma, Tokei, Tokage and Yamaneko, and every one is a constellation. Yamaneko is the longest at eight letters, inside the committee’s limit of nine.
The agency gives its reasons for choosing constellations: they are neutral names, not those of a particular person, company, trademark, place or weather phenomenon; they are natural things, so conflicts of interest are less likely; they sit in the sky, which has an association with a storm; and they are familiar. The other conditions are that a name must be no more than nine alphabet letters, must not have too many syllables and must be easy to pronounce, and must not mean something offensive in another member’s language. With the formation normal at 25.1 a year, the agency notes that the list comes round in roughly five to six years, and it adds that a name may be withdrawn from future use at a member’s request after a typhoon that caused major disaster. A system that enters the basin from east of 180 degrees east or west of 100 degrees east keeps the name already given by the agency responsible for that area, so some typhoons carry names that are not on the list at all.
The English class ladder starts exactly where the definition and the strength grades start
In Japanese-language material the agency writes a typhoon as an era year plus a number. In English it uses the Asian name, and it may put a class in front of that name based on the maximum intensity reached during the typhoon’s life. There are three classes, and they are stated in knots.
- Tropical Storm, TS: maximum wind speed 34 knots or more and under 48 knots.
- Severe Tropical Storm, STS: 48 knots or more and under 64 knots.
- Typhoon, T or TY: 64 knots or more.
Read against the agency’s other two definition pages, the two lower boundaries are not arbitrary. The bottom of the Tropical Storm band, 34 knots, is the figure the agency prints in brackets beside 17 m/s as the threshold at which a tropical cyclone becomes a typhoon at all. The bottom of the Typhoon band, 64 knots, is the figure it prints in brackets beside 33 m/s as the bottom of the weakest named strength grade. So the English word Typhoon begins where the Japanese label strong begins, and the whole English ladder sits between the definition and that first grade.
Because names repeat, the agency also allows a year and number to be written after the name to pin down a particular storm. Its own worked example is typhoon No. 1 of 2000, whose class was Typhoon, written as Typhoon Damrey, TY Damrey (0001) or Damrey (T0001). The agency publishes a table of typhoons and their classes covering 1977 to 2025 as a CSV file, and notes that it corrected some names in that file on 30 March 2022.
Sources Checked
Japan Meteorological Agency — typhoon normals
Confirmed that the normals are thirty-year averages for 1991 to 2020, that a blank month means no typhoon of that category occurred in the statistical period, that formation is counted on Coordinated Universal Time, that approach means the centre coming within 300 km of any domestic meteorological observatory, that landfall means the centre reaching the coastline of Hokkaido, Honshu, Shikoku or Kyushu, the monthly and annual figures of 0.3, 0.3, 0.3, 0.6, 1.0, 1.7, 3.7, 5.7, 5.0, 3.4, 2.2 and 1.0 with 25.1 for the year for formations, blank, blank, blank, 0.2, 0.7, 0.8, 2.1, 3.3, 3.3, 1.7, 0.5 and 0.1 with 11.7 for approaches, and blank, blank, blank, blank, 0.0, 0.2, 0.6, 0.9, 1.0, 0.3, blank and blank with 3.0 for landfalls, the notes that approaches can straddle two months and that rounding means monthly figures need not sum to the annual figure, the mainland approach normal of 5.8 defined as within 300 km of a station in Honshu, Hokkaido, Kyushu or Shikoku, and the Okinawa and Amami approach normal of 7.9 defined as within 300 km of a station in the Okinawa or Amami regions.
https://www.data.jma.go.jp/yoho/typhoon/statistics/average/average.html
Japan Meteorological Agency — what a typhoon is
Confirmed that a typhoon is a tropical cyclone in the northwest Pacific north of the equator and west of 180 degrees east, or in the South China Sea, with a maximum wind speed as a ten-minute mean of roughly 17 m/s, 34 knots or force 8 or above, and the statements that it develops on heat released by condensing water vapour from the warm sea surface, would disappear in two or three days without that supply, and changes into an extratropical cyclone when cold air flows in aloft.
https://www.jma.go.jp/jma/kishou/know/typhoon/1-1.html
Japan Meteorological Agency — typhoon size and strength
Confirmed the three strength grades of 33 m/s or 64 knots to under 44 m/s or 85 knots, 44 m/s or 85 knots to under 54 m/s or 105 knots, and 54 m/s or 105 knots and above; the two size grades of a 15 m/s radius of 500 km to under 800 km and of 800 km and above; that the size label uses the radius of the area with winds of 15 m/s or more and the storm-wind area is the area with winds of 25 m/s or more; that size is not expressed when the radius is under 500 km and strength is not expressed when the maximum wind speed is under 33 m/s; and the worked reading of a strong typhoon as a radius under 500 km, a maximum wind of 33 to 43 m/s and the presence of a storm-wind area.
https://www.jma.go.jp/jma/kishou/know/typhoon/1-3.html
Japan Meteorological Agency — how typhoon numbers and Asian names are assigned
Confirmed that numbering runs by order of formation from 1 January each year with No. 1 the earliest, that a typhoon which weakens to a tropical depression and redevelops keeps the same number, that the Typhoon Committee has 14 member countries and territories including Japan and decided from 2000 to use a common set of regional names, that 140 names are used in order before returning to Damrey, that Damrey was given to typhoon No. 1 of 2000 and was proposed by Cambodia meaning elephant, that Japan proposed ten names derived from constellations and the stated reasons for that choice, the conditions of no more than nine alphabet letters, few syllables and no offensive meaning in another member’s language, that the annual formation normal of 25.1 means the list comes round in roughly five to six years, that a name may be withdrawn from future use at a member’s request after a major disaster, that a system entering from east of 180 degrees east or west of 100 degrees east keeps its existing name so some typhoons carry names not in the table, the three English classes of Tropical Storm at 34 to under 48 knots, Severe Tropical Storm at 48 to under 64 knots and Typhoon at 64 knots or more, the worked example of Typhoon Damrey, TY Damrey (0001) and Damrey (T0001), and the CSV table of typhoon classes for 1977 to 2025 with its name corrections dated 30 March 2022. We also counted the published name table and found 140 numbered rows with exactly ten from each of the 14 members, Japan’s at positions 5, 19, 33, 47, 61, 75, 89, 103, 117 and 131, named Koinu, Tomo, Hebi, Kajiki, Koto, Kujira, Koguma, Tokei, Tokage and Yamaneko.
https://www.jma.go.jp/jma/kishou/know/typhoon/1-5.html
Last updated: September 2026
