The 2026 Typhoon Season So Far (What the Official Record Shows)
Five named storms formed in the western Pacific between January and May 2026. The ten-year average for those months is 1.6.
That is a little over three times the normal rate, and it happened with an unusual regularity: exactly one named storm in each of January, February, March, April and May. Two of the last ten years — 2016 and 2025 — produced none at all in that window.
The early season, counted
| Season | Named storms, January–May |
|---|---|
| 2016 | 0 |
| 2017 | 1 |
| 2018 | 3 |
| 2019 | 2 |
| 2020 | 1 |
| 2021 | 3 |
| 2022 | 2 |
| 2023 | 2 |
| 2024 | 2 |
| 2025 | 0 |
| 2016–2025 average | 1.6 |
| 2026 | 5 (Nokaen, Penha, Nuri, Sinlaku, Hagupit — one per month) |
Counts computed for this page from the RSMC Tokyo best track archive, downloaded 26 September 2026. Method: every system reaching tropical-storm grade or higher, counted once, by the year and month in which it first reached that grade. The 2026 file is finalised only to 19 May 2026 — see the section below.
For scale, the same archive gives a ten-year average of 25.1 named storms per full season, with a range from 17 in 2023 to 29 in 2018 and 2019. Five before June is a fast start rather than an extraordinary total, and it says nothing reliable about where the season finishes.
Why the official record stops in May
This is the part that makes an honest month-by-month account of 2026 difficult, and it is worth explaining rather than working around.
The best track archive is not a live feed. It is the post-season reanalysis: after a storm is over, forecasters go back through all the available data — satellite, aircraft where there is any, surface observations, later research — and produce a definitive record of where the centre actually was and how strong it actually got. That record often differs from what was issued in real time, and it is the version that goes into the climatology.
Reanalysis takes months. As of this writing the 2026 file runs to 19 May. Everything after that — the whole northern summer, the part of the season that produced the storms most people actually heard about — is not in the official record yet.
So a page claiming to give you 2026's month-by-month totals through September would be claiming something the source does not support. What follows instead is a clear separation between the two.
What is not yet in the record
This site covered a number of systems through the northern summer of 2026, and none of them appear in the finalised archive above. Listing them as provisional is the honest way to present them.
| Period | Status | What this site reported at the time |
|---|---|---|
| Jan–May 2026 | Finalised | 5 named storms, one per month |
| Late Jul–Aug 2026 | Provisional | Dolphin, Chan-hom, Saudel and others covered here as they happened, from operational bulletins |
| Aug 2026 | Provisional | A stretch with four systems running at once in the western Pacific |
| Sep 2026 | Provisional | Not yet summarised |
The practical difference between the two rows is not small. Operational intensity estimates are revised in reanalysis, sometimes by a full category, and a storm's peak in the final record is frequently not the number that appeared in the headlines at the time. Anyone comparing this season with a previous one should compare finalised against finalised, and this season cannot be compared that way yet.
Why the count is of named storms specifically
Every number on this page counts systems that reached tropical-storm grade or higher, which is the threshold at which a system is given a name. That choice is worth explaining, because a different threshold produces a different season.
Counting every tropical depression inflates the total substantially and includes a great many systems that never affected anybody. Counting only typhoons deflates it and hides the tropical storms that do most of the flooding damage in the Philippines and Vietnam. The named-storm threshold sits between the two and has the practical advantage of being the point at which agencies start issuing the bulletins a traveller would actually read.
It also makes the figures comparable across years, which matters more than it sounds. Detection of weak systems has improved with satellite coverage, so a depression count from 2016 and one from 2026 are not measuring quite the same thing. The named-storm threshold is high enough to be largely immune to that drift.
Ten years of totals, for scale
Any statement about this season needs something to be measured against. Here is the full-season count for the last decade from the same archive.
| Season | Named storms (full year) |
|---|---|
| 2016 | 26 |
| 2017 | 27 |
| 2018 | 29 |
| 2019 | 29 |
| 2020 | 23 |
| 2021 | 22 |
| 2022 | 25 |
| 2023 | 17 |
| 2024 | 26 |
| 2025 | 27 |
| Average | 25.1 |
The range is 17 to 29 across ten years, which is narrower than most people assume. The western Pacific produces roughly two dozen named storms a year with fairly limited variation, and the years that are remembered as bad are not generally the years with the highest counts.
2023 is the clearest illustration. It produced 17 named storms, the lowest total in the decade by a margin, and it was not a quiet year on the ground. Where storms go matters far more than how many form, and a basin count captures none of that.
How to read a season count without being misled
| The claim | What to check |
|---|---|
| "This season is above average" | Above average to date, or projected? Five before June is a fast start, not a forecast. |
| "A record number of storms" | Counted how — named storms, typhoons, or all depressions? The three give very different totals. |
| "The strongest storm since…" | Operational or reanalysed intensity? They differ, and only one is comparable across years. |
| "Quiet season" | Storm count and damage are only loosely related. 2023 had 17 named storms and was not a quiet year for the Philippines. |
| Any 2026 month-by-month table through September | Ask where it came from. The official archive does not contain it yet. |
| A count for your own trip | Season totals are close to useless at that scale. Approaches to one place in one month is the number that matters. |
That last row is the one this site keeps returning to. A basin-wide total tells you almost nothing about a specific destination, which is why the regional pages here count approaches to a single reference point instead. The comparison across five regions makes the point directly: Taiwan's worst month is the Philippines' quietest.
One per month is the odd part
The total is unusual. The distribution is stranger.
Five named storms arriving as exactly one in each of January, February, March, April and May is not how the early western Pacific normally behaves. Off-season activity in this basin tends to cluster — two systems in a fortnight and then nothing for two months — because it depends on short-lived favourable windows rather than on a steady background state. A clean one-a-month sequence implies those windows kept reopening at regular intervals, which is not a pattern with an obvious cause.
It is worth resisting the urge to read meaning into it. Five is a small number, and small numbers produce tidy-looking sequences by chance more often than intuition suggests. The honest description is that 2026 had an unusually active and unusually evenly spread first five months, and that the second half of the sentence is probably coincidence.
What the sequence does confirm is that the basin was capable of producing storms continuously through what is normally its quiet season, and that is a statement about ocean and atmospheric conditions rather than about luck.
What an active early season does and does not predict
The honest answer is: not much, and it is worth saying so plainly because the temptation to extrapolate is strong.
Early-season activity in the western Pacific is driven substantially by conditions — ocean heat, the state of the El Niño and La Niña cycle, the position of the monsoon trough — that can change by the time the peak months arrive. A busy January to May tells you the basin was favourable in January to May. The correlation with the eventual seasonal total is weak enough that forecasters do not rely on it.
What it does affect is timing rather than volume. An early start can mean the season's activity is spread more widely across the year, which matters for places at the edges of the normal window — and that includes the Philippines in December, which this site has covered separately.
The one thing five early storms definitely does not mean is that the total will be five times higher. The 2026 total will land somewhere in the range the last decade has produced, and nothing observable in May narrowed that range much.
Current positions and five-day tracks are published by the Japan Meteorological Agency, and the finalised archive used for the counts above is the RSMC Tokyo best track. When this page and a live bulletin disagree, the bulletin is right.
Comments
Post a Comment