Philippines Late Typhoon Season (October, November, December)
Haima and Nock-ten. Khanun and Tembin. Kammuri and Phanfone. Molave, Goni and Vamco. Kompasu and Rai. Nalgae. Yinxing. Fengshen and Kalmaegi.
Fifteen storms, and every one of them came within 100 km of a major Philippine city in October, November or December. That is the last ten seasons, counted from the Japan Meteorological Agency's own track archive, and it is the part of the Philippine year that almost every English-language travel guide treats as the tail end of something.
The season does not wind down. It relocates
The common picture of the western Pacific typhoon season has it building through the summer, peaking in August, and fading through autumn. For Japan, Korea, Taiwan and Okinawa that picture is roughly right. For the Philippines it is wrong in a way that puts people in the country at exactly the wrong time.
The reason is steering, not storm production. Systems form in much the same part of the ocean all season. What changes is the pattern that carries them west. Through the early season the subtropical ridge sits in a position that pushes storms north of the archipelago and on toward the Ryukyus and Japan. As the ridge shifts south through autumn, the same conveyor delivers storms into Luzon and the Visayas instead of above them.
That is why the Philippine curve is shaped differently from its neighbours'. This site has counted the same thing for Jeju, Naha, Taipei and Hong Kong, and all four peak between July and September. Manila's two worst months are October and November. There is no single Asian typhoon season, and the Philippines is the clearest proof of it.
Fifteen storms in ten Octobers, Novembers and Decembers
Here is the count. Each row is one season, and each storm is named as the JMA archive names it.
| Season | Storms | Which ones, and when |
|---|---|---|
| 2016 | 2 | Haima (Oct), Nock-ten (Dec) |
| 2017 | 2 | Khanun (Oct), Tembin (Dec) |
| 2018 | 0 | — |
| 2019 | 2 | Kammuri (Dec), Phanfone (Dec) |
| 2020 | 3 | Molave (Oct), Goni (Nov), Vamco (Nov) |
| 2021 | 2 | Kompasu (Oct), Rai (Dec) |
| 2022 | 1 | Nalgae (Oct) |
| 2023 | 0 | — |
| 2024 | 1 | Yinxing (Nov) |
| 2025 | 2 | Fengshen (Oct), Kalmaegi (Nov) |
| Total | 15 | October 6 · November 4 · December 5 |
Counts computed for this page from the RSMC Tokyo best track archive (Japan Meteorological Agency), downloaded 26 September 2026. Method: every storm of tropical-storm grade or higher whose best-track position came within 100 km of one of six Philippine reference points (Laoag, Manila, Legazpi, Tacloban, Cebu, Davao), counted once per storm, for the ten complete seasons 2016–2025. This measures close approach, not landfall — landfall cannot be determined from track positions alone.
Three things in that table are worth pulling out.
The average is 1.5 a season, and the average is misleading. Two of the ten seasons had none at all. One had three in six weeks. A traveller is not exposed to 1.5 storms; they are exposed to a coin-flip that occasionally lands on three.
December is not the safe month. Five of the fifteen arrived in December, which is more than November. The idea that the season is finished by the end of November is not supported by the record.
The worst single stretch was 2020. Molave in late October, then Goni and Vamco within a fortnight of each other in November. Any plan built on the assumption that one storm clears the way for a settled week is built on the wrong assumption.
What is actually different about a late-season storm
It is not only the timing. Late-season systems in the Philippines tend to arrive with characteristics that change what breaks.
They come in on a more westerly track and often cross the archipelago rather than clipping the north, which means the exposed side is the eastern seaboard — Bicol, Samar, Leyte — rather than the Batanes and Cagayan coast that takes the summer systems. Those are the regions with the least alternative transport when a road or a ferry goes.
They also arrive on ground that has already taken the southwest monsoon. By late October the catchments feeding Marikina, Pampanga and the Cagayan valley have absorbed months of rain, and the rainfall totals that produce serious flooding are lower than they would be in June. The same storm does more damage in November than in July, which is the part the wind signal never tells you. This site has a longer piece on how the rainfall colours work and why they are local, and it applies with particular force at this end of the year.
And they arrive into shorter daylight and a busier travel calendar. Late November and December carry both the tail of the season and the start of the peak holiday period, which is the worst combination for rebooking: the storms are still coming and the aircraft are full.
There is a fourth difference that matters for anyone moving between islands. The inter-island ferry network is the transport layer with the least redundancy in the country, and it suspends on sea state rather than on wind at the pier. Swell generated by a system that is still hundreds of kilometres away will stop sailings before any signal is raised locally, and it keeps them stopped for a day or two after the wind has gone. In the late season, when one system can follow another within a fortnight, those suspension windows start to overlap. A traveller who has planned a route that depends on two separate ferry crossings is not exposed to one storm; they are exposed to the gap between two.
What the late season looks like from inside it
I lived in the Philippines for seven years, and the thing that took longest to learn was that the late-season storms are socially different from the summer ones, not just meteorologically different.
By October people have been through several rounds of it. The preparation is faster and quieter, the announcements are read more carefully, and the questions people ask each other change from "is it coming here" to "which route is still open". Shops restock differently. The early-season habit of buying everything at once has worn off, and the actual scarce items are narrower and more predictable than a first-time visitor expects.
The other difference is the calendar. A storm in the second week of December lands on a country that is already moving — people travelling to family, flights full, buses full. The storm itself may be no worse than an August system, and the disruption is worse because there is no slack anywhere in the network to absorb it.
None of that shows up in a track map, and it is the part that decides how a week actually goes.
Deciding about an October to December trip
The record does not say do not go. Eight of the ten seasons had two storms or fewer across three whole months, and most visits in that window pass without incident. What the record says is that the risk is not zero and is not falling through November, so the trip should be built to absorb one.
| If this is true | Then |
|---|---|
| A signal is raised over your province and your flight is 48 hours out | Rebook now, before the waiver queue forms. Airline waivers open early and close when the aircraft fill. |
| A signal is raised and you are due to take a ferry between islands | Assume it is gone for two days after the wind drops. Sea state outlasts the storm. |
| No signal anywhere, but an orange or red rainfall warning over your city | Do not move around the city. This is the case that does most of the damage and raises no signal at all. |
| You have a same-day international connection through Manila or Cebu | Add a buffer night. This single change removes most of the real risk in a late-season trip. |
| Your itinerary is Bicol, Samar or Leyte in November | Keep the last two days unbooked. That is the exposed side on a late-season track. |
| It is December and you assumed the season was over | It is not. Five of the last fifteen late storms arrived in December. |
The short version
October, November and December are the Philippine typhoon season, not its aftermath. Fifteen storms in ten years came within 100 km of a major city in those three months, December included. The exposure is real but modest in any single season, and it is almost entirely manageable by keeping one buffer day and treating the rainfall warning as the dangerous one.
The official forecasts and signals are published by PAGASA, and the regional track guidance used for the counts above comes from the Japan Meteorological Agency. Those are the authorities; this page is background.
Comments
Post a Comment