Open data · cyclones & rainfall · data to July 2026 · last checked 5 October 2026
Andaman cyclone season: what 35 years of storm data show
Guests ask us two weather questions above all others: will a cyclone hit my trip, and will it rain the whole time. Both usually get answered with folklore, so we went to the archives instead — every tropical system recorded in the North Indian Ocean since 1991, measured against Port Blair, plus ten years of readings from the Port Blair airport rain gauge. The answer to the first question surprised us.
The finding
In 35 years, not one tropical cyclone has passed within 300 km of Port Blair during June, July, August or September.
The monsoon — the season travellers are most nervous about — is the only part of the Andaman year with no cyclone record at all. The storms come in October, November and December, with a smaller peak in April and May: the months that look best in a brochure. November alone accounts for 12 of the 34 systems.
The cyclone year, month by month
Every tropical system that passed within 300 km of Port Blair between 1991 and 2025 — 34 in 35 seasons, about one a year. The darker bar counts the closer passes, within 100 km (10 in total).
within 300 km within 100 km · 300 km is our own cut-off, not an official one — the 100 km series is shown so you can see how much the picture depends on where the line is drawn. It does not change the shape: the monsoon is empty either way.
The last five years, storm by storm
Every system that came within 500 km of Port Blair since 2021. Two came genuinely close: Midhili at 28 km in November 2023, and an unnamed March system at 82 km.
| Storm | Date | Closest approach | Peak intensity |
|---|---|---|---|
| Unnamed | 2 Apr 2021 | 396 km | — |
| Jawad | 2 Dec 2021 | 308 km | 35 kt |
| Unnamed | 20 Mar 2022 | 82 km | — |
| Asani | 7 May 2022 | 296 km | 65 kt |
| Sitrang | 21 Oct 2022 | 125 km | 45 kt |
| Mandous | 5 Dec 2022 | 421 km | 55 kt |
| Mocha | 8 May 2023 | 460 km | 145 kt |
| Hamoon | 20 Oct 2023 | 457 km | 90 kt |
| Midhili | 13 Nov 2023 | 28 km | 40 kt |
| Dana | 21 Oct 2024 | 394 km | 65 kt |
In the storm summary we generated on 28 July 2026, the 2025 season has no track within 500 km. The latest closest approach of any system in that summary is dated 30 November 2025, and we have not checked the archive for anything later. Systems did form in the North Indian Ocean that season — at least seven — and none of their tracks came near the islands. Going further back, the closest approaches on record are Lehar at 14 km (November 2013), an unnamed system at 16 km (November 1998) and Gaja at 26 km (November 2018).
What a nearby system actually means for your trip
This is the part the distances above do not tell you: the disruption radius is far wider than the damage radius. A cyclone tracking 200 km away can leave Port Blair itself completely untouched — no damage, barely a bad day ashore — and still put enough swell into the channels to stop inter-island ferries. So the realistic question for a traveller is not "will a storm hit my hotel", it is "could a storm several hundred kilometres away cost me a crossing".
Distance alone does not tell you what the sea will do. Our sister site's Andaman sea conditions page has month-by-month wave data for these waters, from NOAA's WAVEWATCH III model for 2015–2024, and gives the same advice: do not read a cancellation, or a safe crossing, off a storm's distance on a map.
How much notice you actually get: in practice, ferry operators here suspend sailings about one to two days before a system arrives, usually once the India Meteorological Department issues a red alert — the highest level in its four-step green, yellow, orange, red warning scale. That is our operational experience, and the track data is consistent with it. For the 14 systems that were tracked from more than 500 km out before coming within 300 km of Port Blair, the gap between the last archived track position outside that 500 km ring and the closest approach was typically about a day and a half — the upper of the two middle values — and under a day for the shortest. The archive stores positions a few hours apart, some of them interpolated, so treat these as rough. How long a suspension then lasts depends on the storm, and we have no count of that.
For current warnings go to the source, not to us: the India Meteorological Department's cyclone information page and its regional cyclone centre in New Delhi. We do not forecast.
When a system is tracking near the islands we contact affected guests, rework the itinerary around the sea state, and rebook crossings. It is also why our example itineraries keep the last night in Port Blair before the flight home. Our ferry cancellations study counts how often travellers cancel their own ferry bookings; it cannot count sailings pulled by an operator, which our booking desk says come mostly in the off season and end in a seat on another sailing.
And the other question: how much does it actually rain?
Rain days per 30-day month as recorded by the Port Blair airport gauge over ten years (2015–24), set against NASA POWER, a gridded reanalysis, for the same years and with the same threshold: a rain day is one with 1 mm or more. A reanalysis is a retrospective estimate for a grid cell. It is not a forecast and it is not a gauge.
| Month | Rain days (gauge) | Dry days (gauge) | NASA POWER (gridded) |
|---|---|---|---|
| Jan | 4 | 26 | 8 |
| Feb | 2 | 28 | 2 |
| Mar | 2 | 28 | 3 |
| Apr | 4 | 26 | 9 |
| May | 16 | 14 | 23 |
| Jun | 20 | 10 | 26 |
| Jul | 20 | 10 | 28 |
| Aug | 21 | 9 | 26 |
| Sep | 21 | 9 | 27 |
| Oct | 17 | 13 | 24 |
| Nov | 14 | 16 | 20 |
| Dec | 9 | 21 | 13 |
The gap is widest in the wet months. For July the gauge records about 20 rain days; the gridded dataset gives 28. Across June to September it gives about 30 percent more rain days than the gauge. That is the difference between ten dry days in July and two. We have not established why the two differ. The dataset describes a grid cell roughly 55 by 70 km that takes in open sea as well as the island, while the gauge measures what fell at one point in Port Blair. This is one dataset set against one gauge; we have not checked any other gridded product, or the data behind any forecast service, and make no claim about them.
One thing the daily totals cannot show, which we will state as operators rather than as data: monsoon rain here tends to arrive in heavy bursts and pass, rather than settling in for the day. A rain day is rarely a lost day. What the monsoon reliably costs you is sea time, not beach time — see our month-by-month weather guide for how that plays out across the year.
Sources and method
Cyclone tracks are from IBTrACS v04r01, the international best-track archive maintained by NOAA's National Centers for Environmental Information, North Indian Ocean basin. We computed the great-circle distance from every archived track position to Port Blair and kept each system's closest approach. Rainfall is NOAA GSOD daily records from the Port Blair airport station (2015–24), normalised to a 30-day month because the station does not report every day. The comparison column is NASA POWER daily precipitation (parameter PRECTOTCORR, from the MERRA-2 reanalysis) for the grid point at Port Blair airport, 2015–24, counted with the same 1 mm threshold. All three are published by United States government agencies as open data. Every figure on this page is generated by scripts we keep in the repository, so the annual update is a re-run rather than a re-typing. The warning-window figure is computed the same way: for the 14 systems with a position more than 500 km out before a closest approach within 300 km, the time from the last archived position outside the 500 km ring to the closest approach. The figure quoted is the upper of the two middle values of those 14 gaps, not an interpolated crossing time. The alert-scale description is India Meteorological Department's own published system — we describe how it works, we do not interpret or forecast alerts.
Three limits worth stating plainly. The storm summary was generated on 28 July 2026. The latest closest approach in it is dated 30 November 2025, and we have not checked the archive for storms after that. The 300 km radius is our own choice, which is why the 100 km series sits beside it. And we have deliberately published no climate-trend claims here — we did compute long-run changes in temperature and rainfall, but a tour operator comparing two blocks of years is not a trend analysis, and we would rather show you nothing than dress up arithmetic as climate science.
Using these figures
You are welcome to quote any number on this page — in an article, a guide, a dive-school briefing or a hotel's travel-advice section — with a link back so readers can check the workings. No permission needed and nothing to sign. The underlying archives are public and we would rather the analysis was used than admired. If you are writing about this and want the month-by-month figures in a usable form, or a sanity check on a claim before you publish it, just ask.
Suggested credit: Tropical Andamans analysis of NOAA IBTrACS and GSOD records, July 2026 — and please cite the year, because the figures are recomputed each February when the previous cyclone season closes.
Cyclone and rainfall questions, answered
Is there a cyclone season in the Andaman Islands?
Yes, and it is not the season most people expect. Across 35 years of NOAA best-track records, 34 tropical systems have passed within 300 km of Port Blair — about one a year. They cluster in October, November and December, with a smaller secondary peak in April and May. November alone accounts for twelve of the thirty-four. The monsoon months of June to September, which travellers worry about most, have recorded none at all in that entire period.
Has a cyclone ever hit Port Blair directly?
Systems have passed very close. Cyclone Lehar came within 14 km of Port Blair in November 2013, an unnamed system within 16 km in November 1998, and Gaja within 26 km in November 2018. Hudhud passed 66 km away in October 2014 at a peak intensity of 115 knots. Close approaches are uncommon but real — ten systems have come within 100 km in 35 years, roughly one every three or four years, and eight of the ten in October or November.
Is the Andaman monsoon dangerous for tourists?
It is wet, not stormy. The monsoon is the rainiest part of the year and the sea gets rough, which is why ferry schedules are less dependable then. But it is also the only part of the year with no cyclone record at all near Port Blair across 35 seasons. The disruption the monsoon brings is a cancelled crossing or a lost beach day. Build spare days into the itinerary.
How many days does it actually rain in the Andamans in July?
About 20 days in a 30-day month, according to ten years of readings from the Port Blair airport gauge — which leaves roughly 10 dry days even in the wettest stretch of the year. A NASA gridded dataset for the same years gives 28 rain days for July. It estimates rainfall over a grid cell roughly 55 by 70 km, not at a point, and it is not a forecast. The gauge is measuring rain that actually fell at the airport.
Will my ferry be cancelled if there is a cyclone nearby?
Quite possibly, even if the system stays a long way off. A cyclone 200 km away is far enough to leave Port Blair itself untouched and still make the sea too rough for inter-island sailings — the disruption radius is much wider than the damage radius. Operators typically suspend sailings one to two days ahead, usually once the India Meteorological Department issues a red alert. The track data is consistent with that short notice: among the 14 systems tracked from more than 500 km out before coming within 300 km of Port Blair, the typical gap between their last position outside that ring and their closest approach was about a day and a half. When a system is tracking near the islands we contact affected guests, rework the itinerary around the sea state and rebook crossings.
Where does this data come from?
Three public datasets, all published by United States government agencies. The cyclone tracks are IBTrACS, the international best-track archive maintained by NOAA, covering the North Indian Ocean basin — we measured the distance from every archived track position to Port Blair. The rainfall is NOAA GSOD daily records from the Port Blair airport weather station itself. The comparison column is NASA POWER, a gridded reanalysis. We wrote the analysis; the observations are theirs, and our figures regenerate from committed scripts rather than being typed in by hand.
Turn the data into a trip
- Andaman weather, month by month — temperatures, sea state and the verdict for each month.
- Ferry cancellations study — what 28,482 booked legs show about cancellations.
- Best time to visit — the honest verdict, season by season.
- The Andaman itinerary — including the return overnight and departure day.
How this page stays true
Compiled by the Tropical Andamans team in Port Blair, July 2026, from the public datasets named above, and corrected on 5 October 2026: the comparison column is now labelled as what it is, a NASA gridded reanalysis, and recomputed from a committed script; a storm-by-storm wave comparison we could not source was removed; and the 2025 result and the warning-window figure now say how they were measured. The numbers are computed, not estimated, and they regenerate from a committed script each February once the previous cyclone season closes. Nothing here is a forecast: it is a record of what has already happened, which is a useful thing to plan against and a useless thing to predict with. If you want to know what the sea is likely to do on your actual dates, ask us — that is a conversation, not a table.
Planning around the weather?
Tell us your dates and we will tell you honestly what that month tends to do — the sea, the rain and the crowds — from the people who work the season.