Kyoto's 2026 autumn leaves are running late: what that does to a November trip

Kyoto's maples do not run on a calendar. They run on night temperatures, and the temperatures have been absurd since July. On 3 August the Japan Meteorological Agency reported that mid-to-late July in western Japan was the hottest since its records began in 1946. Warm autumns delay red maple, so every aggregator with a Japan page is already printing peak dates for late November and early December. Here is what those posts leave out. No Japanese forecaster has published a 2026 koyo forecast yet. The first ones come out in September. Mid-August is also when November rooms in Kyoto stop being available, so the useful move right now is not picking a date. It is building a week that survives being wrong about the date.

What the forecasters have actually published

Checked on 10 August 2026, the three outfits that matter say the same thing: not yet.

tenki.jp, run by the Japan Weather Association, carries a one-line notice that 2026 foliage information begins around September. Weathernews says its next season goes up in September 2026. The Japan Meteorological Corporation still shows its final 2025 forecast, issued in early November.

So the numbers circulating now, 25 November to 7 December with the densest colour around 29 November to 3 December, are not forecasts. They are last year's outcome plus a hot summer, arithmetic done by content sites in July. Nobody should book on them.

The long-run pattern is solid enough to use. tenki.jp puts Arashiyama's usual window at mid-November to early December, and its final 2025 forecast placed Arashiyama in late November, with much of western Japan colouring later than average.

Forecasts checked 2026-08-10. The first official 2026 koyo forecasts arrive in September and get revised through October, so parts of this post will age fast.

Why a hot summer moves the peak later

A maple turns red when it stops feeding the leaf and cold nights trap sugar in the tissue. Cold nights are the trigger, not the date. A September and October that stay mild push the change back, and a long mild autumn tends to make it patchier across a hillside. That is the reasoning behind every "later than average" line in the Japanese forecasts, and the reason summer heat gets read as an autumn clue at all.

The same reasoning argues against doing anything today. July heat is weak evidence about late November. The two months that actually set the date are September and October, and neither has happened. What has been consistent lately is direction rather than precision: recent autumns have run late, and the Japan Meteorological Agency expects western Japan to stay well above normal through August. Plan for late. Do not plan for the 29th.

The 23 November collision nobody puts on the same page

In 2026, Labour Thanksgiving Day (勤労感謝の日) falls on Monday 23 November. Per the National Astronomical Observatory of Japan's official 2026 calendar, that makes Saturday 21 to Monday 23 a three-day weekend nationwide. Culture Day, 3 November, falls on a Tuesday and creates no long weekend at all. November 2026 gets exactly one, and it sits on the leading edge of the window every forecaster is pointing at.

Now put the rooms next to it. The Kyoto City Tourism Association's monthly hotel report for November 2025 covered 110 hotels: 86.9 percent occupancy and an average room rate of 29,085 yen, up 4.0 percent on the previous November. November is already the most expensive month to sleep in Kyoto. A late peak pulls the colour onto the same days as a domestic long weekend and the tail of international autumn travel, in a city already close to full.

Find your arrival week

If your flights are booked, the useful question is not when the peak is. It is where your week sits relative to it.

• Arriving 8 to 15 November: you land ahead of central Kyoto. Go up instead of in. Mount Hiei, Takao and Ohara colour earlier than the city floor, while the temples along the Kamo river will still be green and gold.
• Arriving 16 to 22 November: the awkward middle, and you walk into the long weekend on the 21st. Put the out-of-town half of your plan on Saturday to Monday and keep Tofuku-ji or Eikan-do for Tuesday.
• Arriving 23 to 30 November: probably your best colour and certainly your worst queues. Change your alarm, not your dates. One famous temple at opening, one in the last hour of light, nothing famous in between.
• Arriving 1 to 7 December: the quiet win in a late year. Crowds thin once the calendar says December, rates ease, and in a delayed autumn the lower gardens are often still holding.

Five places within an hour that colour on a different schedule

The queue is not spread evenly. It concentrates in roughly six places, and elevation moves the date, so a 50-minute bus buys a different week rather than the same week with fewer people. All of it checked 2026-08-10.

Reading the table before you commit

The Ohara row is the one to act on first, because JNTO says its leaves turn about one week earlier than the rest of Kyoto, and a week is the entire margin of error everyone is arguing about. One fare is unsettled: West JR Bus says the Kyoto flat-fare section runs out to Toganoo, which puts Takao at the 230 yen city rate, while the Takao tourism association's own access page still lists 520 yen for a different routing. Carry 600 yen and stop thinking about it.

Two entries need a plan rather than a ticket. Jingo-ji is roughly 400 stone steps up from the bridge and runs 9:00 to 16:00, so it is a morning in its own right. Enryaku-ji also closes at 16:00 in November, desk at 15:45, and its 1,000 yen ticket covers three pagoda areas a shuttle ride apart. Shugakuin is free, but the Imperial Household Agency rations it: 18 or over, advance slots capped at 200 per session, same-day numbered tickets from 08:40 until they run out, photo ID checked at the gate.

Should you move your dates?

No. You would be cancelling a confirmed room in the tightest month of Kyoto's year to chase a forecast nobody has written yet, then rebooking into rates that climbed 4 percent last November. The colour is the variable you cannot control. The itinerary is the one you can. Here is the position I will defend: in the last week of November, skip Arashiyama and Kiyomizu-dera completely. A 230 yen bus to Takao at nine in the morning gives you more maple and fewer elbows than any early-alarm strategy at a famous gate.

What the week actually costs

At roughly 160 yen to the dollar, that 29,085 yen average room night is about 180 dollars, well above the 18,000 yen mid-range night we used in our 2026 Japan cost breakdown. November is where the premium sits. Nothing else in this post is expensive: buses run 230 to 360 yen, admissions 700 to 1,000, Nara is 720 yen each way on the rapid, and two of the five entries above cost nothing. If you plan to shop, the tax-free rules changed this year and are worth ten minutes. And if your trip actually starts in September or early October, the weather question in front of you is a different one, which we worked through in the 2026 typhoon post.

Build a week that can move

The plan that survives a late peak is the one you can reorder on the morning. Leave two mornings unassigned in the middle of the Kyoto leg and fill them once the September and October forecasts are out and the temples start posting real conditions.

Travolp is built for that shape of week. Describe the trip and it drafts a day-by-day plan from real places, you download it so it works on the bus to Takao with no data, and when a temple is mobbed you re-plan the day and the rest of it rebases around you. Share it and everyone travelling with you sees the same version. Our 3 days in Kyoto itinerary clusters each day tightly enough that swapping one morning does not break the others, the 5-day Kyoto plan is a starting point you can copy and edit, and our Kyoto plans page is where to build from scratch.

Book the room now. Pick the temples in October, when the forecasts finally exist.