Camping in Japan from September through early December fails in different ways than summer camping does. Fire moves closer to where people sleep, darkness arrives earlier each week, and the air dries out. Those three shifts sit behind the season's recurring incidents: carbon monoxide poisoning in tents and cars, hypothermia, and fires that escape a fire pit. This article works from material published by the Japan Pediatric Society, the Fire and Disaster Management Agency, the Japanese Association for Acute Medicine, the Japan Meteorological Agency, JAF and the National Astronomical Observatory of Japan. No equipment was field-tested for it, so no specific product or campground is assessed here.

What actually changes after summer

Once the temperature drops, heating moves under the fabric. Stoves, charcoal, gas burners, water boiled for a hot bottle: handling of fire that stayed outside in July drifts under the fly sheet or into the porch. That single change — fire getting closer to the space people sleep in — sits at the centre of how autumn and early-winter incidents happen.

The second change is daylight. According to the calendar service of the National Astronomical Observatory of Japan, sunset in Tokyo in 2026 falls at 18:09 on 1 September, 17:26 on 1 October and 16:28 on 1 December. That is a swing of one hour and 41 minutes between early September and early December. Arriving at 17:00 means pitching in daylight in September and pitching in the dark in December.

The third is dryness and wind. The FY2025 Fire and Disaster Management White Paper, published by Japan's Fire and Disaster Management Agency on 20 January 2026, records 831 forest fires during 2024 and notes that they cluster in April, then January and May — the months with low rainfall, dry air and strong winds. Late autumn heads straight into that window.

The sections below take carbon monoxide, hypothermia and fire handling in turn, using what government bodies and medical societies have published.

Carbon monoxide in a tent: leaving a vent open is not a defence

Injury Alert No. 83, published by the Japan Pediatric Society in its journal (vol. 123, no. 11, 1 November 2019), documents carbon monoxide poisoning inside a two-room tent. In November 2017 a family of four was camping at an outdoor campground. The charcoal cooking was done under the tarp — outdoors — for roughly two hours, and neither the used grill nor any heater was brought inside the tent afterwards.

The poisoning happened anyway. During cooking, the divider between the living space and the inner tent was fully open while every side vent was shut. After the meal, feeling cold, the family closed the main entrance leaving a gap of about 30 cm at ground level. At 19:30 a six-year-old girl and an eight-year-old boy went into the inner tent, complained of headache and collapsed unconscious minutes later. Carboxyhaemoglobin (COHb) measured at hospital was 20.8% for the girl and 18.5% for the boy. The report estimates the CO concentration inside the inner tent at 1,700 ppm or above. No CO detector was in use.

The case matters because both common defences were in place: the fire was outside, and a ventilation gap was left. The same report cites warnings from the US CDC and CPSC and from Japan's National Institute of Technology and Evaluation (NITE) stating that opening a window, a door or a tent flap is not enough to prevent CO poisoning. It also notes experiments in which burning a stove inside a tent raised CO levels within about ten minutes across several different configurations, and a tent manufacturer's test in which a shichirin charcoal grill pushed levels past 400 ppm in roughly 15 minutes.

For severity, the report gives the general picture as headache at 10–20% COHb, dizziness at 20–40%, coma and convulsions at 50–60%, and death around 70%. Children have a higher basal metabolic rate and show symptoms at roughly half the adult COHb level — headache at 7%, coma at 25%. Carbon monoxide is colourless, odourless and non-irritant, and binds to haemoglobin around 200 times more readily than oxygen. It is not a gas anyone can smell coming.

CO alarms: placement, limits, and sleeping in a vehicle

The same report lists four countermeasures. Keep the cooker at least 6 m from the tent and position it with the wind direction in mind; do not cook over fire in the porch area under the tarp either, not only inside the tent; open every tent vent while cooking; and always fit a CO detector near the ceiling when a tent is in use. The 6 m figure matches the 20 feet the American Academy of Pediatrics recommends between CO-producing appliances and living space.

An alarm helps, but it is not a licence. By the time it sounds the concentration has already risen, and a sleeping person reacts late. Battery charge and cold-weather behaviour both affect it. The safer reading is that an alarm exists to catch an inflow nobody planned for — not to make burning appliances inside a tent acceptable.

One practical point deserves rehearsing in advance: early CO symptoms are hard to separate from ordinary tiredness. Headache, nausea and drowsiness all look normal after a long drive and a pitch in the cold. If more than one person under the same fly reports feeling unwell at the same time, treat CO as the working assumption, get everyone outside into moving air, and call 119. Deciding that order beforehand removes the hesitation.

Vehicles carry the same problem. In a user test run by JAF on 16 February 2015 in Asahikawa, Hokkaido, cars were buried in snow up to the bonnet with the engine running while cabin CO was measured. With no countermeasure, the cabin reached 400 ppm after 16 minutes and hit the meter's 1,000 ppm ceiling at 22 minutes. Even with a window open 5 cm, levels reached the low hundreds within five minutes, 400 ppm at around 40 minutes and the 800 ppm range immediately after. The car whose exhaust area had been cleared of snow registered almost no CO. Cracking a window is not a countermeasure. The test used snow, but any blocked exhaust outlet produces the same result — drifted leaves, mud, a sloped parking spot, prolonged idling in a confined space. Autumn supplies those conditions too.

Hypothermia is set by wet and wind, not by the thermometer

The medical glossary of the Japanese Association for Acute Medicine defines accidental hypothermia as a core body temperature at or below 35°C, and grades it as mild (35–32°C), moderate (32–28°C) and severe (28°C or below). Shivering is present in the mild range, disappears in the moderate range, and gives way to muscle rigidity in the severe range. The listed causes include drowning and mountaineering accidents, which is the clue: nothing here requires sub-zero air. The definition is anchored to core temperature, not to the outside reading.

At an autumn campsite the heat is taken mainly by moisture and wind rather than by air temperature alone. Sweat, rain, condensation or river water soaks clothing, and moving air carries the heat away through evaporation. The classic sequence is a base layer sweated through during the day and never changed before the evening breeze arrives. Treat “the shivering stopped” as deterioration, not relief — it means the body's own heat production is failing.

The cold nights are predictable too. The Japan Meteorological Agency's forecast terminology defines radiative cooling as heat being lost from the ground surface by radiation, lowering the air temperature; clear, low-wind nights therefore produce the sharpest pre-dawn drop, so “it was not cold at bedtime” proves nothing. Altitude compounds it. The same JMA glossary puts the lapse rate at a fall of 0.5–1°C per 100 m in the troposphere. A plateau at 1,000 m therefore sits somewhere in a 5–10°C band below the lowlands — a range to prepare for, not a single number.

Treatment principles are consistent: remove wet clothing, block the wind, warm the torso from the outside, and give warm drinks only to someone fully alert. Confused answers, an inability to stand, or shivering that has stopped are grounds to call 119. Waiting out an ambiguous case until morning is the wrong call in a season when the temperature keeps falling all night.

Reading a sleeping bag's comfort and extreme ratings, and dealing with condensation

Sleeping bag temperature figures come from the international standard ISO 23537. QTEC, the Japan Textile Products Quality and Technology Center that performs the testing, defines comfort temperature as the temperature at which a standard adult woman can sleep soundly in a relaxed posture; limit temperature as the lowest temperature at which a standard adult man can sleep eight hours without waking, in a curled posture; and extreme temperature as the limit at which a standard adult woman can hold out for six hours, described as the temperature at which risk of harm from hypothermia arises.

The extreme rating is the one most often misread. It is not a temperature to sleep at; it is defined as the point where hypothermia risk begins. Reading “extreme −5°C, so 0°C is fine” works against the standard's intent. The test bodies are specified as well: QTEC gives the standard man as 25 years old, 173 cm and 73 kg, and the standard woman as 25 years old, 160 cm and 60 kg. The further a person sits from those figures, the wider the gap between label and reality.

The rating also describes the bag alone. Heat lost downward depends on the sleeping mat's insulation, and clothing, the last meal, fatigue and whether the bag is damp all shift how it actually feels. Swap a thick mat for a thin one and the same bag performs differently.

Condensation is where that last factor bites. Water vapour from breath and skin meets a cold fabric wall and turns to liquid. In a double-wall tent the vapour passes through the inner and condenses on the underside of the fly, so less of it lands on the bag; in a single-wall tent, or with the vents closed down, droplets form on the wall the bag touches. Note the coupling: shutting the vents raises the condensation problem and the carbon monoxide problem at the same time. Getting the bag out into the air first thing, and keeping wet clothing out of it overnight, stops the damp carrying over to the next night.

There is no single national campfire rule — and since 2026 there is a wildfire warning

Whether ground fires are allowed, whether a fire stand is mandatory, where ash may be dumped: campfire rules are set campground by campground, and no nationwide rule exists. On top of that, the Fire and Disaster Management Agency notes that municipal fire prevention ordinances may require prior notification to the fire service before handling fire outdoors, including campfires. The rules come in two layers — the site's and the municipality's.

A third layer arrived in 2026. The agency created a forest fire advisory (rinya kasai chuihou) and a forest fire warning (rinya kasai keihou), and operation began in many municipalities across Japan from January 2026. The legal basis is a partial revision of the model fire prevention ordinance dated 29 August 2025; the actual start date and the specific thresholds are set by each municipality, so whether a destination operates the system has to be checked against that local authority.

According to the agency, a forest fire advisory marks conditions of rainfall and dryness under which forest fires start and spread easily; the illustrative indicator combines rainfall of 1 mm or less over the preceding three days with either 30 mm or less over the preceding 30 days or an active dry-air advisory. A forest fire warning adds a strong-wind advisory to those conditions, marking a state in which a fire that does start is likely to grow large. While an advisory is in force, people are expected to make an effort to refrain from outdoor use of fire. While a warning is in force, outdoor use of fire is prohibited, and failing to comply can bring a fine of up to 300,000 yen or detention. Anyone planning a campfire now has one more thing to check besides the forecast.

The white paper explains the origin. It cites the forest fire that broke out in Ofunato, Iwate Prefecture on 26 February 2025, which damaged 226 buildings and burned roughly 3,370 hectares, and records that the advisory and warning were created to make forest fire prevention effective, along with clarification that campfires fall under the notification requirement. The statistics back the emphasis: of 37,141 fires during 2024, cigarettes caused 3,058, campfires 2,781 and cooking stoves 2,718, and dead grass was the most commonly ignited material at 17.0% of all fires. Among the 831 forest fires alone, campfires were the leading cause at 245 cases, or 29.5%. If a fire is lit in the season of fallen leaves and dry grass, the basics — clear combustible material from around the fire, have water ready before lighting, and stay until it is fully out — are what those numbers argue for.

Darkness arrives early: build the day backwards

Following Tokyo's 2026 sunsets through the calendar service of the National Astronomical Observatory of Japan gives 18:09 on 1 September, 17:27 on 30 September, 17:26 on 1 October and 16:28 on 1 December — one hour and 41 minutes of difference between early September and early December. At higher latitudes, and at valley campgrounds that fall into mountain shadow, useful light ends earlier still.

The practical move is to work backwards. Look up sunset for a point near the destination, subtract the time a pitch takes, and set the arrival time from that. Driving pegs, tensioning guy lines and handling fire are all measurably harder after dark. Starting the fire and the cooking before sunset, and clearing up after it, buys margin.

Use a head torch as the primary light so both hands stay free, and keep a separate light for inside the tent. Batteries lose capacity in the cold, so carry spares and keep the unit somewhere warm — inside the sleeping bag — overnight. Packing up runs on the same logic: the fly comes down wet with condensation during the coldest hour of the morning, so build drying time into the plan.

Bears in autumn, and four places to check before leaving

Autumn also changes bear behaviour. Feeding intensifies before hibernation, and the beech and oak mast crop swings widely from year to year, which moves the sighting picture with it. The Ministry of the Environment publishes provisional figures on bear sightings and injuries, and prefectures issue their own sighting maps and advisories. Concrete campground practice — food and rubbish handling, layout at night, what to do on an encounter — is covered in separate articles on this site, so this piece stops at the seasonal point.

Everything above can be established before leaving home. Four places cover it, with the Ministry of the Environment and the destination prefecture added for bears.

One closing caveat. This article is compiled from published material and does not assess any individual piece of equipment, any campground, or the weather on a given day. Rules and thresholds get revised. For anything carrying a penalty — campfire permission and outdoor fire restrictions above all — confirm it against the destination's municipality, fire service and campground before setting off.

  • Japan Meteorological Agency — warnings and advisories for the destination (dry-air advisory, strong-wind advisory, wind, overnight low)
  • The destination's municipality and fire service — whether forest fire advisories and warnings are in operation, and whether a campfire notification is required
  • The campground's own website — ground fires, fire stands, ash disposal, winter facilities and closures
  • National Astronomical Observatory of Japan, calendar service — sunset time for a point near the destination

Sources & References