Indonesia's Fire Season Ignites Early: 5,000 Hotspots, Drying Peat and the Return of Transboundary Haze Risk

Indonesia enters the peak of its dry season with the fire indicators already flashing. BMKG, the national meteorology and climate agency, recorded 5,019 hotspots nationally through the end of July 2026, with the largest concentrations as of 28 July in West Kalimantan, Riau, South Papua, Aceh and East Kalimantan. The agency's forecast is explicit: karhutla — forest and land fire — risk peaks in August–September 2026, with Riau, South Sumatra, Jambi, West Kalimantan, Central Kalimantan and South Papua named as priority high-risk areas. The watchdog Pantau Gambut counts a larger number that matters even more: 23,546 hotspots detected in peatland areas since January. Haze has already been reported in West Kalimantan.

The mechanism: why El Niño finds Indonesia's weak point

El Niño's first move is to shift tropical convection east, off the maritime continent — the rainfall Indonesia loses is, almost literally, the rainfall the central Pacific gains. That deepens the June–October dry season into severe deficit. BMKG's June update quantifies it: 482 climate zones covering 56.18% of Indonesia's land area are forecast to receive below-normal rainfall, and 437 zones (48.77%) face a longer-than-normal dry season, with the peak in August.

The second stage is what makes Indonesia's fires unlike grass or forest fires elsewhere. Vast areas of peatland have been drained for agriculture and plantations. As groundwater falls in a rainless season, the peat — organic soil storing roughly 20 times more carbon than typical mineral soils — becomes combustible to depth. Fires ignited at the surface burrow downward and smoulder, resisting suppression for weeks and releasing enormous quantities of PM2.5 and carbon dioxide. Peatland groundwater level is a documented alert indicator for El Niño and positive-IOD years precisely because it measures how deep the fuel goes. This is the chain — Pacific SSTs to Bornean groundwater — that makes Indonesia a signature case of the science of teleconnections.

The escalation risk: haze without borders

The difference between a bad fire season and a regional crisis is wind direction and scale. In 1997–98, drought and forest fires produced haze that sickened millions across Southeast Asia; in 2015, peat fires inflicted losses the World Bank put at roughly $16 billion on Indonesia alone, alongside a public-health emergency whose respiratory toll is still being studied. The escalation marker to watch in 2026 is transboundary spread — smoke reaching Malaysia and Singapore, which converts an Indonesian emergency into a diplomatic and economic one. With haze already detected in West Kalimantan before the forecast peak, the early-warning value of the next six weeks is high.

One honest complication belongs in the record: BMKG's ENSO forecasts ran more conservative than NOAA's through spring 2026 — a legitimate divergence between respected agencies. NOAA's July diagnostic now puts weekly Niño 3.4 at +2.1 °C with an 81% chance of a very strong event in October–December 2026; BMKG's own late-July messaging describes Indonesia entering a strong El Niño phase with drought and karhutla danger looming. The agencies have converged on the risk, if not always on the numbers.

The lagged damage: palm oil in 2027

Not all of the cost arrives with the smoke. Oil-palm trees respond to drought stress with reduced fresh-fruit-bunch yields on a 6–12 month lag, meaning the production impact of the 2026 dry season lands in mid-2027 — a slow-burning consequence for the world's dominant vegetable-oil supply that will surface long after the haze clears. It is the same lag structure that defines this event globally, from Amazon fires to the 2027 temperature record risk tracked in our event timeline.

What to watch and when

Through August–September: daily hotspot counts against the 5,019 July baseline; peatland groundwater levels in Riau, South Sumatra and Kalimantan; air-quality readings in Pekanbaru, Palembang and Pontianak; and any detection of haze crossing the Malacca Strait. Watch, too, the state response — water-bombing deployments, cloud-seeding operations ahead of dry spells, and enforcement against burning on concessions. October–November brings the pivotal question: does the monsoon return on time, or does El Niño delay the wet-season onset and stretch the fire window? BMKG's dry-season update points to the peak in August, but a delayed onset is the classic strong-El Niño sting in the tail. The season's severity is a probability distribution, not a script — 2015 is the downside scenario, not the central one — but with more than half the country forecast dry and the fuel already primed, the next eight weeks are the ones that decide.