Very strong El Niño and a positive IOD are suppressing rain over Indonesia. They raise the odds of fire weather and trapped smoke, but people and land use still determine where the fires start.
Two ocean signals are now taking rain away from Indonesia, and fire is what turns the dryness into smoke. BMKG, Indonesia’s weather and climate agency, reports a very strong El Niño (its Niño 3.4 index was +2.70 for July to September) and a positive Indian Ocean Dipole, or IOD (+0.72 for September). Those signals work on rainfall. The smoke needs a second step that the ocean cannot supply.
That step is now visible across borders. ASEAN’s smoke centre, ASMC, has held its highest alert for southern ASEAN since 26 August. Singapore’s National Environment Agency (NEA) put the 24-hour Pollutant Standards Index (PSI) at 118 to 168 at 4pm on 14 October. Malaysia’s air-pollutant system showed 34 areas in the Unhealthy range at 9am on 15 October, and the Education Minister has closed schools in eight Selangor districts for 16 October.
We have covered this season before, and this post does not repeat that background: the 19 July piece on the “Godzilla” nickname, the 3 August post on cloud seeding, the 19 August post on July’s burned area, our June IOD explainer and Market Watch 5 on the delayed rainy season.
Three things are new. BMKG’s operational weekly products now state the two-driver diagnosis for the current dry spell, where our July post could only say the IOD was forecast. The first ten days of October are observed rather than projected: BMKG puts 87.72 percent of Indonesia’s area below normal rainfall. And the smoke has escalated across borders, which in July was only a thing to watch.
| Signal | What it changes | What it does not prove |
|---|---|---|
| Very strong El Niño BMKG Niño 3.4, +2.70 for July to September | BMKG says it reduces the potential for rain-cloud formation, and that it has produced dry periods in some zones that normally have year-round rain. | That it started or sized any fire, burned area, death, school closure or pollution reading. BMKG assigns it no share of the drought. |
| Positive IOD BMKG, +0.72 for September | BMKG says it contributes to reduced water-vapour supply, with air masses over the Indian Ocean moving away from Indonesia. | That it adds a measured amount on top of El Niño. No source read splits the drying between the two. |
| Fire Hotspots, land clearing, drained peat | Turns dry fuel into smoke. Officials urge the public not to clear land with fire. | That any particular fire is natural, or that any particular company lit it. A hotspot is a heat detection, not a hectare. |
| Wind Southeast to south, per ASMC and NEA | Carries smoke from Sumatra and Kalimantan toward Singapore and Peninsular Malaysia. | That every reading abroad traces to a named fire. No back-trajectory product was read. |
El Niño gets the headlines because it is very strong, and NOAA’s Climate Prediction Center said on 8 October that a strong to very strong event is likely through January to March 2027. CPC’s discussion also says anomalies “remained suppressed over Indonesia,” but it does not mention the IOD. BMKG does. Its outlook for 2 to 8 October says a positive IOD also contributes to reduced water-vapour supply, and its 22 to 28 September outlook says the weekly dipole index of +0.7 shows air masses over the Indian Ocean moving away from Indonesia, which could significantly reduce water-vapour supply from western Indonesia. On 22 September BMKG’s head said the combination of El Niño and IOD worsens and prolongs the drought.
BMKG’s IOD value depends on the window: +0.376 for August, +0.72 for September, +0.467 for July to September. Its El Niño values are its own indices, not NOAA’s relative index. The pages we read give no category thresholds and no numerical share for either driver, so this post assigns none.
BMKG’s analysis for 1 to 10 October, published 13 October, puts 83.85 percent of Indonesia’s area in the low rainfall category and only 53 seasonal zones, 7.6 percent, in the rainy season. Its meteorological drought warnings for 11 to 20 October reach the top category, Awas, in areas including South Sumatra, West Java, East Java and East Nusa Tenggara, with Siaga and Waspada in many more. Warnings are issued by regency, not for whole provinces.
A hotspot is a satellite heat detection. Burned area is a mapped extent confirmed later. PM2.5 and the PSI measure the air people breathe, and respiratory illness counts measure what clinics saw. They are four different questions and none converts into another.
BMKG’s high-confidence hotspot counts as of 11 October put Papua Selatan first at 482, ahead of Sumatera Selatan at 234 and Kalimantan Tengah at 213. That ranking is not the smoke story: NEA and ASMC trace the smoke reaching Singapore to Sumatra and Kalimantan. Counting systems also disagree on size. On 10 October the Forestry Ministry’s SiPongi showed 11,772 detections in 24 hours, of which 578 were high confidence, as reported by Katadata.
Severe smoke is not a measure of a record season. The latest national burned-area total we could verify is 202,004 hectares for January to July, reported by Suara on 14 October from the Forestry Ministry, against more than 600,000 hectares through July in 2015 (see our 19 August post). A full-season figure is not in.
Whether the current fires are mainly peat, forest or agricultural clearing is unresolved in what we read. BNPB’s head said on 7 September that 72,008 hectares of peatland had burned (IDN Times). The Ministry’s South Kalimantan release for January to June counted 29.44 of 383.07 burned hectares as peat. Neither is a national mix. On cause, the Ministry asks the public not to clear land with fire and describes sanctions and criminal action against those proven to have burned forest, and as reported on 7 September police were investigating eight corporations with none named a suspect. None of that is a finding about a specific fire.
ASMC’s Level 3 criteria require significant, persistent hotspots, moderate to dense haze for two or more consecutive days, dry weather and prevailing winds blowing smoke toward neighbours. Its 7 October page reported moderate to dense transboundary haze over Peninsular Malaysia, Singapore, Kalimantan, Sumatra and western Sarawak. NEA’s 14 October update described dense smoke haze from fires in Kalimantan drifting toward Singapore, with 1-hour PM2.5 at 139 to 251 µg/m³. On 15 October at 9am Malaysia’s system showed Johan Setia in Selangor highest at 192 on its air-pollutant index, with 34 areas Unhealthy, up from 23 the morning before.
Those are different indices on different clocks, so we do not tabulate them. What the official products support is smoke from Sumatra and Kalimantan drifting toward Singapore and Peninsular Malaysia under prevailing winds. Malaysia’s published readings name no source, and we read no back-trajectory product, so a given Malaysian reading is not traced to a given fire.
Reporting, not agency data: Malaysia’s Deputy Prime Minister said Malaysia’s offer of an aircraft, three helicopters and 55 personnel had no response as of 11 October, and Indonesia’s security minister said on 12 October that Indonesia can still cope and has not declared a national disaster (Suara.com, 14 October). The Health Ministry’s 174,694 acute respiratory infections in seven provinces, 18 August to 21 September, count infections from any cause, not smoke illness.
BMKG’s 22 September forecast puts rainy-season onset later than normal in 529 of its seasonal zones, 61.08 percent of land area, with the start in November for 185 zones (24.66 percent) and December for 266 (17.31 percent). At the end of September only 40 zones, 5.7 percent, had entered the rainy season. BMKG’s list of areas likely to exceed 300 mm in November includes Aceh, Riau and West Kalimantan but not South Sumatra or Central Kalimantan, which lead recent Ministry hotspot counts. That omission is our reading of the list, not a BMKG statement.
ASMC gives the same timing from the regional side. Its 12 to 25 October outlook is drier than usual for most of southern ASEAN, and late-October inter-monsoon showers could start to ease the haze, but hotspots and haze are likely to stay elevated in fire-prone areas with prolonged dryness.
Our reading. This is not an El Niño-only story. Two ocean signals reduce the rain, the deficit raises the odds of fire weather and trapped smoke, and people and land use decide where fires start. The dryness, the fire and the smoke are all documented. How much of any one fire, hectare or reading belongs to the ocean is not, and no source we read says.