ASMC lowered the southern ASEAN haze alert to Level 1 on 5 November, a week after Level 2. BMKG’s count of zones in the rainy season is 59, 8.4 percent, nearly unchanged in ten days, with 73.8 percent in its below-normal rainfall category. Relief is arriving unevenly.
The ASEAN Specialised Meteorological Centre (ASMC) lowered its southern ASEAN haze alert to Level 1 on 5 November, a week after it lowered it to Level 2. It said showers continued, hotspot counts stayed low and the haze improved further. The rainy-season record is less settled. BMKG, Indonesia’s weather and climate agency, counts 59 of its seasonal zones, 8.4 percent, as having entered the rainy season in its assessment of the last ten days of October, published on 3 November. That is nearly unchanged from 58 ten days earlier, and 73.80 percent of the rainfall analysis fell in BMKG’s below-normal category. The 8.4 percent counts zones, not land area, and for the 73.80 percent BMKG’s page does not specify the denominator.
The question here is narrow: has the delayed rainy season brought sustained relief, or are showers arriving while the larger deficit stays in place? The record to date says the relief is real but uneven, and not yet shown to be sustained. The clearest sign of easing is in fire and haze, not in the onset count.
| Question | What the record shows | Status |
|---|---|---|
| Have fire and haze eased? | ASMC lowered its southern ASEAN alert from Level 3 to Level 2 on 29 October and to Level 1 on 5 November, citing continued showers and low hotspot counts. | Yes, through 4 November |
| Has the rainy season arrived? | 59 zones, 8.4% by zone count, in BMKG’s late-October count, nearly unchanged from 58 ten days earlier. BMKG forecast most onset for November and December. | Early. Too soon to call the forecast wrong. |
| Is rainfall recovering? | The below-normal category share fell from 87.82% to 73.80% in the last ten days, near the 70.41% of late September. For October, BMKG reports 85.79% in its below-normal category. The denominator is not specified. | Partial |
| Is the relief sustained? | BMKG’s November outlook is below normal to normal, and it expects too little rain-cloud growth over most of Indonesia until about mid-November. | Not established |
| ASMC alert | Level | Hotspots reported | Reason given |
|---|---|---|---|
| 26 August | 3 (activated) | Not in the material read | Not in the material read |
| 29 October | 2 (downgrade) | Sumatra 1 and 0; Kalimantan 64 and 54 (27 and 28 Oct) | Showers had returned in recent days, hotspot counts fell, haze somewhat improved |
| 5 November | 1 (downgrade) | Sumatra 64 and 7; Kalimantan 13 and 34 (3 and 4 Nov) | Showers continued, hotspot counts stayed low, haze improved further; risk of transboundary haze likely low |
Level 1 is the level ASMC listed at the start of the dry season on 10 July. ASMC’s 5 November alert says isolated hotspots and local smoke may still be expected over drier areas, and the counts it lists are not a smooth decline: Sumatra reported 1 and 0 hotspots on 27 and 28 October and 64 and 7 on 3 and 4 November. Four days of counts are not a trend. ASMC expects the hotspot and haze situation to stay subdued with more rain in the coming weeks, which is its expectation, not a measurement.
That settles one of our own tests. Our 15 October haze post said ASMC would still list Alert Level 3 on 1 November. It did not: the level fell to 2 on 29 October and to 1 on 5 November. Our 1 November alert-level prediction failed. That post also described BMKG’s 87.72 percent below-normal figure for 1 to 10 October as a share of Indonesia’s area, which BMKG’s pages do not say. We have added a dated note to the post and an entry to the corrections log covering both items, and the original prediction stays in place.
The fire signal is not the river signal. We have no verified gauge reading or shipping update for the Kapuas, the river behind our palm-oil shipping post, and we do not treat the haze downgrades as evidence about it.
On 22 September BMKG forecast that the 2026 to 2027 rainy season would start later than normal in 529 seasonal zones, 61.08 percent of land area, with 11 zones, 1.12 percent, starting early and 26, 2.15 percent, on time. It put onset in November for 185 zones (24.66 percent of land area) and in December for 266 (17.31 percent). On 26 October BMKG’s head said most of Indonesia would enter the rainy season in November and December, arriving gradually. Market Watch 5 covered that forecast when it was issued.
Those land-area shares and the 8.4 percent are different measures. The 8.4 percent counts zones; the forecast’s 61 percent weights each zone by its area. The 59 zones cannot be read as 8.4 percent of land area, and 8.4 percent is not a score against a 61 percent forecast.
BMKG’s 3 November slides list some zones in parts of 30 provinces, from Sumatra through Java, Kalimantan, Sulawesi and Nusa Tenggara to Maluku and Papua, as expected to enter the rainy season between 1 and 30 November. November’s forecast window has barely opened, so 5 November is too early to call the forecast wrong.
| BMKG ten-day period (release date) | Zones in rainy season | Below normal | Normal | Above normal |
|---|---|---|---|---|
| 21 to 30 September (5 Oct) | 40 (5.7%) | 70.41% | 9.16% | 20.43% |
| 1 to 10 October (13 Oct) | 53 (7.6%) | 87.72% | 5.19% | 7.09% |
| 11 to 20 October (23 Oct) | 58 (8.3%) | 87.82% | 5.71% | 6.47% |
| 21 to 31 October (3 Nov) | 59 (8.4%) | 73.80% | 11.08% | 15.12% |
The count of zones in the rainy season rose by 13, then 5, then 1 across the last three releases, and the latest change is nearly nothing. Rainfall changed differently. The below-normal category stayed near 88 percent for the first two ten-day periods of October, then fell to 73.80 percent, with the above-normal category rising from 6.47 to 15.12 percent. That is a real shift, but it takes the below-normal share back to roughly its end-of-September level, not below it. For October, BMKG reports 85.79 percent in its below-normal rainfall category; the cited material does not specify the percentage’s denominator. BMKG’s slides also describe the month’s rainfall as the fourth driest since October 1991, without saying whether that ranking is among Octobers or among all months.
The named regions differ between releases. BMKG’s late-September list of regions in the rainy season included parts of Jambi, South Sumatra, Banten, West Java, West and East Kalimantan, West Sulawesi, Maluku and Papua. Its October lists name Aceh, North Sumatra, Riau, West Sumatra, Bengkulu and the Riau Islands, and by late October a small part of Bogor and part of West Kalimantan. BMKG does not explain the difference, and we have not resolved it. We report the counts as published and do not infer what the changing lists mean. A departure from normal also compares each place with its own climatology, so in places where October is normally dry, below normal can mean very little rain in absolute terms.
BMKG’s warnings for 1 to 10 November apply to particular districts, not to the country. Heavy-rain warnings at the highest level, Awas, cover parts of Aceh, North Sumatra and West Sumatra, the same northwestern corner where the rainy season has arrived. Drought warnings at the Awas level cover parts of Banten, East Java, West and East Nusa Tenggara, South and Southeast Sulawesi, Maluku, North Maluku and South Papua. The two lists mostly describe different places. Banten appears on both, with drought at the Siaga and Awas levels and heavy rain at Siaga, and BMKG’s slides name provinces rather than districts, so we cannot say which districts.
A heavy-rain warning is about an event; seasonal onset is about persistence, and a downpour is weaker evidence of recovery than rain that keeps coming. BMKG makes the point itself. In mid-October, as iNews reported it, BMKG said significant daily rain was still possible where atmospheric conditions supported it even though most of Indonesia stayed dry on a ten-day scale. Its 3 November slides expect ten-day rainfall of mostly low to medium amounts (0 to 150 mm) for November, with the Madden-Julian Oscillation (MJO), a tropical rain-cloud pulse, in phases 2 to 3 until about mid-November and too little rain-cloud growth over most of the country. Its monthly outlook for November is below normal to normal, and for December through April normal to above normal.
It does not show that El Niño or the positive Indian Ocean Dipole caused this month’s rainfall, or that either will end the dry pattern. BMKG reports an ENSO index of +3.09 for August to October and an Indian Ocean Dipole index of +0.659 for the last ten days of October, and those are BMKG’s own indices, not NOAA’s. It does not say how large any rainfall deficit is in millimetres, because the percentages above are shares in categories whose denominator BMKG does not state. It does not show why ASMC’s counts vary from day to day. It says nothing about crops or prices, which we leave out until evidence connects them to the observed weather.
BMKG publishes each ten-day Analisis Dinamika Atmosfer on its website with a PDF update, and ASMC lists its alerts on its website. All the BMKG and ASMC pages used here are linked below, along with our earlier Indonesia posts.
Our reading. Fire and haze have eased fastest: two alert downgrades in a week. The wet season is beginning in northwestern Sumatra, the south and east remain under drought warnings, and the count of zones in the rainy season is nearly unchanged in ten days. That is uneven relief, not yet shown to be sustained. BMKG’s analyses for 1 to 10 and 11 to 20 November are the next evidence on whether onset arrives in the places and periods it forecast.