Friday, 4 September 2026, two in the afternoon.

At that moment the Air Pollution Index (API) in Serian, Sarawak, read 521, according to The Star and the NST. The Star said it was the highest reading ever recorded there.

The Star's headline the next day used the word "hazardous". That day, nearby Kuching read around 310, Samarahan roughly 280 to 290, and Sri Aman about 240.

The same day, the Prime Minister's Office issued a statement: the Yang di-Pertuan Agong, Sultan Ibrahim, had consented to the declaration of an emergency across the whole of Serian district. ^1,2 The statement said the decision was made after taking into account the advice of Prime Minister Anwar, Sarawak Chief Minister Abang Johari and others.

By the following Monday, 647 schools were closed. ^3,4

What This Piece Asks#

A word first: this piece is not looking for someone to blame.

It wants to ask three things. First, on what basis was this emergency declared? Second, where did this smoke come from, and how much do we know and not know? Third, how many times has this happened, and after the last time, what system did we leave behind?

Half of the answers are "not found in the public record". We will write it as it is.

An Order Without a Provision#

Start with the statement.

Every report this site read — the reproduction of the Prime Minister's Office statement, the New Straits Times, The Vibes, Harian Metro and Al Jazeera — says one and the same thing: the Yang di-Pertuan Agong gave his "consent" (perkenan). ^1,2,5,4 Not one of them names any provision. No Article 150 of the Constitution, no Act, and no word on whether the emergency must be tabled in Parliament.

We need to pause here, because the word "emergency" carries heavy weight in Malaysia.

Article 150 of the Constitution is the best-known source of emergency powers. This site found no report linking it to the Serian case, and did not obtain the text of the Article. Another name often mentioned alongside is section 29A of the Environmental Quality Act 1974: the summary this site found describes it as a criminal provision against open burning, with a maximum fine of one million ringgit. That is an enforcement tool, not the basis for declaring an emergency, and the two should not be confused.

Deputy Minister Uggah, in a Bernama report, said this emergency is "not like the MCO", and that detailed SOPs would be issued following the directions of the National Disaster Management Agency (NADMA).

That sentence does not answer the question of legal basis; it only says what the emergency is not.

So the honest statement is this: the legal basis for this emergency is not found in the public record this site read. It may be a constitutional-level declaration, or it may be an administrative decision under a disaster-management mechanism that borrowed the word "emergency". The reports do not say, and this site will not write it in on the authorities' behalf.

The Disagreement Over the Numbers#

Now the reading itself.

On the same day, different reports gave different APIs. The Star and the New Straits Times wrote 521 at two in the afternoon; the reproduced PMO statement gave 514, from the police headquarters monitoring station; The Vibes wrote 518; The Jakarta Post, in a 6 September retrospective, said the peak was 512.

So the safer way to put it: above 500, with reports ranging from 512 to 521 depending on the station and the time.

The Department of Environment's (DOE) raw hourly series was not obtained by this site.

The Jakarta Post also wrote that by 6 September the Serian reading had dropped to around 219. In the same report, a resident said a durian seller's income had fallen by about half; a teacher worried that days like these felt like a "lockdown". These two statements are personal accounts, and this site did not verify them separately.

Where the Smoke Comes From#

The official explanation is that it is transboundary.

The ASEAN Specialised Meteorological Centre (ASMC), in a bulletin quoted by The Star, said moderate to dense transboundary haze was observed over northern Borneo, with hotspots concentrated in Kalimantan and central and southern Sumatra.

In search summaries, this site also found a set of DOE figures: at the end of September, roughly 327 hotspots in Kalimantan, roughly 191 in Sumatra, and just 1 inside Sarawak. These figures come from pages that were blocked and whose original text this site could not read, so they are not treated as fact, only as a lead.

What the lead suggests: this haze was most likely from outside, though local risk is not nil. The same summaries say the DOE also stepped up inspections of peatlands and landfills in Sarawak and Johor.

"Outside" and "local" are not an either-or. They are two faces of the same smoke.

Not the First Time#

The first face in the mirror is 1997.

That year Sarawak also declared an emergency. ^6,7 The peak reading commonly cited for Kuching is 838 or 839: the DOE figure relayed by the Korea Times is 838, Lim Kit Siang wrote 839 in his statement of 25 September 1997, and The Star also wrote 839 in its 2026 retrospective. ^8,6,7 The two figures differ by one, and both are secondary relays; the DOE's original series is not found in the public record.

There is also a lower figure, 658, from a search summary; this site could not trace its source, so it is not used.

According to the Korea Times relaying the DOE, the emergency was lifted on Sunday 28 September, schools and workplaces reopened from 29 September, and the Kuching reading fell to 31. From over 800 to 31 — this smoke left faster, and more completely, than people thought.

Then 2005. The Star's retrospective this year wrote "over 500"; this site also saw a secondary compilation that put the peak below 500. The two accounts contradict each other and the contradiction is unresolved.

Then 2013. The same retrospective wrote "over 750". Search summaries say that time it was Muar, in Johor.

Four times, four readings, four emergencies. Each time it was called an "emergency", and each time, in the reports this site found, no provision appears. Note: "each time" here means "each time we found", not "every time".

The Number Counted Afterwards#

The second face in the mirror is 2015.

The haze of that year was later recalculated by researchers using an atmospheric chemistry model. A 2016 Harvard paper entry (the study by Koplitz and colleagues) says the smoke was estimated to have caused 100,300 excess deaths.

This figure has three boundaries that must be stated plainly. It is a model estimate, not a count taken from death certificates. What this site read is only the paper entry, not the full journal text. The country-by-country breakdown was not verified by this site.

What it means: the cost of smoke is often calculated after the smoke clears. When it is calculated, no one is at the scene.

Treaty, Law and Gap#

So what did the system leave behind?

On 10 June 2002, ASEAN countries signed the ASEAN Agreement on Transboundary Haze Pollution (AATHP) in Kuala Lumpur. Article 3 states that parties have the sovereign right to exploit their own resources while ensuring that activities under their jurisdiction do not harm other states; Article 27 provides that disputes be settled amicably by "consultation or negotiation"; Article 29 provides that the agreement enters into force on the 60th day after the sixth instrument of ratification is deposited. Varkkey, a researcher on the issue, argues in her monograph that the regional mechanism lacks binding instruments, so each state can pick the initiatives that fit its own economic interests.

This is not to say the treaty is useless. It is to say: it places the pressure on negotiation, not adjudication.

Indonesia was the last country to ratify. According to a search summary, the Indonesian parliament passed the ratification law in September 2014, and the instrument of ratification was deposited on 20 January 2015. Parliamentary passage and deposit are two different dates and must not be conflated; this section comes from a summary, and this site did not read the original.

Singapore enacted a separate Transboundary Haze Pollution Act in 2014, which can reach entities abroad. The range of fines this site saw only in a summary; the provisions were not opened.

And Malaysia?

In 2019, the then minister, Yeo Bee Yin, was awaiting Cabinet approval to draft a Transboundary Haze Act, with aims including extraterritorial prosecution of overseas subsidiaries and Malaysian-owned plantations. According to a search summary, the plan was shelved in August 2020. Any new progress in 2026 is not found in the public record, and this site did not check Parliament's Hansard.

The Other End of the Smoke#

Reading this far, you may ask: is there really no one to be held responsible?

There is, but it is not a simple answer.

In Haze Problem in Southeast Asia: Palm Oil and Patronage, published in 2016, Varkkey argues that state-led agribusiness development drove palm oil expansion and land clearing by fire; many of the companies involved are Malaysian and Singaporean, with close ties to the politicians and officials responsible for dealing with the problem, creating conflicts of interest.

This is one researcher's argument; what this site read is only the publisher's blurb, not specific chapters.

But the argument deserves a place here, because it shifts the question. Not "which country's people are burning", but "who profits from clearing the land, who enforces the law, and who is overseen by whom". The latter is a systemic problem, and all three countries have a share.

Evidence Boundary#

Sources visible to the reader: event-level facts (the declaration, school closures, the lifting, the 1997 emergency) were reported separately by several outlets and carry markers; exact API readings and hotspot figures are single releases, written in the text as "according to a newspaper" with no marker.
AATHP text and Varkkey's argument: the full treaty text and the publisher's blurb are each a single source, attributed directly in the text and not listed in the source table.
Legal basis of the emergency: not found in the public record. Every report read says only "consent". The description of section 29A of the Environmental Quality Act 1974 comes from a search summary; the original text was not opened.
API readings: between 512 and 521, from five outlets and the PMO statement; the DOE raw series was not obtained.
Hotspot figures: the DOE's 327, 191 and 1 appear only in search summaries of blocked pages.
Lifting of the emergency: 7 September (Monday) around midday, Harian Metro only; the statement said the haze was "tidak lagi dilihat mengancam keselamatan dan kesihatan orang ramai", but the same statement admitted that the Serian and Samarahan readings were still "amat tidak sihat". The Serian reading that day, 222, comes from a summary.
Residents' accounts (durian seller, teacher): The Jakarta Post, a single source.
1997: the lifting date is a single source; the other figure, 658, cannot be traced. The economic loss figure (a search summary says RM802 million for Malaysia) was not read from the book page and is not used.
2005 and 2013: only The Star's retrospective is usable; Muar in 2013 and the claim of the "Prime Minister signing" come from summaries; the date and place for 2005 were not verified by this site.
Indonesia's ratification date, the fines under Singapore's law, the shelved Malaysian law: all from search summaries.
Varkkey: only the publisher's blurb was read; Koplitz et al.'s 100,300: only the thesis entry was read, the full journal text was not opened.
Indonesian enforcement figures: only secondary relays; this site did not obtain a primary source, so they are not in the text.
October Malaysia-Indonesia joint firefighting negotiations: only seen in search summaries, not treated as fact.

The Writer's Verdict

A fair word first: to look at this matter, we do not need to find an enemy first. Smoke has no nationality. It rises from one country's peatland, drifts across a strait, and settles on a schoolyard in another. The people of Serian do not ask which map the air they breathe belongs to; children kept at home do not cough any less for it. What truly unsettles in this matter is the gap in the system, not the ill will of any party. An emergency is declared, yet no one in the reports says its provision; a treaty is signed over twenty years ago, yet disputes can only be "negotiated"; a country's own transboundary haze law has gone from an idea in 2019 to a shelving in 2020, and nothing since. Each link, taken alone, makes sense; put together, they let the same smoke come back again and again. "Emergency" is supposed to be an exception. But when a country declares an emergency again and again, lifts it again and again, without spelling out the legal basis and without closing the enforcement gap, it quietly turns the exception into a season on the calendar. What our country needs is more than declaring an emergency when the smoke arrives; it is to write the provisions clearly and divide the responsibility clearly when the smoke has not come. The standard we demand of our neighbours, we should apply to ourselves: who is clearing the land, who is approving it, who is enforcing, who is profiting — all of it should be sayable and verifiable. Under the same sky, no one is truly outside.

Around midday on 7 September, the Yang di-Pertuan Agong consented to the ending of the emergency. ^1,5 The statement said the haze was no longer seen as a threat to public safety and health.

In the same statement there was one sentence less easy to quote: the Serian and Samarahan readings were still "very unhealthy".

So, the next time a reading climbs past 500, will someone read the provision out?

Note: At the time of writing (October 2026), the legal basis for the Serian emergency is still not found in the public record; the joint Malaysia-Indonesia firefighting operation is still under negotiation. This site will keep watching and will update when a primary source appears.