The fires that came from everywhere: how the 2015 Southeast Asian burn defied a single culprit

On 17 July 2026, a team of fire scientists published a finding that complicates a decade of received wisdom about Southeast Asia's worst modern haze crisis. The 2015 equatorial fire season, which smothered Sumatra, Kalimantan and southern Malaysia in smoke for months, was not driven by a single dominant ignition source. Most of its largest, most damaging fires, the researchers found, emerged from clusters of small, geographically separate ignition points that converged into one blaze. The shift matters because every policy lever pulled since 2015, from Indonesian moratoria to satellite early-warning pipelines, has assumed a recognisable culprit and a recognisable fix.

The practical question behind the paper is older and more uncomfortable than the haze itself. How does a regional fire season scale into a humanitarian, economic and diplomatic emergency? The answer the team offers is procedural: by aggregating many small ignitions across a fragmented landscape, each individually too small to register on a satellite hotspot feed, until they coalesce into a fire too large to suppress. The implication is that counting hotspots, the standard metric used by scientists and governments to allocate blame and budget, systematically understates the early signal of a coming catastrophe.

A different kind of fire atlas

The research team, whose work appeared in a peer-reviewed outlet covered by Phys.org on 17 July 2026, used satellite-derived active-fire detections to reconstruct the earliest origins of large blazes recorded during the 2015 season. Rather than asking where fires ended up, they asked where each fire began and what surrounded it at ignition. Most large fires, they concluded, did not grow outward from a single starting point. They accreted, the way a thunderstorm builds, from multiple discrete origins that happened to fall inside the same drainage, plantation block or drained peat dome. Several small fires had merged into one.

That observation, mundane as it sounds, cuts against the dominant story of the 2015 crisis. The narrative that took hold in newspapers, foreign ministry briefings and academic papers alike centred on a recognisable villain: industrial plantation operators clearing land by burning during the dry season, on peat soils that should never have been drained. The paper does not absolve that actor. Drainage and plantation expansion remain the precondition for the disaster. But it argues that conflating the supply of ignition (the plantation, the slash-and-burn clearance) with the supply of catastrophe is the analytical mistake. Many fires were lit deliberately; few were the work of one operator.

Why "a fire" is not a fire

Fire managers have long known that conflating detection with cause is treacherous. A thermal anomaly on a satellite pass can be a campfire, a clearing burn or a peat dome that has been smouldering for weeks. The new work pushes the problem further back, into the fire's own biography: a blaze that ends the season as a single, contiguous scar on the landscape may have begun as several unrelated events that only became one event by accident of geography and weather.

That detail reshapes how blame is assigned and how intervention is timed. Indonesia's moratorium on new palm oil and logging concessions, the peat restoration agency BRG's hydrological work, Singapore's transboundary haze agreement: all are aimed at the large operator, the obvious ignition. A multiple-origin regime suggests a different failure surface. It is the small ignitions, in aggregate, that the existing detection-and-response architecture misses until the fire is already too large to fight.

The haze economy, again

The 2015 fires are estimated to have released more than a gigatonne of carbon and cost the regional economy upwards of sixteen billion dollars, with Indonesia bearing the largest share of public-health damage. Singapore's Pollutant Standards Index crossed three hundred on multiple days; schoolchildren were kept indoors in Kuala Lumpur; commercial flights were diverted. The episode is also the single most-cited data point in arguments about whether palm oil supply chains can ever be made "fire-free", a position that now anchors European Union deforestation rules and several major trader commitments.

This paper does not soften those conclusions. Drainage and concession policy remain necessary conditions of the disaster. What it softens is the implicit assumption that locating the worst fires is a reliable proxy for locating the worst actors. Under a multiple-origin regime, the largest scars are an emergent property, not a fingerprint.

What the new finding does not yet tell us

The reconstruction is retrospective and limited to one season. The authors note that whether the multiple-origin pattern holds in years dominated by different weather regimes, notably the mild El Niño years, remains untested. The dataset also cannot distinguish, at ignition, between a deliberate land-clearing burn and an accidental roadside or agricultural fire; both look like a thermal anomaly until ground-truthed.

What this publication finds most consequential is the operational one. Satellite hotspot counts, the metric that diplomats and journalists both rely on, become a lagging indicator of where the next catastrophe will form. If early-warning systems are rebuilt around the smallest detectable ignitions, not the largest ongoing scars, the window for a peat-fire fight in 2026 or 2027 might open weeks earlier than it did in 2015. Whether governments and trading houses have an appetite for that more granular, less blame-friendly model is a different question, and one the paper does not pretend to answer.

Monexus framed this as a fire-science finding with direct operational consequences for Southeast Asia, rather than as a re-litigation of the palm-oil blame chain. The Western wire line tends to anchor on industrial plantation operators; the paper's contribution is that the unit of analysis should be smaller and more distributed.

Wire provenance

This editorial synthesis draws on the following public wire/social posts:

• https://en.wikipedia.org/wiki/2015_Southeast_Asian_haze
• https://en.wikipedia.org/wiki/Peatland_fires_in_Indonesia