By This Hour Science Desk
Fires burning in Indonesia’s peatlands were sending thick smoke into the skies over parts of the archipelago in early September, with satellite imagery showing a concentration of fire detections on the southern side of Borneo, NASA reported. The immediate concern is not simply the number of visible fires. In peat landscapes, flames at the surface can be accompanied by smoldering combustion below ground, where the fire can persist and spread through dried organic soil.
Aqua satellite observations from September 1 showed gray smoke plumes extending from numerous locations on Borneo. NASA’s mapped detections, generated from thermal signals identified by the satellite’s MODIS instrument and an associated algorithm, were clustered largely in the island’s south. Those detections indicate pixels with heat anomalies consistent with fire; they do not necessarily represent one separate blaze apiece. A single fire can produce more than one detection.
The reports arrive during a period of severe dryness. NASA said Indonesia was about a month into its 2026 fire season by September 2, after conditions in early August left roughly 90 percent of the country with little or no rain, based on Indonesian meteorological-agency data cited by the space agency. That dryness has particular consequences in landscapes that would more commonly retain enough water to limit the spread of fire beneath the surface.
Smoke points to a problem beyond the visible flames
Peat is accumulated organic material in wetland soils. When sufficiently dry, it can burn slowly at comparatively low temperatures, including underground. That makes peat fires unusually difficult to put out: suppression of flames above ground does not necessarily end combustion below it. NASA said fires can continue moving through dry peat deposits out of sight, creating a gap between what is apparent at the surface and what may still be active in the soil.
The satellite view offers an important window on the scale and location of heat and smoke, but it also has defined limits. Thick smoke and cloud can obscure active burning from MODIS and similar satellite systems. Fires below a forest canopy may evade detection as well, and underground peat fires are especially hard to identify from space. NASA cautioned that the apparent count of satellite detections can fall during severe smoke episodes precisely because the view of the surface becomes more obscured.
Indonesia’s Ministry of Forestry fire-monitoring platform, SiPongi, recorded 946 hotspots on August 31, NASA said. That number is a measure of identified hotspots, not a complete census of all fires or of the land burned. It must be read alongside the sensing limitations described by NASA. The published imagery is therefore strong evidence that fires and smoke were present over the observed areas, but it cannot by itself settle the full extent, intensity, duration or underground reach of the burning.
That distinction matters for authorities and communities dealing with haze. A count derived from satellite heat signals can help direct attention and resources, yet it cannot provide a direct measure of exposure for every community, nor can it establish how much smoke any particular person inhaled. NASA reported that Indonesian officials had warned of hazardous smoke exposure across broad areas, while disruptions were already affecting daily routines.
Drought and climate patterns raise the fire risk
The early-August rainfall shortfall described by NASA formed the proximate backdrop to the fires. Dry weather can reduce water levels in peatlands and leave deposits that are ordinarily too wet to sustain deep-burning fires more combustible. NASA’s account said that peat areas in Kalimantan, Sumatra and Papua generally do not allow fire to spread underground when conditions are wet, but can do so during drought.
NASA also linked the season’s conditions partly to strong El Niño conditions. A Columbia University researcher cited in the agency’s account said fire activity was rising in a pattern resembling 2015 and attributed part of the unusual dryness to El Niño. The agency said the climate pattern had been assessed as present and strengthening in August, and that it typically brings sharp reductions in Indonesian rainfall. The same account noted a positive Indian Ocean Dipole as another factor present in the regional climate setting.
Climate patterns can help explain why a season is unusually dry without accounting for every fire’s origin or path. The information supplied by NASA does not identify ignition sources for the individual detections seen on September 1. Nor does it demonstrate that El Niño alone caused the 2026 fires. The reported relationship is one of heightened dryness and heightened fire vulnerability, particularly in landscapes where dry peat permits combustion to move beneath the ground.
Indonesia holds about 36 percent of the world’s tropical peatlands, NASA said, giving the issue weight far beyond the footprint of any one satellite image. The scale of the country’s peat area means dry-season fire management is closely tied to the condition of wetland soils, rainfall and the ability to find burning that may be concealed by vegetation, smoke or the ground itself.
2015 remains the comparison, not a forecast
The developing 2026 season has drawn comparison with Indonesia’s major 2015 fire episode, an earlier El Niño year that NASA described as among the most extreme periods of burning in recent decades. In its account, NASA said the 2015 fires burned for more than three months and released 1.75 billion tons of greenhouse-gas equivalents. It reported that, as of September 2, the 2026 fires had released roughly 10 percent of the greenhouse-gas-equivalent emissions attributed to the 2015 fires.
That comparison gives a sense of why scientists are watching the fires closely, but it is not a projection that the current year will reproduce 2015. The two figures refer to different points in time and different durations of burning. NASA’s estimate for 2026 was made around a month into the fire period described in its article, while the 2015 total followed more than three months of fires. Weather changes, rainfall, fire-control efforts and the progression of underground combustion could all alter the season’s eventual outcome.
Peat fire emissions also differ from emissions from other tropical forest fires. NASA cited an estimate finding that peat fires can generate about three times as much fine particulate matter, five times as much sulfur dioxide, three times as much organic carbon, and twice as much methane and carbon monoxide. Such comparisons explain why peat combustion can have outsized implications for air quality and atmospheric pollution even where flames are difficult to see.
Still, these are estimated relative emissions cited in NASA’s article, not measurements from each active Indonesian fire in 2026. They should not be treated as a direct calculation of the pollution coming from the fires visible in the September 1 image. The supplied information also does not include local air-quality readings, hospital data, a verified total area burned, or a comprehensive emissions estimate for the season beyond the reported comparison with 2015.
Closures and disrupted routines show the haze burden
NASA said the smoke was already producing operational and social consequences on the ground. Officials had warned that large parts of the population were exposed to hazardous haze, some schools had begun moving to remote learning, nine national parks had closed, and several flights had been delayed because of heavy smoke. These measures indicate that the effects had moved beyond remote peatlands into education, travel and access to public places.
The closures and delays should not be read as a complete inventory of the disruption. The available report does not specify which schools changed their arrangements, how long the national parks would remain closed, which flights were delayed, or the number of people affected. It also does not provide a province-by-province account of smoke conditions. What it does establish is that NASA reported a broad enough haze problem for Indonesian officials and institutions to take protective or operational action.
The challenge facing monitoring efforts is therefore twofold. Fire managers need to follow visible surface activity and smoke plumes while also accounting for hidden combustion in drained or dried peat. NASA said Indonesian authorities use observations from MODIS and VIIRS sensors to track active fires in near real time. The agency’s account also described efforts to improve detection of understory fires with other satellite observations, reflecting the known shortcomings of relying on any single sensor or a clear view from above.
The next consequential change will be in moisture conditions. The NASA account said underground peat fires can keep burning until rains return, with October or November identified as the expected period for rain in the cited assessment. That is conditional, not a confirmed timetable for extinguishment. Rainfall may alter fire behavior, but the source material does not establish when, where or how fully it would end the present burning.
This report is based on NASA’s published account and the satellite and monitoring information it summarized. The reported fires, hotspot count, emissions comparison, rainfall figures and disruptions have not been independently corroborated for this article. The limits of satellite detection, the absence of a complete ground-level assessment in the available material, and the evolving nature of a fire season leave important uncertainty about the total scale of the 2026 event.
