El Niño-driven drought is exposing a larger Amazon fire emergency
A severe drought tied to El Niño is driving a fast-moving wildfire crisis across the tropics, and the Amazon rainforest is now absorbing one of its most dangerous blows. NASA satellite monitoring has picked up a sharp rise in fires and smoke across central and eastern Amazonia beginning in September 2026, with activity spreading near Manaus and Santarém and extending toward the mouth of the Amazon River. In practical terms, the fire season is not confined to isolated hotspots; it is unfolding across a vast forest system already stressed by prior burn events.
The public health burden is immediate. In Manaus, thick smoke pushed fine particulate matter, or PM2.5, above 135 micrograms per cubic meter, a level associated with hazardous air pollution. The city’s air quality fell into the second-worst category of the regional index, schools closed, public events were canceled, and children’s sports were suspended. For a city of more than 2 million people, the smoke is not a distant environmental indicator. It is a direct respiratory threat affecting daily movement, schooling, and outdoor work.
Smoke plumes and satellite systems are tracking the burn front
On October 12, 2026, MODIS aboard NASA’s Aqua satellite captured enormous smoke plumes rising from fires on both sides of the Amazon River. Those plumes were visible in multiple monitoring platforms, including the Fire Information for Resource Management System, the Global Fire Emissions Database, and NASA Worldview. The geography matters because the active burn zone sits between major Brazilian commercial and transport centers and the eastern edge of the basin, where forest, road access, and land-use pressure often overlap.
Fire in tropical forest behaves differently from fire in dry grassland or temperate woodland. The Amazon normally holds enough moisture to suppress ignition, which is why lightning accounts for less than 1 percent of fires in Brazil’s rainforest. Most of the current blazes began as human activity: land clearing, removal of shrubs on grazing land, and burning crop residue. Once drought strips the understory of moisture, those managed burns can escape containment and move into the forest floor, where they become difficult to stop and even harder to see.
By late August and September, rainfall around Manaus had already collapsed far below normal, with precipitation reported at 95 percent below normal in August and 60 percent below normal in September. A forest that is usually humid enough to suppress flame becomes vulnerable when the understory dries out. In that setting, a small ignition can turn into a large fire front with little warning.
The Amazon is repeating a pattern seen in previous El Niño years
Douglas Morton of NASA’s Goddard Space Flight Center described the present sequence as a tropical fire cascade driven by rainfall disruption. The pattern begins near the equator, where drought first intensifies peat fires in Indonesia, then expands as dry weather develops elsewhere. As the chain reaction moves westward, the Amazon becomes the next major region to catch. Morton’s comparison is not speculative; it draws on prior El Niño periods, including 2015–2016 and 2023, when similar transitions occurred across the tropics.
One sentence captures the scale of the current divergence from normal conditions: the Amazon is not facing a routine fire season.
Forests that burned during prior El Niño episodes are especially at risk now. Many of the areas threatened in 2026 already experienced fire in 2015 or 2023, which leaves trees weakened, ground fuels altered, and canopy gaps more likely to admit heat and wind. In rainforest ecology, repeated burning is especially destructive because most species are not adapted to frequent fire. Unlike fire-prone ecosystems, the Amazon loses resilience after each burn, and the second or third passage of flame can do more damage than the first.
Hidden understory fires suggest the count may still be too low
NASA’s current analysis suggests that standard monitoring may be missing a meaningful share of active fires. Shane Coffield, an assistant research scientist at the University of Maryland working at NASA Goddard, is using observations from VIIRS, the Visible Infrared Imaging Radiometer Suite, to find candidate fires that do not meet the detection thresholds used in official FIRMS data products. In the Santarém area alone, the initial findings suggest about 40 percent more fires than appear in active fire counts from FIRMS.
This is not a trivial technical difference. Fire detection from orbit depends on infrared heat signals, and those signals can be blocked by smoke, cloud cover, and the dense canopy itself. Slow-burning understory fires are especially elusive because the hotter material sits under vegetation that masks the signal. On the smokiest days, MODIS and VIIRS can show very few detections even when the ground is still burning. For researchers, land managers, and civil authorities, that means the visible fire count may lag behind the actual footprint of damage.
Coffield’s work is aimed at passing those additional detections, called candidate fires, to MapBiomas, a collaborative network that tracks fires and land-use change across the Amazon. The value is straightforward: a more complete fire inventory gives a better estimate of where the basin is under active stress and where suppression resources or community alerts may be needed next.
Fire danger may climb northward before shifting south again
The current surge may be only one stage of a longer season. Morton and colleagues expect fire activity to expand northward during November and December 2026, reaching Roraima, Amapá, northern Pará, and parts of Venezuela, Guyana, and Suriname. Later, the danger is expected to shift southward, with drought and wildfire risk peaking again in the southern Amazon regions of Brazil and Paraguay by August and September 2027.
That forecast matters because it extends the operational window far beyond a single smoke event. Communities farther north and south still have time to limit ignition, especially by avoiding land burning during the driest months. Morton’s warning is blunt: during a strong El Niño, burning should be avoided at all costs. In a rainforest already primed by drought, one escaped fire can seed a broader emergency that outlasts the season in which it started.
The current Amazon crisis is therefore both an environmental and health story. It is about fire ecology, but it is also about smoke exposure, school closures, hazardous air, and the inadequacy of present-day monitoring when flames hide beneath the canopy. The next few months will show whether the basin’s fire counts continue to rise faster than the instruments designed to measure them.