Amazon analysis finds more candidate fire detections than official counts
The story at a glance
NASA reports increased fire activity and smoke emissions in the central and eastern Amazon from September 2026. Its analysis found about 40% more candidate fire detections around Santarém than in FIRMS active-fire counts.1
The comparison concerns signals below the threshold for official FIRMS products, not a verified count of additional fires. The report does not quantify any extra burned area or forest damage associated with them.1
The story in brief
What you need to know
- NASA says the candidate detections came from VIIRS observations around Santarém. They were not included in FIRMS active-fire counts because their signals did not meet the threshold for those official products.1
- NASA says smoke, clouds or canopy cover may help explain why some fire signals are missed. The reported 40% difference is a comparison of candidate detections with FIRMS counts, not confirmation of 40% more distinct fires.1
- NASA also cites an estimate of 50 to 70 uncontrolled fires in the Santarém region. Its account describes increased activity from September, while a separate report from the Associated Press says fire counts across the wider Amazon had been near historic lows during 2026 before tripling after 4 October. These reports refer to different periods and geographic scales.14
- The Associated Press reported that smoke disrupted daily life in Manaus. It also said Brazilian authorities were investigating whether some recent fires had been deliberately set, but who was responsible and why remained unclear.4
- The reported detections should not be treated as a measure of additional forest loss. NASA describes them as candidate signals, and its account does not give an associated estimate of burned area or damage.1
Conservation context
Why this matters
Fire counts depend in part on which signals meet a satellite product’s detection threshold. The findings highlight a potential gap in official counts, but do not establish how many distinct fires were missed or how much forest they affected.1
The developing story
What happens next
A key unresolved step is whether NASA or its partners will validate and publish the locations and number of candidate detections, including how many represent distinct fires. The available reporting also leaves the course of the 2026 fire season uncertain.1
Established evidence
What we know
- A 2013 study mapped understory fires in southern Amazonia from 1999 to 2010. It estimated that they affected more than 85,500 square kilometres, or 2.8% of forests in the study area. This is historical context, not an estimate for 2026.2
- A 2017 study using satellite data from six El Niño and six La Niña events between 1997 and 2016 reported higher average pan-tropical forest fire emissions during and following El Niño. It provides historical context, not evidence of the cause or future course of the current fires.3
- The Associated Press reported smoke impacts in Manaus and an investigation into possible deliberate ignition. It said the responsibility and motives behind the fires were not clear.4
Open questions
What we don't know
- The reporting does not establish how many candidate detections correspond to distinct fires or whether they will be validated and published.1
- The additional burned area, if any, associated with the candidate signals has not been quantified in the supplied reporting.1
- It remains uncertain how the 2026 fire season will develop or whether the regional seasonal pattern described by NASA will be borne out.1
- The available reporting does not establish who started the fires under investigation or their motives.4
Source coverage
