On 27 September 2026 the NOAA-20 VIIRS sensor logged 6 fire detections with radiative power at or above 20 MW inside the watched boundary of Virunga (COD), tile 2/2, over a 2-day window, with the strongest at 117.9 MW. That is dense for the area: 4.68 hot detections per deg² against 2.136 in the surrounding 30 km ring, a ratio of 2.191. A later pass on 29 September picked up 3 hot detections again, which is the independent confirmation. Counts kept climbing on the same rolling window: 19 detections (max 126.5 MW) by 29 September and 23 detections (max 126.5 MW) by 30 September.
What this is not: heat is not clearing, and the rule cannot separate a wildfire from managed, agricultural or legally prescribed burning. The region sits in an El Niño phase with an ONI of 1.8, a setting that favours drier fuel. The boundary used is the watchlist bounding box, not the legal park polygon, so some detections may fall outside the protected area. No natural-event record is logged nearby.
What would change our mind: a burn-permit or prescribed-fire record, a corrected boundary placing the cluster outside the park, or clear-sky imagery showing cropland rather than forest.
Written by AI, then checked by a separate reviewer before publication. Every claim points at the evidence below.