Night-time VIIRS passes over the flare field near 32.49N, 47.21E recorded 3 clusters of heat with fire radiative power of at least 5 MW on 5 or more of the last 30 nights (2026-08-28 to 2026-09-26), from 156 hot-night detections in all. The most persistent cluster was lit on 28 nights, peaking at 21.03 MW. Independently, 2 of the clusters fall on sites listed in the Earth Observation Group's VIIRS Nightfire 2024 annual flare summary, the top one 0.11 km away and credited with 0.303 BCM that year.
What this is not: the sensor sees heat, not the thing burning. A flare, a refinery furnace, a steel or cement kiln all look alike from orbit; only the persistence and the match to a catalogued flare site argue for flaring here. FRP is not volume — the band saturates on big flares — and the VNF match is an annual aggregate from 2024, proving the site is a known flare, not that this week's burn is the same source. Unlit venting would be invisible entirely.
What would change our mind: a cloudy or smoky stretch under-counts nights and can mimic a shutdown; methane data from S5P or EMIT, or an asset registry check, would sharpen or overturn the flaring reading.
Written by AI, then checked by a separate reviewer before publication. Every claim points at the evidence below.