How to attribute a methane plume to a specific well pad

A regional advisory lands in your inbox naming a 25 km grid cell with an elevated methane signal. You operate four pads in that cell. The advisory doesn't say which one, and your fund's LPs or your ops team on call are both asking the same question by lunchtime.

Attribution is the step between "something is emitting in this area" and "this specific pad, this specific tank battery, is the source." It takes a few deliberate checks, and a coarse sensor can't get you there on its own.

Why area-level detections stop at the basin

Most public methane alerts come from sensors built for broad coverage, not site discrimination. A pixel footprint of a kilometer or more can hold several pads, a compressor station, and a pipeline right-of-way all in the same cell. The detection tells you the cell lit up. It can't tell you which lease line the source sits inside, because the pixel is bigger than the distance between your pads.

Getting from "this cell" to "this pad" needs two things the basin product usually doesn't carry: a source point resolved at a scale smaller than the gap between facilities, and a site boundary layer to match that point against.

The actual attribution workflow

1. Resolve a source point down to the pad. Imaging spectrometry reads the SWIR absorption signature of the plume itself, which locates where the gas is actually leaving the ground rather than where a wide pixel happened to average out. At 0.5 to 2 m resolution, a source point sitting on your pad reads differently from one sitting on a neighbor's, even when the two are a few hundred meters apart.

2. Overlay the site boundary. Once you have a source coordinate, check whether that point falls inside your pad's lease polygon, your neighbor's, or the right-of-way between them. Plume shape drifts downwind, so a sloppy overlay can put a source a hundred meters from where it actually left the ground, enough to land it on the wrong side of a lease line.

3. Check the plume orientation against wind. A plume stretches downwind from its source rather than sitting in a neat circle over it. Confirm the wind direction at the time of overpass lines up with the plume's long axis pointing back to your boundary before you call a source point final.

4. Look for repeats, not just a single pass. One detection on one day could be a transient event, a blowdown, a stuck valve that got fixed an hour later. A second detection on the same pad during a later pass rules out a one-off event and points you toward something worth a crew visit. Daily tasking, where the sensor schedule allows it, turns a one-off flag into a pattern you can act on or rule out.

5. Rule out adjacent infrastructure. Pads near gathering lines, meter stations, or a neighbor's battery need an extra look. If the source point sits near a boundary line rather than deep inside your polygon, don't round up. Flag it as ambiguous and wait for the next pass.

None of this replaces ground truth. A site visit or an OGI survey still confirms the actual component. What attribution does is tell your team, or your analysts covering a portfolio, where to send that visit first instead of checking four pads because a regional bulletin named a county.

Where an annual flyover falls short

Even a well-flown annual survey has the same problem in a different form: it's a single snapshot. A pad that emits in March and is fixed by the time the plane comes through in October never shows up in that year's record. Catching the emission while it's still happening is as much a cadence question as a resolution one, and resolving the right pad only matters if you catch it in time to act.

A per-site plume map built for pad-level attribution answers "is this pad emitting today," which is a different question than "did this basin have an event last quarter." You can see how FlyPix frames that use case on the Methane Plume Map home page.

Point it at the pad in question and see what today's pass shows.