What is minimum detectable emission rate in satellite methane monitoring?
If you've spent any time reading sensor spec sheets for methane monitoring, you've run into the term minimum detectable emission rate, usually shortened to MDL or MDR. It sounds like a single clean number. It isn't, and knowing why matters if you're deciding whether a given satellite product can actually catch the leak you're worried about.
What MDL measures
Minimum detectable emission rate is the smallest plume, expressed in mass per hour (typically kg/hr), that a given sensor can reliably pick out from background noise under stated conditions. The key phrase is "under stated conditions." MDL is a function of four things working together: the sensor's noise floor in the shortwave infrared, the spatial resolution (ground sample distance), the wind speed at the time of overpass, and the surface albedo underneath the plume.
Methane absorbs light in specific SWIR bands. A plume shows up as a dip in reflected radiance at those wavelengths relative to the surrounding pixels. A sensor detects that dip when it's bigger than the instrument's own measurement noise. Push the wind speed up and the same emission rate gets diluted into a fainter, more spread-out signal, so the detection floor rises. Drop the albedo (water, dark soil, certain vegetation) and the available light signal shrinks, so the floor rises again. That's why you'll see the same instrument quoted with MDL ranging from well under 100 kg/hr on a calm day over bright desert to several hundred kg/hr over a darker, breezier scene.
Why MDL differs so much between sensor types
This is where a lot of confusion comes from when people compare products. Wide-swath multispectral sensors not built specifically for methane, repurposed from mineral or agricultural bands, tend to have MDLs in the high hundreds to low thousands of kg/hr. They're good for finding the biggest, most obvious releases but will miss a mid-size leak entirely.
Dedicated imaging spectrometers, built with narrow, well-calibrated bands tuned to the methane absorption signature, push that floor down substantially. Published figures for current point-source imaging spectrometer missions generally land somewhere in the tens to low hundreds of kg/hr range under favorable conditions, climbing from there as wind and albedo get worse. Pixel size matters too: a 0.5 m to 2 m ground sample distance lets the plume's spatial shape help confirm a detection that a coarser pixel would average away into noise.
A smaller advertised MDL doesn't automatically win the comparison. A sensor with a lower nominal floor but infrequent tasking over your acreage is less useful for MRV reporting than one that revisits your sites often enough to catch an intermittent release in the act.
MDL versus attribution: two different problems
MDL answers "could this sensor see a plume this size." It doesn't answer "whose plume is this." Those are separate questions, and conflating them is how a lot of MRV workflows get stuck. A regional advisory can confirm something above a basin-scale threshold is venting somewhere in a county. That's a detection statement. Tying the same SWIR signature to a specific pad, compressor station, or tank battery boundary is an attribution problem, and it needs per-site imagery, not just a sensitivity spec.
If your job is closing out a monthly MRV report or screening a portfolio of assets for a fund, the practical question is simpler than a spec sheet makes it look: can you point at this specific site and get an answer today? That's the gap Methane Plume Map is built around: per-site plume imagery that resolves the SWIR signature of an individual emitter and ties it back to a site boundary, rather than a basin-wide advisory you then have to go chase down yourself.
What this means when you're screening a portfolio
When you're evaluating a monitoring product, ask for the MDL under the wind speed and albedo conditions you operate in, not the best-case number on the data sheet. Ask how often the sensor revisits your sites, since a low MDL on a satellite that passes over once a year won't catch an intermittent release. Separate the detection question from the attribution question explicitly: a lot of vendors blur the two to make their numbers look better than they are in practice.
If you're ready to see what a per-site plume map looks like for your own acreage, get in touch and we'll walk through what Methane Plume Map can show for your sites.