Methodology

Every number the platform surfaces is derived from published, peer-reviewed methods with explicit uncertainty. This page is the reference for the parameters and equations behind detection, attribution, quantification and CO₂-equivalence conversion.

Detection

A detection is a Sentinel-5P TROPOMI XCH₄ column enhancement above the local background baseline, associated with an overpass within a monitored facility's bounding box. We use the operational Copernicus Level-2 CH₄ product from the Copernicus Data Space Ecosystem (CDSE), filtered on the vendor-supplied quality-assurance flag (qa_value ≥ 0.5). Baseline is computed as the trailing 90-day median column at the same location, following common practice in the TROPOMI methane literature (Lauvaux et al. 2022, Science).

Attribution

Search radius: 25 km around each detection centroid.
Time window: ±6 h around the overpass timestamp.
Confidence threshold: 0.60 — detections below this remain unattributed and are surfaced for analyst review.
Scoring: a weighted combination of (a) inverse distance from the plume centroid to each candidate facility, (b) alignment of the plume major-axis with the wind-vector back-trajectory at the overpass hour, and (c) asset-type prior (well, tank, compressor, pipeline, terminal).

Quantification

We use the Integrated Mass Enhancement (IME) method as formulated by Varon et al. (2018, Atmospheric Measurement Techniques):

Q = U_eff · IME / L

where IME is the plume-integrated mass excess of methane (kg), L is a characteristic plume length scale (m), and U_eff is the effective transport wind (m/s). We adopt Varon's empirically calibrated scaling U_eff = 1.4 · U10, with U10 taken from ERA5 reanalysis (via Open-Meteo archive) at hourly cadence on a 25 km grid.

Reported 1-σ uncertainty on Q is ±50%. This reflects the joint uncertainty of column retrieval, plume delineation and effective wind — consistent with the ranges reported in Varon 2018 and Sherwin et al. 2024 (Nature) blind-test results.

CO₂-equivalence conversion

82.5× (20-year GWP) — IPCC AR6 WG1, Table 7.15, fossil methane.
29.8× (100-year GWP) — IPCC AR6 WG1, Table 7.15, fossil methane.

Reports state which horizon is applied. The 100-year value is the default for regulatory submissions consistent with UNFCCC / EU practice. The 20-year value is available for climate-risk framings that emphasise near-term forcing.

Reconciliation

Measured mass over a reporting window is the time integral of quantified rate over that window, per facility. It is reconciled against the latest operator-declared inventory row for the same facility and prorated to the scan window length. Deltas exceeding 25% are flagged; deltas exceeding one order of magnitude are escalated as critical.

Limitations we are honest about

TROPOMI's detection floor is instrument-limited to roughly 5 t/hr in favourable conditions; below that we depend on higher-resolution sensors as they come online.
Cloud cover, high solar zenith angles and low surface albedo (e.g. water, snow, dense forest) degrade retrieval quality. We surface qa_value alongside every detection.
Attribution is probabilistic. In dense infrastructure (mid-stream corridors, offshore fields) the correct answer may be a facility cluster, not a single asset.
Wind reanalysis at 25 km cannot resolve local topographic or platform-wake effects. This is the dominant term in the ±50% quantification uncertainty.

References

Varon D. J. et al. (2018) Quantifying methane point sources from fine-scale satellite observations of atmospheric methane plumes. Atmos. Meas. Tech. 11, 5673–5686. doi:10.5194/amt-11-5673-2018
Lauvaux T. et al. (2022) Global assessment of oil and gas methane ultra-emitters. Science 375, 557–561.
Sherwin E. et al. (2024) Single-blind test of nine methane-sensing satellite systems. Nature.
IPCC AR6 WG1 (2021), Chapter 7 and Table 7.15.