Accountability — CERES Analysis Tools

Accountability — Where top-down meets bottom-up. Enter run · d drain · j/k navigate gaps · deep link: ?cell=31.75N_102.25W&species=CH4

Pending Monitor

No events yet

Stream Log

Waiting for pipeline run…

Accountability BETA

This tool asks a simple question: do the greenhouse gas emissions that US facilities report to regulators match what satellites actually observe from space? Where they don't, it flags the gap — and estimates how confident that finding is.

Press Understood to go straight to the tool, or read on.

How it works

The pipeline runs three independent data streams simultaneously — bottom-up facility reports, satellite column observations, and top-down energy budget analysis — then joins them at the same 0.5° grid cell (~55 km) that the CEDS emissions inventory uses.

For each grid cell, a Reconciler compares the total mass reported by all facilities against the mass estimated from satellite column density. A gap beyond the satellite's measurement uncertainty triggers a discrepancy. If the CERES energy budget independently shows anomalous forcing at the same location, confidence is upgraded from MEDIUM to HIGH. This three-way corroboration — facility report, satellite column, energy budget — is the core of the accountability model.

Data sources

EPA GHGRP (bottom-up · Tier 1)
US facility self-reported emissions under the Greenhouse Gas Reporting Program. ~21,000 facility-gas records for 2023, covering CO₂, CH₄, N₂O and SF₆ for facilities emitting above 25,000 tCO₂e/year — roughly 50% of total US emissions.

TROPOMI / Sentinel-5P (satellite · Tier 2)
ESA Copernicus satellite methane column density (XCH₄), exported monthly from Google Earth Engine at 0.5° resolution. Background-subtracted to isolate local emission anomalies above the global mean. 2023 annual coverage.

CERES EBAF (top-down · Tier 3)
NASA CERES Energy Balanced and Filled TOA flux data. Cells showing anomalous net energy imbalance (σ ≥ 1.0 over a 10-year baseline) within the US bounding box are used to independently corroborate bottom-up discrepancies.

A note on the numbers

The gap percentages shown — some exceeding 1,000% — reflect a real and known limitation: EPA GHGRP captures only facilities above the reporting threshold. TROPOMI sees all methane in the column, including diffuse agricultural sources, landfills, pipeline leaks, and the thousands of smaller operations below the 25,000-tonne threshold. The gap is not evidence of fraud — it is evidence of structural under-counting that is well-documented in the scientific literature.

We invite you to examine these numbers critically. Cross-reference HIGH-confidence gaps with known industrial regions, check whether the seasonal heatmap shows the signal persisting across months or appearing only in winter (suggesting atmospheric stability artefacts), and compare reconciled cells against cells with large gaps to understand what the satellite is and isn't capturing. The data and the full audit trail — every event, every gate decision — are available in the Stream Log.