Provenance
Monitoring evidence is retained as structured station/day records with source records, downloadable evidence and release context. The accepted 30-day product contained 9,696 evidence rows.
Environment
England Air-Quality Evidence Observatory
Evidence-led monitoring and transparent review of England air-quality data, with provenance, limitations and reproducible release controls.
Continuous evidence system
A weekly evidence cycle with public and protected layers.
AirQuality26 repeatedly harvests and re-checks monitoring evidence rather than publishing a static snapshot. Source/integrity checks, validation gates, public-surface reconciliation, review packs and release/archive controls form the operating cycle. The public observatory is bounded for responsible publication, while protected/unredacted review material can be retained separately where unrestricted publication is inappropriate.
Recurring automationValidation & reviewVersioned evidenceProtected/public separation
See the evidence activity record →Purpose
What it is trying to create.
Create a public evidence surface that makes official monitoring data easier to inspect without turning descriptive screening into unsupported source attribution or health conclusions.
Intended outcomes
What success looks like.
- Clear public monitoring summaries with explicit scope and limitations.
- Downloadable evidence and provenance suitable for independent checking.
- Automated validation gates before publication.
- A repeatable framework for longitudinal policy and environmental review.
Evidence status
Provenance · integrity · verification · validity
Integrity
Publication is gated by limitation statements, source/integrity checks and public-surface reconciliation. The project explicitly avoids unsupported source attribution, health findings or regulatory conclusions.
Verification
Accepted panel: 238 valid NO₂ station-days of 240 expected (99.2% completeness), 8 reporting stations across 8 of 9 English regions; station-panel coverage 80% and regional coverage 88.9%.
Validity
Valid for descriptive monitoring and contextual comparison within the stated panel/window. It is not presented as causal attribution, regulatory compliance assessment or individual health diagnosis.
Evidence record: Accepted 30-day product: 22 Jun–21 Jul 2026
Public boundary
Transparency without exposing operational secrets.
Public pages describe purpose, evidence principles and intended outputs. Credentials, private repository links, source-specific collection tactics, restricted implementation details, anti-abuse controls and other security-sensitive material are deliberately excluded.