Public-health interpretation boundary
AQ26 keeps emissions, ambient concentrations, exposure and health outcomes separate. WHO-style pollutant benchmarks are used only as health-context prompts; they do not identify the source of pollution.
WHO-style pollutant benchmark scope
fine particulates / proxy evidence context
benchmark_ready_no_attributionparticulates; UKHSA notes PM10 can be a proxy for other emissions in some evidence
benchmark_ready_no_attributioncombustion-related nitrogen oxides context
benchmark_ready_no_attributionacid gas / combustion context
benchmark_ready_no_attributionsecondary pollutant context, not direct stack marker
benchmark_ready_no_attributioncombustion-efficiency context
benchmark_ready_no_attributionchlorinated/fluorinated acidic gases
permit_matrix_neededabatement/slip and secondary particulate chemistry context
permit_matrix_neededmetals including lead, zinc, chromium, mercury and others where reported
permit_matrix_neededrequires specialist sampling, stack/permitting and biomonitoring evidence
manual_review_requiredBenchmark cards are public-health context only. They are not source-attribution findings.
Source-attribution caution
The automated site must not collapse these evidence classes into a single causal claim.
UKHSA-style balance
The public output is framed to support evidence review and challenge. It does not override regulator, public-health or permit determinations, and it does not assert that a modern regulated facility is or is not a significant public-health risk.
