Using council diffusion tube data in an assessment

Local authority nitrogen dioxide diffusion tube data is the only measured air quality information most sites will ever have, and it arrives with four adjustments already applied or waiting to be applied. Use it raw and the verification is wrong. Use the council's published annual mean without checking what was done to it and you inherit somebody else's assumptions without knowing it.

The governing document is Defra's local air quality management technical guidance, and almost every question below has an answer in it. What follows is how the guidance plays out on a real assessment.

Where the data lives and how to request it

Every local authority reviewing air quality publishes an annual status report to Defra. That report contains the monitoring network, the site identifiers and coordinates, the site classification, monthly and annual results, the data capture percentage, the bias adjustment factor applied and its source, and any annualisation.

Start there rather than with a data request, because the report often answers the question. Where you need the monthly raw values, the site photographs or the exact tube position, ask the authority's environmental health or pollution control team directly and allow time, since response speeds vary considerably.

Ask for three things in the same email: the monthly unadjusted results for the relevant years, the site metadata including coordinates, height and distance from the kerb, and confirmation of the bias adjustment factor with its derivation. Getting all three at once saves a fortnight.

Bias adjustment and annualisation

Diffusion tubes do not measure the same as a reference analyser. They over-read or under-read depending on the laboratory, the preparation method and the exposure conditions, and the correction is a bias adjustment factor applied to the raw annual mean.

Factors come from one of two places: a local co-location study, where tubes are exposed alongside a continuous analyser in the same authority, or the national bias adjustment spreadsheet that pools co-location studies by laboratory and preparation method. A local factor is preferred where the study is sound and the sample size adequate. State which was used and for which year, because factors change annually and applying the current year's factor to an earlier year's data is a common error.

Annualisation is the second correction. Where a tube achieves less than the threshold for annual data capture set in the technical guidance, the period mean has to be scaled to represent a full year using the ratio observed at nearby long-term monitoring sites. Below a lower capture threshold the data should not be used at all.

Two practical points. Annualisation using a single reference site is weaker than using several, so report how many were used and where they are. And a tube with poor capture in the winter months is a different problem from one with poor capture in summer, because nitrogen dioxide has a seasonal profile.

Choosing comparable monitoring locations

The site classification is not a label, it is a statement about exposure geometry. Kerbside, roadside, urban background, urban centre and suburban sites sit at different distances from traffic and experience different concentrations for the same emissions.

Match the classification to what you need the data for. Verifying a modelled roadside concentration needs roadside or kerbside tubes on comparable roads. Establishing background needs an urban background site away from the local road influence, and using the national background maps alongside it as a cross-check.

Look at the site photographs and the coordinates rather than trusting the label. Tubes get mounted on the nearest lamppost, which may be behind a bus shelter, above a taxi rank, or on the opposite side of a junction from the modelled link. Any of those makes the site unsuitable for verification even though the classification reads correctly.

Where a tube sits in a street canyon and your receptor does not, the comparison breaks down entirely. Say so rather than including it and hoping the regression absorbs it.

Distance correction to the receptor

Nitrogen dioxide concentrations fall with distance from the carriageway, steeply at first. Where the monitoring site and the receptor are at different distances, the measured value has to be corrected before the two can be compared.

The technical guidance provides a calculator that separates the measured concentration into background and road contributions, scales the road contribution to the new distance, and recombines them. It relies on a national relationship rather than a local one, so it works acceptably in open road conditions and much less well in canyons, at junctions and where a barrier or building line interrupts dispersion.

Report the input distances, the background used and the corrected value, and state where the correction was not applied because conditions made it unreliable. A corrected figure quoted without its inputs cannot be checked by anybody, which defeats the point.

Using the data to verify a model

Verification compares the modelled road contribution against the measured road contribution at the same locations, not total against total. Subtract the background from the measured value to get the measured road increment, take the modelled road increment at the same point, and derive an adjustment factor.

The reporting that makes a verification credible includes the number of sites used and why any were excluded, the derived adjustment factor, the fit statistics, and the residuals at each site after adjustment. An adjustment factor derived from two tubes is not verification, and an assessment that reports the factor without the underlying comparison invites the officer to ask for it.

Apply the factor to the modelled road contribution only, and then re-add the background. Applying it to the total concentration inflates the background as well, which is both wrong and usually conservative in the unhelpful direction.

Where the traffic inputs come from a separate document, confirm they match. Modelled flows should be the same figures as the Transport Assessment uses, on the same links, for the same year.

Reporting uncertainty honestly

An Air Quality Assessment that presents concentrations to two decimal places while relying on a tube with seventy per cent data capture is telling the reader something untrue about its precision.

State the data capture for each site used, the source and year of the bias adjustment factor, whether and how annualisation was applied, the number of verification sites and the spread of the residuals, and the background year and map version. Then present the results in a way that reflects that, with a stated range or a sensitivity where the answer sits close to an objective.

The honest position on verification is that adjustment factors outside a moderate range indicate a problem with the model setup or the input data rather than a correction to be applied. Where that happens, fix the model instead of scaling the output, and say what was changed.

When a site has moved or closed

Monitoring networks are not static. Tubes are relocated when lampposts are replaced, sites close when an area falls out of exceedance, and identifiers are reused for new locations, which is the trap.

Check the coordinates year by year rather than relying on the site identifier. A site reference that appears in three consecutive annual status reports may describe two different locations, and a trend built across that discontinuity is meaningless.

Where the nearest suitable site closed before your assessment year, options are limited. Use an older year with a stated background correction, use a more distant comparable site with the distance correction applied, or commission a short survey and annualise it. All three are acceptable if the reasoning is explained. Silently using a site that no longer exists is not.

For a reviewer signing off a draft, these are the checks worth running first, and PlanScope's professional directory is the route through which review requests reach a qualified reviewer. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required.

Browse all 66 PlanScope report types to see where air quality work sits alongside the rest of an application.