Reading Defra background pollutant maps for an assessment

Background pollutant concentration maps are the starting point for the baseline in almost every air quality assessment in England, and the two mistakes that recur are choosing the wrong grid square and double counting a road that is already in the map. Defra publishes the maps on its UK-AIR service at one kilometre resolution, modelled for a base year and projected forward. Getting the selection, the adjustment and the version reference right takes an afternoon. Getting any of them wrong undermines every concentration in the report.

Where the background maps live and what years they cover

The maps sit in the local air quality management section of the UK-AIR website. They are published as downloadable datasets giving annual mean concentrations by one kilometre grid square for nitrogen oxides, nitrogen dioxide, particulate matter as PM10 and PM2.5, and several other pollutants.

Each release has a base year, which is the year the modelling was calibrated against measured data, and then a series of projected years running forward from it. A release is superseded when Defra publishes a new base year, and the values for a given square can move between releases, sometimes appreciably.

Two habits follow. Download the dataset rather than reading values off a web page, so that you have a fixed copy of what you used. And check whether a newer release has appeared between commissioning and reporting, because a reviewing officer or a third party consultant will be working from the current one.

Selecting the right grid square

Use the receptor, not the site.

A one kilometre square is large enough to contain very different environments, and an assessment with receptors on both sides of a main road may legitimately draw on two or three different squares. Take the easting and northing of each individual receptor, look up the square containing it, and record the grid reference alongside the concentration in the receptor schedule.

Where a receptor sits close to a square boundary, say so and test the sensitivity of the result using the adjacent square. That single paragraph removes an easy line of attack.

For proposed development the receptor positions are drawn from the layout, which means the air quality work cannot be finalised before the layout is. Where a Transport Assessment is also feeding traffic flows into the model, sequence the two so that the traffic data and the receptor positions come from the same version of the scheme.

Why the maps are projected forward and what that implies

Defra projects the maps forward using national emissions projections, which assume fleet turnover, cleaner vehicles and reductions from other sectors over time.

Those projections have historically been optimistic about the pace of decline in roadside nitrogen dioxide, and measured concentrations have on occasion fallen more slowly than modelled. Professional guidance therefore counsels caution about relying on projected future-year backgrounds to deliver compliance on their own.

The practical response is a sensitivity test. Run the assessment year on the projected background and again on the base year background, and report both. If the conclusion only holds on the projected figure, say so plainly rather than presenting a single number. An assessment that acknowledges this reads as competent. One that quietly banks the projection invites a challenge that is easy to make.

Combining background with modelled road contributions

The background map value is the sum of contributions from all sources affecting that square, and that is the point at which double counting creeps in.

If you model a road explicitly in a dispersion model and then add the full background map value to the modelled road contribution, you count that road twice, because the map already contains a contribution from the major roads within the square. The result overstates total concentrations and can push a receptor over an objective that it does not in fact exceed.

Defra addresses this directly. Alongside the background maps it publishes sector contribution data that lets you extract the contribution to a given square from specified road sources, so that it can be removed before the modelled contribution is added back. The process is straightforward: identify which of the roads in the square you are modelling explicitly, extract their contribution from the mapped value, and document what you removed.

Double counting and how to avoid it

A short procedure that survives review.

1. List every road source modelled explicitly, with its road link reference.
2. For each receptor's grid square, extract the mapped contribution attributable to those sources.
3. Subtract that contribution to give an adjusted background.
4. Add the modelled road contribution to the adjusted background.
5. Record the unadjusted value, the amount removed and the adjusted value for each square in an appendix table.

Step five is what separates a defensible assessment from one that has to be redone. Reviewers cannot check arithmetic they cannot see, and an appendix table of three columns removes an entire category of query.

One caution. Only remove sources you have actually modelled. Removing a motorway contribution and then failing to model the motorway leaves a hole in the baseline, which understates concentrations and is the mirror image of the original error.

Model verification against monitoring data

Verification is the step that turns a modelled road contribution into a number anyone will accept.

The method compares modelled concentrations at monitoring locations against the measured values from those monitors, usually diffusion tubes and any automatic analyser in the study area, and derives an adjustment factor applied to the modelled road contribution. Local air quality management technical guidance sets out the procedure and the diagnostics.

Three things to check in someone else's verification. Whether the monitoring sites used are representative of the modelled locations, because verifying an urban canyon model against a suburban background tube proves little. Whether diffusion tube bias adjustment has been applied and which factor was used. And whether the adjustment factor is reported explicitly rather than buried, along with the number of sites it was derived from.

Where no local monitoring exists, say so and explain what was done instead. An unverified model is not worthless, but its uncertainty has to be stated.

Recording the map version in the report

Put the provenance in one place, early, and in a form someone can reproduce.

The minimum is the map release identified by its base year, the date the dataset was downloaded, the grid square reference for every receptor, the assessment year used, the sector contributions removed with the road links they relate to, and the verification factor with the monitoring sites behind it.

That paragraph does more for the credibility of an Air Quality Assessment than any amount of description elsewhere in it. It also makes the report usable two years later, when the scheme changes and someone has to decide whether the baseline still stands.

Reviewers checking this kind of work will look for exactly those items, and a reviewer signing off a draft can only do so if the assumptions are visible. The professional directory is the route for arranging that review, and the wider set of supporting documents sits in the report index.

Enter your postcode on PlanScope to see which reports a site is likely to need. This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it. 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.