Building a significance matrix in an Environmental Statement

A significance matrix in an Environmental Statement is a presentational device, not an arithmetic engine. It sets magnitude of change against receptor sensitivity and shows the resulting level of effect, and its value lies entirely in whether the two axes were defined before the assessment began. The Town and Country Planning (Environmental Impact Assessment) Regulations 2017 require the Environmental Statement to identify and describe the likely significant effects of the development, and the matrix is a way of getting from evidence to that conclusion in a form a reader can audit.

What it cannot do is replace the judgement. Two competent assessors using the same matrix on the same data can reach different answers, and the honest Environmental Statement explains why its answer is the one it reached.

Why significance is a judgement, not a calculation

The Regulations do not define a scale. They ask for identification and description of effects that reach the threshold of significance, and they require the Environmental Statement to be prepared by competent experts.

That framing matters. The matrix is a tool chosen by the assessor to structure a judgement, so the criteria behind it are the assessor's to set and to defend. Where topic guidance supplies criteria, from the landscape, ecology, air quality, noise or transport disciplines, those criteria should be used and cited rather than reinvented.

An Environmental Impact Assessment is a report type where a competent expert is legally required under the 2017 Regulations, so an AI-assisted draft is a starting point for that expert rather than a substitute for their judgement.

Defining magnitude of change for each topic

Magnitude describes the change the development causes, independently of who or what experiences it. Five attributes do the work.

Direction, meaning adverse or beneficial. Scale, meaning how much change relative to the baseline. Extent, meaning the spatial area affected. Duration, distinguishing construction from operation and temporary from permanent. And reversibility, which often separates a high magnitude construction effect from a high magnitude permanent one.

Define each magnitude band in terms a reader can test. "High" should mean something measurable for that topic: a stated percentage change in a modelled concentration, a loss of a defined proportion of a habitat parcel, a change of a stated number of decibels, a described change in the composition of a view. Where the discipline has published thresholds, use them.

Avoid bands defined by their own adjectives. "Medium magnitude: a moderate change" is a circular definition and a reviewer will say so.

Defining receptor sensitivity and value

Sensitivity is where the matrices diverge most between topics, because the disciplines really do treat it differently.

The landscape disciplines separate value, which reflects designation and recognised quality, from susceptibility, which reflects the ability of the receptor to accommodate the particular change proposed. Ecology typically expresses importance using a geographic frame of reference, from international down to site level. Air quality and noise work from the exposure and vulnerability of the people or habitats affected, and the reasoning in receptor sensitivity in an air quality assessment sets out how that is normally handled.

The error worth naming is the collapse of susceptibility into value. A nationally designated landscape may be highly valued and still able to absorb a particular change, and an undesignated residential street may be highly susceptible to a change in outlook. A matrix that only has one sensitivity axis hides that distinction, which is why the assessor has to say in the text which judgement the single axis represents.

The matrix and where it misleads

Four failure modes recur.

The first is false precision. A five by five grid implies twenty-five distinguishable outcomes, when in practice the useful distinctions are fewer and the boundaries between them are judgements.

The second is the cliff edge. Two receptors sitting either side of a band boundary receive different significance labels although their circumstances are nearly identical, and the text has to explain that rather than let the table assert it.

The third is double counting. Where value has already been used to set the magnitude band, using it again on the sensitivity axis inflates the result.

The fourth is cross-topic drift. "Moderate" in the noise chapter and "moderate" in the landscape chapter frequently mean different things, and the cumulative effects chapter then adds them together as if they did not.

Stating criteria before applying them

The methodology section should publish the criteria, in full, for each topic. Magnitude bands with definitions, sensitivity bands with definitions, the matrix itself, and the point on the output scale at which an effect is treated as reaching the regulatory threshold.

Agree them at scoping where there is a scoping opinion. An authority that has seen and commented on the criteria is far less likely to challenge the output, and a statutory consultee who objects to the criteria at scoping has done you a favour compared with objecting at determination.

Then apply them consistently. The assessment should read as though the criteria were written first, because they were.

Professional judgement and departures from the matrix

Departures are legitimate and sometimes necessary. A matrix cell may give moderate where the assessor's judgement, on the specific receptor, is minor or major.

When that happens, say so explicitly in the text for that receptor, give the reason, and keep the matrix output visible alongside the final judgement so the reader can see both. An unflagged departure is the single fastest way to lose the credibility of a whole chapter at an inquiry.

What does not work is departing in one direction only. If every departure reduces the level of effect, a reviewer will read the chapter as advocacy, and the topic specialist will be asked about it under cross-examination.

Keeping topic chapters consistent with one another

Run a cross-chapter review before the Environmental Statement is issued, with one person reading only the methodology and conclusion sections of every chapter.

That reviewer is checking four things: that the output scale uses the same words across topics, that the stated threshold for the regulatory test is the same, that construction and operational phases are treated on the same temporal basis, and that the cumulative chapter can actually combine what the topic chapters produced.

The baseline sections need the same treatment, because a significance judgement is only as good as the baseline it measures against. The discipline described in the baseline section of a landscape and visual impact assessment applies across the document, and so does the honesty expected in the survey limitations section of an ecology report.

One closing admission. The matrix is a communication device that makes a contestable judgement legible, and it works best when the Environmental Statement treats it that way rather than presenting it as a result. Readers trust a chapter that shows its reasoning more than one that shows a tidy grid.

To see which environmental reports a scheme is likely to need alongside the Environmental Statement, browse all 66 PlanScope report types.

This is general guidance, not legal advice; check the position with your local planning authority or a qualified professional before relying on it. For specialist technical reports, qualified professional input may be required. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.