Setting a mirror image baseline in a daylight assessment

A mirror image baseline daylight assessment compares the daylight reaching a neighbouring window against what that window would receive if the site opposite were built to a form that mirrors the neighbour's own building, rather than against the site as it stands today. It exists because the standard comparison, measuring against a cleared or under-developed site, sets a target that no urban scheme can meet. The 2022 edition of the Building Research Establishment guidance on site layout planning for daylight and sunlight makes room for alternative baselines of this kind, provided the choice is explained and justified.

Used carefully it is a legitimate piece of analysis. Used to bury a poor result it is the first thing a reviewer will pull apart.

Why a baseline is needed

The numerical tests in the guidance are relative, not absolute. Vertical sky component is judged against the existing value at the same window, with a material reduction commonly taken at a retained value below 27 per cent and below 0.8 times the former value. Daylight distribution and annual probable sunlight hours work the same way.

That relativity is the whole problem in dense areas. A window facing a surface car park currently receives far more sky than the surrounding urban grain would ever give it. Judged against that accident of history, any development of the car park fails.

So the question becomes what the comparison should be. The guidance is clear that the numerical targets are advice rather than a fixed rule and that they should be applied flexibly in dense urban contexts. The mirror image approach is one way of giving that flexibility a defensible shape.

When the mirror image approach is used

The conditions where it holds up are narrow, and stating them plainly in the report is what earns it credibility.

Use it where the site is genuinely under-developed relative to its context, where a consistent urban form exists on the surrounding streets, and where the neighbouring building whose windows are being tested is itself part of that form. A terrace of four-storey townhouses facing a cleared plot is the clean case. The mirrored form is simply the same terrace, reflected across the street.

It does not hold up where the surrounding grain is inconsistent, where the neighbouring building is a low bungalow in a suburban street, or where the proposed scheme is substantially taller than the mirrored form. In that last case the mirror image tells you about a building you are not proposing, which is the criticism that sinks the analysis at inquiry.

Setting out the illustrative model

The following figures are illustrative and are given to show the method, not as results for any real site.

Take a site on the north side of a street 14 metres wide, facing a terrace of four-storey buildings with a parapet 13 metres above pavement level and a shallow pitched roof behind. The site currently holds a single-storey workshop 4 metres high.

The three models you need:

1. Existing. The workshop as built, with ground levels surveyed rather than assumed.
2. Mirror image baseline. A block on the site footprint with a parapet at 13 metres and the same roof pitch, set at the same distance back from the pavement as the terrace opposite.
3. Proposed. The scheme as designed, at whatever height and massing it reaches.

Run the assessment twice: proposed against existing, and proposed against mirror image. Report both. An assessment that only reports the flattering comparison invites the assumption that the other one is bad.

That third row is the one people forget, and it is often the most persuasive. If the mirror image alone already takes a window below the target, the loss is a function of the street, not of your design.

Existing obstructions and their treatment

Model what is there. Boundary walls above 1.5 metres, mature trees in leaf, projecting bays, balconies, canopies and existing plant all reduce the existing value, and leaving them out inflates the baseline and makes your proposal look worse than it is.

Trees are contentious. The guidance treats deciduous trees as partially transmitting rather than solid, and practice varies on whether to model them at all. State what you did. If you have excluded trees, say so and explain that the result is conservative.

Rooflights and windows to non-habitable rooms need a decision too. Bathrooms, circulation space, stairwells and store rooms carry less weight, and the guidance says so. List which windows you have assessed and which you have set aside, with the room use for each, drawn from a survey or from the planning history rather than guessed.

Reporting results against the baseline

Present the numbers window by window, not as an aggregate percentage. Reviewers want a table with the window reference, the room it serves, the room use, the existing value, the mirror image value, the proposed value and both ratios.

Then write a paragraph of interpretation for the windows that fail. Say which floor they are on, what room they serve, whether the room has other windows, and what the retained absolute value is. A ratio of 0.72 on a window serving a room with two other openings is a different matter from 0.72 on a single-aspect bedroom.

Keep the Daylight and Sunlight Assessment and any separate Overshadowing Study consistent on geometry. The two documents are frequently produced from different models, and a reviewer who spots a discrepancy in the massing will discount both.

Where the approach is criticised

Three criticisms recur, and you should meet them in the report rather than wait for the objector's consultant to raise them.

The first is that the mirror image is a hypothetical building with no planning status. That is true. The answer is that it is a reference form used to interpret a relative test, not a consented scheme, and the existing-condition results are reported alongside.

The second is selection bias in choosing which building to mirror. Answer it by stating the rule you applied before running the analysis, for example that the mirrored form matches the closest building directly opposite, and by showing the street elevation that supports it.

The third is that a resident experiences the actual loss, not the notional one. That criticism is correct and cannot be argued away. The honest position is that the mirror image informs the planning judgement about whether the design is reasonable, while the existing-condition figures describe the lived change. Say both.

Alternative baselines and sensitivity

The mirror image is not the only option. Others in use include a baseline drawn from the height limit in an adopted site allocation, a baseline based on the prevailing datum of the street, and, where a scheme has recently been consented nearby, that consented envelope.

Whichever you pick, run a sensitivity test. Model the mirrored form one storey lower and one storey higher and report how the results move. If the conclusion survives a storey either way, it is a conclusion worth defending. If it flips, the baseline is doing too much of the work and the design needs to change instead of the analysis.

Explaining the choice to an officer

Officers are not daylight specialists and the report needs to work for a planner reading it once. Lead with a plain sentence saying which baseline you used and why. Put the reasoning on a single page near the front, with the street elevation showing the mirrored form drawn over the existing context.

Offer the officer the sentence they need for the committee report: the scheme performs within the range a contextual form on this site would produce, and the windows falling below the standard targets are identified with the reasons. Then let the appendices carry the modelling. A Design and Access Statement that describes the massing in the same terms as the daylight report makes the whole submission easier to follow, and you can see the full set of 66 report types by category if you are still scoping the evidence base.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. Daylight and sunlight modelling is specialist work and councils usually expect a suitably qualified author, so professional input on the modelling itself may be required.