Reviewing an opponent's daylight report

Most daylight reports you are asked to review are competently produced and still reach a conclusion you would not. The difference almost never lies in the arithmetic. It lies in four decisions made before any calculation happened: which rooms were included, what baseline was modelled, how far the context model extended, and which test was applied to which window. Check those four first, and the rest of the review takes an afternoon rather than a week.

Review in a fixed order, and record what you could not check as carefully as what you could.

Start with the scope and the excluded rooms

Open the report at the schedule of assessed windows and compare it against the floor plans of the affected properties. You are looking for rooms that are absent.

Non-habitable rooms are properly excluded, and bathrooms, hallways, landings and storage are the usual list. The questions worth asking are about the borderline cases. A kitchen treated as non-habitable, where it is the only daylit space serving an open plan room. A bedroom described as a study. A room in a basement flat omitted because its window is below the assessment plane.

Where the internal layout of a neighbouring property was assumed rather than surveyed, the assumption decides the result. Find the statement of where the layout came from. If the report does not say, that is a finding in itself, because an assumed layout that happens to place the window in a shallow room produces a more favourable no sky line result than a deep room would.

Then check the properties, not just the rooms. A report that assesses the two adjoining houses and omits the flat above the shop on the corner has made a scoping decision that needs explaining.

Check the baseline against the drawings

The existing situation is the baseline in the ordinary case, and three variations on it cause arguments.

A consented but unimplemented scheme on the application site, modelled as the baseline. That can be legitimate, and it has to be stated and justified rather than assumed, because comparing against a larger consented envelope flatters the proposal.

A recently demolished building modelled as still present, or recently constructed one omitted.

Permitted development potential on the neighbouring site treated as the baseline. That is not a baseline, and it is occasionally slipped in.

Verify the existing envelope against the topographical survey and the existing elevations rather than against the proposed drawings. Then verify the proposed envelope against the drawings actually submitted for determination, checking revision numbers and dates. A model built from an earlier revision is a common and entirely innocent error, and it can change the results.

Where the review relates to a small scheme, the same check applies in miniature: the obstruction angle in daylight evidence for a rear dormer in a terrace depends entirely on which section drawing it came from.

Test the context model extent

The context model is the input most often wrong and least often checked. A report that models only the application site and the immediately affected facades will overstate the loss caused by the proposal, because other obstructions that already limit light to those windows are missing.

Ask for the plan showing the modelled extent. If the report does not contain one, request it, because the extent cannot be verified from the results.

Three specific checks repay the time. Whether buildings behind the receptor, which do not obstruct, have been included unnecessarily, which is harmless. Whether buildings to the side within the field of view have been omitted, which is not. And whether boundary walls, projecting bays, porches, dormers and existing extensions on the receptor property itself are modelled, since these often account for more of the existing restriction than the proposal does.

Balconies and overhangs above a window reduce the vertical sky component substantially, and the recognised guidance offers a way of reporting the effect with and without them. A report that models balconies on the proposed scheme but not on the existing receptor is comparing two different things.

Trees are usually excluded, which is consistent with the guidance for deciduous species. Dense evergreen planting immediately outside a window is a different matter and should at least be discussed.

Look at reflectance and transmittance inputs

Now the numbers. The inputs that move results are glazing transmittance, internal surface reflectances, any maintenance or dirt factor, and the sky model used.

Check that the stated values are the ones actually used, which means asking for a sample calculation if the report gives only outputs. Check that the same values were applied to existing and proposed cases, because applying a generous transmittance to the proposed condition and a standard one to the existing is a quiet way of improving a ratio.

For internal daylight assessed under the current illuminance based approach, the room surface reflectances matter more than they do for vertical sky component, and a report using optimistic reflectances for a room with dark finishes is overstating provision.

Where the report assesses annual probable sunlight hours, confirm that only windows facing within 90 degrees of due south were assessed. Including north facing windows is a harmless error. Excluding a south east facing window is not.

Verify which test was applied to which window

This is where real method errors cluster.

Two errors recur. Reporting only vertical sky component for a room whose depth means the no sky line result is the one that matters. And applying the retained ratio to a window whose existing value is already very low, where a small absolute loss produces a large proportional one, without saying so. The second is not an error of method, but it is an error of presentation, and the honest way to report it is to give both the absolute and the proportional change.

Also check the overshadowing assessment separately. An Overshadowing Study uses a different metric and a different date from the window assessment, and conflating them is common.

Separate method errors from judgement

A review that lists twelve criticisms of equal weight is less useful than one that lists three method errors and then disagrees with one judgement.

Method errors are things that are wrong and can be corrected: a missing building, a superseded drawing, an omitted room, the wrong test, an input inconsistently applied. Quantify the effect where you can, and say where you cannot.

Judgements are things a competent assessor could reasonably decide differently: whether a shortfall is acceptable in a dense urban context, what weight to give an existing low baseline, whether a room's assumed use is the right one. The guidance is explicit that its numerical targets are advisory and that flexibility is appropriate in dense areas, so a report that treats the targets as absolute pass or fail thresholds has made a judgement error rather than a calculation error.

Say which is which. A decision maker who can see the difference will act on the method errors and form their own view on the judgements, which is the most a reviewer can usefully achieve.

Writing the review so it is usable

Keep it to a structure a case officer can lift into a report.

1. What you were asked to review, what you received, and what you asked for and did not receive.
2. The method errors, each with the effect on the results stated or described.
3. The judgement points, clearly labelled as such.
4. What you agree with, which should not be an empty section.
5. What would resolve the outstanding points, with the specific output you would need.

Two habits improve how a review lands. Avoid adjectives. A report described as seriously flawed invites a defensive response, while a report described as having omitted the building to the south east invites a corrected run. And state your own limitations: whether you had access to the properties, whether you built an independent model or checked the submitted one, and which conclusions you could not test.

Where the review is commissioned by an objector rather than by the authority, the same discipline makes it more persuasive, not less. Officers discount advocacy and read analysis. The point applies equally when answering a neighbour's daylight objection from the applicant's side.

If you review this kind of work and would consider reviewing AI-assisted drafts, the professional directory is the route to sign-off requests, and a Daylight and Sunlight Assessment drafted by the platform is a starting point for a qualified reviewer rather than a finished document.

This is general guidance, not legal advice. Daylight and sunlight assessment is a specialist discipline and councils usually expect a suitably qualified author, so professional review is strongly advised.

PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.