The two hour sun on ground test in an overshadowing study

The sun on ground test asks a simple question about an amenity area: on 21 March, does at least half of it receive two hours or more of direct sunlight, and if it does not already, does the development reduce the sunlit area to less than 0.8 times what it was. That is the test in the Building Research Establishment's site layout guidance, and it is the measure that decides whether a garden, a courtyard or a play space is treated as adequately sunlit.

It is a different exercise from the window based measures. Vertical sky component and annual probable sunlight hours describe what reaches a window. Sun on ground describes what reaches the ground, and a scheme can pass one set of tests and fail the other.

What the test measures and why 21 March

The equinox is used because it is representative rather than extreme. On 21 June almost any open space receives sunlight for long periods, and on 21 December almost nothing does, so neither date distinguishes between a good layout and a poor one.

The guidance asks for the proportion of the amenity area receiving at least two hours of direct sunlight on that date, expressed as a percentage of the total area. Half or more is the figure that indicates the space will appear adequately sunlit through the year. Where an existing area already falls below that, the second limb applies: a reduction to less than 0.8 times the previous sunlit area is treated as noticeable.

Report both limbs. A space that was at 38 per cent and drops to 33 has failed the absolute figure both before and after, and the planning question is the proportional change, which is 0.87 and passes.

The Overshadowing Study should state the date, the latitude or location used, whether daylight saving is applied to the timings and whether the two hours must be continuous. The guidance does not require continuity, and assuming it does will fail spaces that are perfectly usable.

Defining the amenity area to be tested

Half the arguments in this field are about the denominator rather than the analysis.

Define the area on a plan, hatched, with its extent in square metres, and say what has been excluded and why. Reasonable exclusions are driveways, bin and cycle stores, service yards, footpaths providing access rather than amenity, and permanently covered areas such as undercrofts. Unreasonable exclusions are the shaded strip along the northern boundary, removed because it fails.

For private gardens serving houses, test each garden separately rather than aggregating them, because an average across a street conceals the two plots that are unusable. For communal space, test the space as designed, including any part intended as a sitting area.

Where a scheme proposes new amenity space, some local standards apply the same half and two hour test to communal amenity and to children's play space. Check the local plan and any play space guidance before choosing a metric, and where a Play Space Assessment is required, keep the sunlight analysis consistent with the play space areas it identifies.

Modelling the surrounding context correctly

Three modelling decisions change the result more than anything in the layout.

The first is context extent. Model surrounding buildings far enough out that any structure capable of casting a shadow across the space at any time between sunrise and sunset on the test date is included. On a tight urban site that can mean a radius considerably larger than the immediate neighbours, particularly to the south east and south west where low morning and afternoon sun throws long shadows.

The second is boundary treatment. Existing and proposed walls, fences and hedges to their real heights, because a 1.8 metre close boarded fence on a southern boundary removes a strip of garden from the two hour figure on its own. Assuming a permeable boundary where a solid one exists overstates the result.

The third is trees. On 21 March deciduous trees are bare, so modelling them as solid volumes is wrong and modelling them as absent is closer to the truth for that date. Evergreens should be included. Where large deciduous trees are the dominant shading influence, report the equinox result without them and add a sensitivity run for a summer date with canopy, and say which is which.

Level changes matter on sloping sites, and the ground model should be the proposed levels rather than existing, because a garden cut into a slope sits lower than the survey suggests.

Reporting before and after percentages

The reporting format that survives scrutiny is a table with one row per amenity area.

Add plots. A pair of plans showing the two hour contour before and after, on the same sheet at the same scale, communicates more to an officer or a member than any table, and the hourly shadow plots behind it belong in an appendix. The conventions for those plots, including which hours to show and how to label them, are covered in shadow plots and dates in an Overshadowing Study.

State the software and the sky and sun model used. Different packages handle the horizon, atmospheric refraction and the sun's disc slightly differently, and where a result sits within a percentage point of a threshold that difference is material enough to mention.

Transient overshadowing and where the argument runs out

Transient overshadowing is the observation that a shadow crossing a space for part of the day is different from permanent shade. It is a fair point and it is already inside the test, because the test measures cumulative duration rather than instantaneous condition.

That is worth stating plainly, because the argument is often made as though the guidance ignored it. A shadow that passes across a garden between 10am and 11am has taken one hour out of that part of the garden's total, and if the remaining sunlit hours still reach two, the point passes the test on its own terms.

Where the argument runs out is on a space that receives sunlight only in a short window. A courtyard that is sunlit from 12pm to 2pm and shaded either side is at exactly the threshold, and a new building that removes twenty minutes takes it below. Describing that as transient is not persuasive, because it removes the only sunlight the space has.

The honest position in a marginal case is to report the figure, describe the pattern through the day, and let the decision maker weigh it rather than characterising the loss with an adjective.

Gardens, courtyards and play space in practice

The three types behave differently and it is worth saying so in the report.

Private rear gardens on a terrace are long and narrow, so shading from a rear extension next door affects the near end while the far end is unaffected. The percentage often passes while the part of the garden closest to the house, which is where people sit, does not. Reporting the sunlit area's location as well as its size addresses this.

Communal courtyards in perimeter block schemes are the hardest case, because the block shades its own courtyard. The variables that move the result are courtyard width, the height of the southern range and whether that range is broken by a gap or lowered.

Children's play space frequently ends up in the residual land at the northern edge of a site, which is the worst position for sunlight. Testing the play space location early, before the layout is fixed, avoids a late relocation.

Design changes that move the result

Five changes do most of the work, in rough order of effectiveness.

1. Reduce the height of the range on the southern side of the space, even by one storey.
2. Introduce a gap in that range, aligned so that midday sun reaches the space through it.
3. Widen the space in the north to south direction rather than east to west.
4. Move the amenity area to the southern or western part of the site and put parking or servicing in the shaded northern strip.
5. Lower or make permeable the southern boundary treatment, which on small gardens can be worth several percentage points on its own.

Test these early. Running the analysis once the layout is fixed turns it into a compliance report, and the value of a sun on ground study is in the two or three iterations before that point. The window based analysis in a Daylight and Sunlight Assessment should be run in the same iterations, because a change that improves ground sunlight sometimes worsens a neighbour's window.

This is general guidance, not legal advice. The Building Research Establishment guidance is advisory and local standards vary, so confirm the metric and the target with the local planning authority or a qualified consultant before relying on it.

PlanScope can draft the assessment narrative and policy context around your modelling outputs, and you can browse all 66 report types to see what else a scheme needs. 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.