Using Environment Agency LiDAR for site levels
Free national LiDAR gives ground levels for almost anywhere in England. Here is how to download it, which model to use, where it misleads and how to cite it in a planning submission.
Environment Agency LiDAR gives you ground levels for almost anywhere in England, free, under the Open Government Licence, usually at one metre resolution or better. For a self-builder working out whether a plot falls away to the rear, how a flood level compares with the ground, or what the neighbours' ridge heights are, it answers the question in an afternoon. It does not replace a topographical survey, and submitting it as though it does will cost you later.
The distinction that matters most is between the terrain model and the surface model. Get those the wrong way round and every level you quote is wrong.
What the data gives you
Aircraft-mounted laser scanning has covered England repeatedly, most recently through a national programme producing consistent coverage at one metre resolution. Older composite datasets exist at finer resolutions in places, down to a quarter of a metre in areas surveyed for flood modelling.
Heights are to Ordnance Datum Newlyn and positions are on the British National Grid, which means the data drops straight into a site drawing without reprojection. The published accuracy varies by dataset and by capture, so read the metadata for the tiles you download rather than assuming a figure.
Coverage is not universal and it is not uniform in age. Some areas have been flown several times, some once, and a few gaps remain. A tile captured in 2017 will not show the earthworks somebody carried out in 2021.
Terrain against surface models
A digital surface model records the first return from the laser, which means the top of whatever the beam hit. Trees, hedges, roofs, parked vehicles, polytunnels.
A digital terrain model is derived from that by an algorithm that attempts to strip out vegetation and buildings and leave bare earth.
Both are useful for different questions. Use the surface model when you want to know how tall the neighbouring building is, what the hedge line does, or where a tree canopy sits. Use the terrain model for ground levels, falls across a plot and flow paths.
The failure mode to watch for is the terrain model under dense vegetation. The algorithm has to guess where the ground is when very few pulses reach it, so levels under woodland, thick scrub or a mature hedge line can be well out. Steep banks and narrow ditches are smoothed. Retaining walls sometimes disappear.
Downloading tiles for a site
The data is published through Defra's survey data download service, which lets you draw an area, choose the product and resolution, and download the tiles as raster files.
- Find the site on the map and draw a boundary that extends beyond the red line, because you will want the surrounding ground.
- Choose both the terrain and surface products for the same year where both exist.
- Note the capture date, the resolution and the product name before you close the page.
- Load the tiles into your drawing software or a free geographic information system to generate contours or a levels grid.
- Save the metadata alongside the data, because you will need it when you write the report.
Most of this works in free desktop mapping software, and the raster files open without a licence. Step three is the one people skip and regret. A levels drawing with no stated source and date is not evidence.
Accuracy limits and when a survey is still needed
LiDAR is a remotely sensed national dataset. A topographical survey is a measured record of your site on a stated date, tied to a benchmark, with features picked up individually.
You need the survey for anything that will be built to. Finished floor levels, retaining structures, drainage falls, boundary positions, threshold levels, levels for a section drawing that will be conditioned. You also need it wherever a small error has a large consequence, which includes a tight access gradient or a flood-sensitive threshold.
LiDAR does the work before that. Feasibility, cut and fill estimates, catchment and flow path analysis, initial sections, and checking a claim made by somebody else.
There is a middle case worth knowing. On a large or awkward site, a full survey of the whole area is expensive, and a sensible compromise is to survey the developable part properly and use LiDAR for the wider context: the fall of the land beyond the boundary, the catchment upslope, the levels of the lane. Say in the drawing notes which part came from which source.
Using it for an initial flood check
This is where the dataset earns its keep for a householder or a small developer.
Take the ground levels from the terrain model, compare them with the flood levels available from the flood map for planning and any detailed data the Environment Agency will supply on request, and you have a first view of whether a site sits above or below a modelled level. That is the core geometry of a Flood Risk Assessment, even though the assessment itself needs far more.
The terrain model is also good at showing overland flow routes across a site, which is the question a Drainage Strategy has to answer. Where water leaves a field and where it collects are usually visible in a hillshade render long before anyone digs a trial pit.
Mind the timing point. A ground level that has been altered since capture invalidates the comparison, and made ground on a former industrial site frequently has been.
Where the model misleads
Four situations catch people out.
Vegetation, as above. Buildings, where the terrain model may retain part of a footprint or leave a depression where one was removed. Water, which absorbs the laser and returns nothing, so water surfaces are interpolated rather than measured. And temporary features, where a stack of pallets or a spoil heap present on the day of capture appears as topography.
None of this is a criticism of the dataset. It is a national product generated by algorithm, and it does that job well. The mistake is treating a derived raster as a measured survey.
Presenting LiDAR-derived levels honestly
Say what it is. One short paragraph in the report is enough: the dataset name, the product, the resolution, the capture year, the datum, and a sentence saying the levels are derived from remotely sensed national data rather than a measured survey.
Then say what you did with it. Contour interval, whether you used the terrain or surface model, how you handled vegetated areas, and whether any spot levels were checked on site with a staff and a level.
Reviewers accept LiDAR readily for the right purposes when it is described this way. They reject it when it appears as unattributed contours on a proposed site plan. The same rule applies to any public dataset used as evidence, including the utility records covered in our note on ordering sewer records and pre-development enquiries.
This is general guidance, not legal advice. Check the position with your local planning authority or a qualified professional before relying on it.
Enter your postcode on PlanScope to see which reports a sloping or low-lying site is likely to need.
PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval.
Frequently asked questions
What is the difference between a DTM and a DSM?
A digital surface model records the first laser return, so it includes trees, hedges, roofs and anything else the beam hit. A digital terrain model is derived from it by an algorithm that strips out vegetation and buildings to estimate bare earth. Use the surface model for building and canopy heights, and the terrain model for ground levels and flow paths.
Is LiDAR accurate enough to design from?
Not for anything built to a level. Finished floor levels, drainage falls, retaining structures, threshold levels and access gradients all need a measured topographical survey tied to a benchmark. LiDAR is well suited to feasibility work, cut and fill estimates, catchment analysis and checking somebody else's claim, and the published accuracy varies by dataset so read the metadata.
Can I use LiDAR in a flood risk assessment?
As part of one, yes. Ground levels from the terrain model can be compared with modelled flood levels to establish the basic geometry, and overland flow routes often show clearly in a hillshade render. The assessment itself needs far more, including detailed flood data from the Environment Agency and, for anything being built, surveyed levels.
Why are levels wrong under trees?
Because very few laser pulses reach the ground through a dense canopy, so the algorithm generating the terrain model has to estimate where the surface lies. Levels under woodland, thick scrub and mature hedge lines can be well out, and narrow ditches and steep banks are smoothed. Treat vegetated areas as indicative and check them on the ground.
How should I describe LiDAR data in a planning submission?
Name the dataset and product, give the resolution and capture year, state the vertical datum, and say plainly that the levels are derived from remotely sensed national data rather than a measured survey. Then describe what you did with it, including the contour interval and whether any spot levels were checked on site.
Sources and further reading
This article is general guidance about the planning system in England and, where stated, the rest of the UK. It is not legal or professional advice. Check your local planning authority’s validation requirements and take professional advice for your specific site.