Where to Actually Get Free UK Site Analysis Data

Nearly everything you need for a UK site analysis is already published, already free, and sitting behind a government logo. The paid environmental searches you keep seeing at the top of Google — Landmark, Groundsure and the rest — are conveyancing products. They bundle public datasets, add indemnity, and sell the package to solicitors. Useful at acquisition. Irrelevant to a Part 2 student with a crit on Thursday, and mostly irrelevant to a small practice doing a desk study.

So here's the sourcing list, layer by layer. For each one: where it comes from, what it will not tell you, and the single design decision it ought to change. If a layer doesn't change a decision, it isn't analysis. It's wallpaper.

Planning History First, Sun Path Last

Everyone starts with the sun because it's the fun one to draw. Start with the planning record instead, because it's the layer most likely to kill a scheme.

Your local authority's planning portal — every English and Welsh LPA runs one, Scotland via the ePlanning service — lets you search by address or by map. Pull the applications on your site, and then pull the applications on every site that touches it. You want the officer's delegated report, not the decision notice. The decision notice says no. The report says why, which is the thing you can design around.

Alongside it: Historic England's National Heritage List for England for listed buildings and scheduled monuments, your council's own GIS layer for conservation area boundaries, Article 4 directions and tree preservation orders. MAGIC Map, run by Natural England, carries designations at a national scale — SSSIs, ancient woodland, agricultural land classification, AONBs — and is the single fastest way to find out your lovely sloping field is Grade 2 agricultural land.

What it won't tell you: pre-application advice, which is confidential and often where the real position lives. It also won't tell you much about the deep past. In practice, most authorities digitised their records somewhere around the late nineties, and anything older is a microfiche or a paper file you have to ask a records officer for — politely, by email, with the address and any reference you already have.

The decision it should change: your massing envelope. Two refusals on the neighbouring plot for a three-storey rear addition is not trivia, it's your ridge height.

Flood Is Three Maps Pretending to Be One

The Environment Agency publishes the Flood Map for Planning — that's the one with Flood Zones 1, 2 and 3, and it covers rivers and the sea only. It is the map a planning officer will open. It is also published with the caveat that it is not a property-level assessment, which people ignore constantly.

Surface water is a separate dataset, on the EA's long-term flood risk service, along with reservoir inundation and a coarse groundwater indication. Then the third map: your LPA's Strategic Flood Risk Assessment, a PDF buried in the local plan evidence base. The SFRA is the one with climate change allowances applied and local knowledge baked in, and when the national map and the SFRA disagree, the SFRA is what the case officer cites. Historic flood outlines — where water has actually been, as opposed to where modelling says it could go — are published by the EA as a separate layer and are worth more than either.

What none of them tell you: a finished floor level. They give you a zone, not a number.

The decision it should change: your FFL and what you put on the ground floor. Habitable rooms, electrical intake and plant all move upstairs long before you draw a plan.

The BGS Gives Away What Consultants Charge For

The British Geological Survey's Geology Viewer gives you bedrock and superficial deposits at your site, free, in a browser. Superficial is the layer that matters for most small projects — made ground, alluvium, river terrace gravels, and the glorious news of London Clay with its shrink-swell behaviour.

Then the BGS onshore borehole index. Scanned borehole logs, mostly from past site investigations, searchable on a map, free to view. If someone drilled a hole within a few hundred metres of your site in 1974 and the log got lodged, you can read what they hit and at what depth. It isn't your site. It is enormously better than nothing.

The Coal Authority — now the Mining Remediation Authority — publishes an interactive viewer showing development high risk areas and reported mine entries. UKradon, run with UKHSA, gives an indicative radon map free; the site-specific report is a paid product. Historic Ordnance Survey maps do your contamination screening for you: gasworks, tanneries, filled ponds, railway sidings, anything with the word works on it.

What it won't tell you: bearing capacity, groundwater level, or whether there is actually contamination. Geology maps are interpretation, not measurement.

The decision it should change: your foundation assumption, and whether you tell the client now that a site investigation is a line in the budget rather than a surprise in month four.

Sun: Get Past the Solstice Arcs

SunCalc gives you azimuth and altitude for any date and time at any coordinate, overlaid on a map, in about four seconds. Andrew Marsh's 3D sun path web apps are the better teaching tool — you can see the sun path dome around the site and read shadow angles off it directly, free in a browser.

For anything quantitative, you want an EPW weather file. EnergyPlus publishes them by location and Climate.OneBuilding hosts a wider set, free, and these are what Ladybug, Climate Studio and most environmental plugins expect as an input. CIBSE's UK weather files are the recognised set for formal design work and they are a paid product — worth knowing before you promise a client a CIBSE-compliant overheating check.

What SunCalc won't tell you: obstruction. It models an empty planet. The sun path is only useful once you've put the neighbouring terrace, the sycamore and the six-storey block opposite into a massing model — which you can build quickly from OS OpenData footprints and a site visit with a tape and a camera.

The decision it should change: where the main living space sits, and what time of day it stops being usable. BRE's BR 209, Site Layout Planning for Daylight and Sunlight, is the document a planning officer will measure you against on overshadowing, and it is not free.

Wind Roses Are Not Pedestrian Comfort Studies

A wind rose tells you prevailing direction and speed distribution from a weather station. meteoblue publishes free wind roses by location; the Met Office publishes observation data from its station network; the EPW file you already downloaded contains hourly wind speed and direction for the year.

Here's the bit nobody says: that station is probably at an airport, on flat open ground, ten or twenty kilometres from your site, at a height and exposure that has nothing to do with a courtyard between two four-storey blocks. The regional wind rose tells you where weather comes from. It tells you nothing about what happens when it hits your corner. Downdraught off a tall slab, the funnelling in a narrow gap between buildings, the gust at the entrance — that's microclimate, and a wind rose doesn't contain it. What we've found is that the honest move at early stage is to draw the prevailing direction, then draw the obstructions and gaps around your site and reason about them with your own eyes. If the building is tall enough or the public realm important enough, that becomes a CFD or wind tunnel study later, with a consultant.

The decision it should change: where the front door goes, and which side of the building you ask people to stand still on — bus stop, café terrace, cycle store, smoking spot.

The Site's Past Is a Free Dataset

The National Library of Scotland's georeferenced historic OS maps cover all of Britain, side-by-side with modern mapping, with a transparency slider. It is the most useful free tool in this entire article and half the profession doesn't know it exists. Britain from Above holds the Aerofilms oblique aerial collection; Historic England's aerial photo explorer adds vertical coverage including wartime RAF sorties. Google Earth's historical imagery slider gets you the last couple of decades.

Four maps across a century tells you what was demolished, where the lost watercourse ran, why that boundary kinks, and which part of the site is made ground.

The decision it should change: usually two — your contamination risk flag, and your argument about grain. "The historic plot width was 5.2 metres and I've reinstated it" is a defensible planning argument. "It felt right" is not.

Movement, From Actual Data

The Department for Transport's Bus Open Data Service publishes timetables, fares and location data for English bus services; NaPTAN gives you every stop with coordinates; TfL's open API covers London in detail. DfT road traffic counts give you annual average daily flows on classified roads. The Propensity to Cycle Tool maps cycling demand by route.

This is what turns the limp isochrone diagram into something a transport officer takes seriously. Not "bus stop nearby" but three services, these frequencies, this walk time.

The decision it should change: parking provision, cycle store size and position, and which boundary the entrance faces.

What You Should Actually Pay For

For a student project: nothing. Everything above is free and a tutor who asks for more is testing whether you'll go and get it.

For a real application, the free desk study still stands — what you pay for is the stuff no public dataset contains. A topographical survey. A measured building survey. A site investigation with real boreholes and lab results. An arboricultural survey to BS 5837. Those exist because the free data stops exactly where liability starts. Pay for measurement; don't pay anyone to re-badge a BGS extract.

Record Provenance Like Somebody Is Going to Check

They will. A tutor will ask where the flood zone came from. A planning officer will ask whether you looked at the SFRA or just the national map. And twelve months later you will open your own file and have no idea which version of a dataset you traced.

Keep a data register alongside the drawings. One row per layer:

Screenshot every viewer with the date visible and file it with the register. Most of these datasets are published under the Open Government Licence, which asks you to attribute the source — so attribute it on the drawing, in small type, next to the layer it belongs to. Three words and a date under a diagram does more for your credibility in a crit than another rendered perspective.

And when the national dataset and the local record disagree — which happens most often with flood and with conservation area boundaries — in practice you lead with the local one and note the conflict. The LPA wrote the local document. Arguing that a national map contradicts their own evidence base is a fight you win in the appeal and lose in the room.

The Test

Go back through your layers and ask what each one changed. Floor level. Entrance position. Ridge height. Foundation budget. Where the living room is. If a layer didn't move a line, it was decoration, and the panel can smell it from the back of the room.

The data is free. The thinking is the bit you're being marked on.