Case study
Maiden Castle in LiDAR
Maiden Castle, south-west of Dorchester in Dorset, is an Iron Age hillfort on a chalk hilltop, built over Neolithic monuments and protected as a scheduled monument. Here it is in eight views of one metre LiDAR.
Last updated By Gitwix Limited
All eight show one frame of the Environment Agency's LiDAR. This page says what each shows of the ramparts, the entrances and the ground around them, and how a student might read them.
What is Maiden Castle?
Maiden Castle stands on a chalk hilltop of the South Dorset Downs, overlooking the valleys of the rivers Frome and South Winterborne, with Dorchester, the Roman town of Durnovaria, to the north-east. It is known above all for its Iron Age hillfort, but Historic England's list entry describes remains of many periods: Early Neolithic pits of around 4000 BC, a causewayed enclosure of about 8 hectares on the eastern plateau, a bank barrow 546 m long built around 3500 BC, two bowl barrows near the centre of the hilltop, the hillfort itself, and a Romano-Celtic temple of the 4th century AD.
The hillfort grew in stages. The first had a single rampart enclosing 6.5 hectares, with entrances to the east and west. In the second half of the 3rd century BC the ramparts were extended around the hilltop to enclose 20 hectares, and in the mid-2nd century BC they were rebuilt on a larger scale. After several remodellings they ended as three banks and two ditches around the hilltop, with an extra bank and ditch along the gentler slope to the south.
Edward Cunnington excavated part of the site in 1884, and Mortimer and Tessa Wheeler excavated there from 1934 to 1937. Further excavation and survey followed in 1985 to 1986, published by Niall Sharples in 1991.
During the Iron Age, the hilltop was progressively developed into one of the largest hillforts in Britain. The hillfort is particularly noted for the scale and extent of the outer ramparts and well developed entrance earthworks.
Those ramparts and entrances are what LiDAR records best, because they are still large earthworks. Much of the rest was found by digging, and lies buried.
| Record | Maiden Castle |
|---|---|
| Designation | Scheduled monument |
| List entry | 1015775 |
| First listed | 9 October 1981, last amended 24 September 1997 |
| National Grid reference | SY 66891 88459 |
| Parishes | Winterborne Monkton and Winterborne St. Martin, Dorset |
| Scheduled area | About 47 hectares on Historic England's map |
What do the pictures show?
All eight pictures show one frame, about 2.6 km from west to east and 1.3 km from north to south, with north at the top; at full size each pixel covers about a metre of ground. The heights are the Environment Agency's LIDAR Composite DTM 1 m, produced in 2022. In this frame the Agency built it from a single survey, flown on 27 February 2021 at 1 m resolution, according to its survey index. A terrain model is the bare ground with buildings and vegetation taken out, so the pictures show the shape of the earth itself.
The LiDAR relief picture is the Agency's own hillshade. The other seven are drawn by Field Archaeology from the same terrain model. How each one is drawn, with its settings and the paper it follows, is set out in the LiDAR visualisations guide.
These scheduled monuments lie inside the frame. The names are Historic England's, and the last column says how each one shows in the Local relief view, which picks out low mounds best.
| Name | List entry | Where in the frame | In Local relief |
|---|---|---|---|
| Maiden Castle | 1015775 | The centre | Bold bands of white and black |
| Four bowl barrows on Hog Hill, west and north west of Maiden Castle | 1015777 | On Hog Hill, west of the fort | Faint pale round spots |
| Four bowl barrows immediately north west of Maiden Castle | 1015776 | Against the ramparts at the north-west corner | Faint pale spots |
| Long barrow immediately north west of Maiden Castle | 1015779 | Just north of the north-west corner | A faint pale smudge |
| Round barrow cemetery immediately south east of Maiden Castle | 1015778 | Just south-east of the eastern entrance | Two clear pale spots |
| Bowl barrow 360m south west of Clandon Farm | 1015782 | The left edge, near the top | A clear pale spot |
| Round barrow N of Ashton Cottages | 1004546 | The bottom left corner | A small pale dot |
What does each view show?
The views run from the most familiar to the most specialised: the Agency's hillshade, which looks like the ground in low sunlight, then six views that do not depend on a single direction of light, then a picture of the shape of the land. Each heading names the view as Field Archaeology lists it.
LiDAR relief, the Agency's hillshade

A hillshade lights the ground from one direction, like a low sun. Here the light falls from the north-west, at the top left, so slopes facing that way are pale and slopes facing away are dark. The ramparts stand out at once as bands of light and shadow, and the fields, the farm at the top left, the roads, and the A354 and the railway at the lower right are easy to place.
The light is also its weakness. A bank is pale on its north-western side and a ditch on its south-eastern side, so the two can be confused, and low features barely show: the bowl barrows on Hog Hill, west of the fort, are hard to find in this view.
Multi-light relief

The same ground lit from six directions and averaged, with heights exaggerated two times. Every steep face comes out dark whichever way it faces, so the ramparts have the same weight on every side of the fort and none is lost in the shadow of a single sun.
Inside the fort it is the cleanest of the eight: a curving bank crosses the interior from north to south, two hollows lie west of it, one square and one round, and a small square mark sits in the eastern half. What the records make of them is set out under How would a student read it?
Local relief

Each point is compared with the average height of a square of ground 41 m across, so hills and valleys drop away and only small rises and dips remain: pale is higher than its surroundings and dark is lower. It is a simple local relief model; compare Hesse (2010).
The ramparts become bands of white and black, which makes them easy to count, but they are far taller than the 0.6 m either way that the grey scale spans, so the bands show where the banks and ditches are, not how big they are. This is the view for the barrows: those on Hog Hill, the group south-east of the fort and the one at the left edge show as pale round spots that the hillshade barely shows.
Sky view

The share of the sky each point can see, shaded from 0.86 (dark) to 0.99 (bright): open ground is pale, and anything enclosed, such as a ditch or a hollow, is dark. The measure is the sky-view factor of Zakšek, Oštir and Kokalj (2011).
With no light direction to allow for, the ditches between the ramparts read as continuous dark rings and the two hollows inside the fort as dark spots, and the routes through the entrances can be followed between the dark ditches. Mounds show less strongly than hollows in this view.
Palaeochannels (slope)

Each metre is shaded by how steep it is, from flat (pale) to 14 degrees and steeper (dark); the atlas lists this view as Palaeochannels. The faces of the ramparts are far steeper than 14 degrees, so each bank shows as two dark bands with a thin pale line along its crest, and each ditch as a pale line along its floor.
Slope cannot tell a rise from a dip, so read it beside Local relief or Red relief. Around the fort it traces the steep sides of the hill and its valleys, and the scarps along field boundaries.
Red relief

Brightness comes from openness and redness from slope, after the red relief image map of Chiba, Kaneta and Suzuki (2008): ridges come out light, hollows dark and steep ground red, with no light direction to choose. The ramparts become red bands with pale crests and dark ditch floors, and the overlapping banks of the entrances are easy to follow.
Because it relies on colour, it reproduces poorly in black and white; for a printed figure, Local relief or Sky view may serve better.
Openness composite

Positive openness in red and negative openness in blue, with the mean of the two, toned by the sky-view factor, in green; openness is the measure Yokoyama, Shirasawa and Pike introduced in 2002. Ground that stands proud reads red to orange and sunken ground reads blue, so the banks and ditches of the ramparts are told apart by colour alone.
It is the busiest of the eight: every field bank, every track and the railway are drawn as strongly as the fort, which helps when following a small feature and hinders when reading the whole.
Height colour

The ground is compared with the average height of a square 601 m across, with 15 m either way across the colours, so this is a picture of the land rather than of its earthworks. The hilltop the fort encloses is pale, and so is Hog Hill to the west; a narrow valley runs between them, and the land falls away to valleys on the north and the south.
It shows how the ramparts follow the ground. Where a small valley heads into the south side, below the middle of the fort, the ramparts bend in around it, and the list entry notes the extra bank and ditch ‘along the gentler slope to the south’.
How would a student read it?
One way to work through the frame, for a seminar, a dissertation or a first look at LiDAR:
- Find your bearings. North is at the top and the frame is about 2.6 km across; the fort itself is a little over a kilometre from end to end. Pick out first what a modern map names: the farm at the top left, the roads, and the A354 and the railway at the lower right.
- Know where the light comes from, and use more than one view. In the Agency's hillshade a bank is pale on its north-western side and a ditch on its south-eastern side, so check each reading in Local relief, where higher ground is pale whatever the light. Historic England's guidance warns that without a record of the light's direction, ‘shadows and highlights can be reversed’, and it cites a study in which no single visualisation recorded more than 77 per cent of features, while any three together recorded more than 90 per cent.
- Count the defences. Follow a line from the interior outwards on the north side, then on the south. The list entry records three banks and two ditches around the hilltop, with an extra bank and ditch along the gentler slope to the south. Measured along one north to south line, easting 366750, the Agency's 1 m terrain model shows just that: three banks and two ditches on the north, spread over about 80 m, and four banks and three ditches on the south, over about 110 m. On the north side the crest of the inner rampart stands about 14 m above the floor of the ditch in front of it.
- Trace the entrances. The first hillfort had entrances to the east and west, and English Heritage notes that they ‘became increasingly complex as ramparts were added and gateways redesigned’. At the west end the ramparts overlap and stagger, so the way in winds between them; at the east end an outer bank curves out in front of the entrance. The list entry's account of large multivallate hillforts names the forms to look for: a single gap in the rampart, inturned or offset ramparts, oblique approaches, guardrooms or outworks.
- Look inside. English Heritage's history of the site uses the Agency's LiDAR to pick out the bank barrow, part of the line of the Neolithic enclosure and early Iron Age defences, a Bronze Age round barrow, and the furrows of medieval and later ploughing. The curving bank across the middle, with its ditch on the west, is the western side of the first hillfort, whose builders followed the line of the Neolithic enclosure. English Heritage notes that the early fort's ‘western rampart and gateway can still be seen’, and that the rest of its circuit was taken into the later ramparts, ‘producing sharp changes of angle in the eventual plan’: look where the inner rampart turns sharply at the southern end of the curving bank.
- Name nothing without a record. The square and round hollows west of the curving bank are probably the ponds English Heritage describes: a square, clay-lined dewpond made for sheep in the 1860s, which replaced an earlier circular pond. The small square mark in the eastern half is about the size of the Romano-Celtic temple's walled precinct, 13 m by 12 m, which the list entry places at the eastern end of the fort. Faint parallel lines in parts of the interior, clearest in Local relief and the Openness composite, may be the cultivation ridges the Dorset Historic Environment Record describes, running east to west on the north side of the bank barrow and north to south on its south side, though lines this faint can also come from the survey itself. In each case the picture suggests and the record decides.
- Look around the fort. In Local relief the barrows show as pale round spots, and the table above gives their records. Some pale spots, on Hog Hill among other places, fall outside every scheduled area: a mark with no scheduled record may still be in the Dorset Historic Environment Record, or in no record at all, so check before you call it a barrow. In Field Archaeology, Find the HER gives the Historic Environment Record for any view and how to ask it for a search.
- Measure. A profile turns a picture into numbers. In Field Archaeology, an elevation profile along a saved line gives the height of the ground every metre, with the climb and the descent, and downloads as CSV; drawn across the ramparts, it shows each bank and ditch as a rise and a fall.
- Record what you saw, then check it on the ground. For each feature note its National Grid reference, the view or views you used and the survey date, and say which identifications come from a record and which are your own. English Heritage lists Maiden Castle as free to enter at any reasonable time during daylight hours. It is a scheduled monument: works there, including any excavation, need Scheduled Monument Consent, and using a metal detector or geophysical survey equipment needs a section 42 licence from Historic England.
What can these pictures not show?
A terrain model records the shape of the surface and nothing beneath it. Much of what the list entry describes was found by digging and lies buried: the Early Neolithic pits, the post holes and floors of houses, the storage pits and the burials. LiDAR shows such things only where they still change the shape of the ground.
Low features are hard to see. The bank barrow is 546 m long but only 0.5 m to 1 m high in the list entry, and English Heritage says later ploughing almost flattened it, ‘leaving it barely visible today’. English Heritage's own LiDAR figure picks it out; in these pictures it is hard to point to with confidence.
The terrain model leaves out buildings and vegetation, so these pictures say nothing about what stands on the ground now; the atlas's Surface model view, drawn from the Agency's surface model, shows trees, hedges and buildings. The Agency gives the vertical accuracy of its surveys as plus or minus 15 cm (root mean square error), and a 1 m grid cannot show detail much smaller than a metre or two. The seven views drawn by Field Archaeology can also sit 1 to 4 m off the Agency's own hillshade, because their positions use the Ordnance Survey Helmert transformation.
How should these pictures be cited?
The LiDAR is the Environment Agency's, released under the Open Government Licence v3.0, which asks for the Agency's attribution statement. Credit the heights to the Environment Agency with that statement and the licence, name the view, and give the date you used it, and the survey date where it matters; How to cite has the exact form.
- Attribution: © Environment Agency copyright and/or database right 2022. All rights reserved.
- A caption: Maiden Castle, Dorset, in the Local relief view, 25 October 2026, from the Environment Agency's LIDAR Composite DTM of 2022 (survey of 27 February 2021). © Environment Agency copyright and/or database right 2022. All rights reserved. Open Government Licence v3.0.
Cite the monument by its record. Field Archaeology's Cite button gives a list entry in this form, followed by its grid reference:
Historic England (2026) National Heritage List for England, list entry 1015775: Maiden Castle. Available at: https://historicengland.org.uk/
listing/ the-list/ list-entry/ 1015775 (Accessed: 25 October 2026). Open Government Licence v3.0.
Forms for the LiDAR itself, for historic maps and for other records are in How to cite.
Questions and answers
How many ramparts does Maiden Castle have?
Historic England's list entry describes three banks and two ditches around the hilltop, with an extra bank and ditch along the gentler slope to the south. Measured along one north to south line on the Environment Agency's 1 m terrain model, the ground shows the same: three banks and two ditches on the north side, and four banks and three ditches on the south.
Which LiDAR view shows Maiden Castle's ramparts most clearly?
Local relief makes the banks and ditches easiest to count, as bands of white and black. Sky view and Red relief also tell banks from ditches without depending on the direction of the light. The Environment Agency's hillshade looks the most natural, but it can make a bank and a ditch look alike.
When was the LiDAR of Maiden Castle flown?
The Environment Agency's 2022 composite terrain model is built here from one survey, flown on 27 February 2021 at 1 m resolution, according to the Agency's survey index.
Can the Neolithic bank barrow be seen on LiDAR?
Only just. The list entry gives its mound as 546 m long but only 0.5 m to 1 m high, and English Heritage says later ploughing almost flattened it. English Heritage's own LiDAR figure picks it out; in the pictures on this page it is hard to point to with confidence.
Sources
- Bennett, R., Welham, K., Hill, R.A. and Ford, A. (2012) ‘A comparison of visualization techniques for models created from airborne laser scanned data’, Archaeological Prospection, 19(1), pp. 41 to 48. Available at: https://doi
.org ./10 .1002 /arp .1414 - Chiba, T., Kaneta, S. and Suzuki, Y. (2008) ‘Red relief image map: new visualization method for three dimensional data’, International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, 37(B2), pp. 1071 to 1076. Available at: https://www
.isprs (Accessed: 25 October 2026)..org /proceedings /XXXVII /congress /2_pdf /11_ThS-6 /08 .pdf - Dorset Council (no date) Dorset Historic Environment Record, MDO3244: Maiden Castle, Winterborne St Martin. Available at: https://heritage
.dorsetcouncil (Accessed: 25 October 2026)..gov .uk /Monument /MDO3244 - Dorset Council (no date) Dorset Historic Environment Record, MDO47524: Medieval settlement, Maiden Castle, Winterborne St Martin. Available at: https://heritage
.dorsetcouncil (Accessed: 25 October 2026)..gov .uk /Monument /MDO47524 - English Heritage (no date) Maiden Castle. Available at: https://www
.english-heritage (Accessed: 25 October 2026)..org .uk /visit /places /maiden-castle / - Environment Agency (2022) LIDAR Composite Digital Terrain Model (DTM), 1m. Available at: https://www
.data (Accessed: 25 October 2026)..gov .uk /dataset /01b3ee39-da3f-47b6-83da-dc98e73a461f /lidar-composite-digital-terrain-model-dtm-1m - Environment Agency (no date) Survey index files: LIDAR Composite 1m DTM 2022 extents. Available at: https://environment
.data (Accessed: 25 October 2026)..gov .uk /spatialdata /survey-index-files /ogc /features /v1 /collections /LIDAR_Composite_1m_DTM_2022_extents - Hesse, R. (2010) ‘LiDAR-derived local relief models: a new tool for archaeological prospection’, Archaeological Prospection, 17(2), pp. 67 to 72. Available at: https://doi
.org ./10 .1002 /arp .374 - Historic England (2018) Using Airborne Lidar in Archaeological Survey: The Light Fantastic. Swindon: Historic England. Available at: https://historicengland
.org (Accessed: 25 October 2026)..uk /images-books /publications /using-airborne-lidar-in-archaeological-survey / - Historic England (2026) National Heritage List for England, list entry 1015775: Maiden Castle. Available at: https://historicengland
.org (Accessed: 25 October 2026)..uk /listing /the-list /list-entry /1015775 - OpenStreetMap contributors (2026) OpenStreetMap. Available at: https://www
.openstreetmap (Accessed: 25 October 2026)..org / - Oswald, A. (no date) History of Maiden Castle. English Heritage. Available at: https://www
.english-heritage (Accessed: 25 October 2026)..org .uk /visit /places /maiden-castle /history / - Sharples, N.M. (1991) Maiden Castle: Excavations and Field Survey 1985 to 6. English Heritage Archaeological Report 19. London: English Heritage.
- Wheeler, R.E.M. (1943) Maiden Castle, Dorset. Reports of the Research Committee of the Society of Antiquaries of London 12. London: Society of Antiquaries of London. Available at: https://doi
.org ./10 .26530 /OAPEN_1002324 - Yokoyama, R., Shirasawa, M. and Pike, R.J. (2002) ‘Visualizing topography by openness: a new application of image processing to digital elevation models’, Photogrammetric Engineering and Remote Sensing, 68(3), pp. 257 to 265.
- Zakšek, K., Oštir, K. and Kokalj, Ž. (2011) ‘Sky-view factor as a relief visualization technique’, Remote Sensing, 3(2), pp. 398 to 415. Available at: https://doi
.org ./10 .3390 /rs3020398