LiDAR
Shape
Sees banks, ditches, terraces and cuttings that still alter the modern surface.
Silent when no relief survivesI · This begins where Ghost Roads ended
The terrain model is silent. The investigation is not over.
LiDAR measures surviving changes in ground height. If centuries of ploughing have flattened the road, or later deposits have buried it, the surface may appear perfectly ordinary. Archaeologists then look for other signals: how plants grow above the road and how radar energy reflects from its buried structure.
Multispectral imaging
Plant response
Detects differences in crop vigour caused by compacted road material and filled ditches.
Powerful in dry arable landscapesGround-penetrating radar
Buried structure
Records reflections from changes between stone, gravel, soil and ditch fill below the surface.
Adds depth—but depends on the groundThe central idea
The past does not leave one kind of trace.
Understanding a lost road means choosing the method that matches what the landscape is still able to preserve.
II · The crop remembers
A buried road can change the plants growing above it
Compacted stone and gravel can restrict roots, moisture and nutrients. Filled roadside ditches may hold deeper, richer soil. In the right season, the road and its ditches write their pattern into the crop. 1
01
Bare field
The buried road is invisible at the surface.
02
Young crop
Early growth may still appear almost uniform.
03
Dry summer
The road produces stressed growth; the ditches may remain greener.
04
After rain
The contrast may weaken or vanish until another favourable season.
What the visitor is seeing
The crop is acting as a temporary detector. It does not show the road directly; it shows how the buried road changes water, roots and plant health. A cropmark may appear clearly in one drought and remain invisible for years afterwards.
III · Beyond visible colour
The crop can be stressed before the eye can see it
Multispectral sensors record selected wavelengths outside ordinary red, green and blue photography. Near-infrared and vegetation indices can strengthen subtle differences in plant vigour. 2
Ordinary colour photograph
The cropmark may be faint: a pale central strip with slightly stronger growth along the possible ditches.
Illustrative comparison—not recorded multispectral data.
Shows colour and height differences already apparent to the camera and human eye.
Responds strongly to plant health and can expose stress before visible colour changes become clear.
Combines spectral bands to emphasise relative plant vigour across the field.
Go deeper: what is NDVI?
The Normalised Difference Vegetation Index compares near-infrared and red reflectance. Healthy vegetation generally reflects more near-infrared energy than stressed vegetation. The calculation is useful, but the archaeological interpretation still depends on soil, crop, season, recent farming and the shape of the anomaly.
NDVI = (Near-infrared − Red) ÷ (Near-infrared + Red)
IV · Radar beneath our feet
Cropmarks suggest a buried road. Radar can ask what survives below.
Ground-penetrating radar sends electromagnetic pulses into the ground. Reflections return when the pulse crosses materials with different electrical properties—such as soil, stone, gravel or ditch fill. The delay helps estimate depth. 3
Typical rural Roman-road survey
A medium-frequency antenna balances detail and penetration, seeking the metalled surface, foundations and flanking ditches within the upper metres.
Simplified explanatory diagram. Actual depth and clarity depend on equipment, soil and moisture.
Higher frequency
Lower frequency
Finer detail
Useful for shallow surfaces and small construction features.
Balanced survey
Common archaeological compromise between resolution and depth.
Greater reach
Can investigate deeper deposits, but individual features become less distinct.
These are principles, not guaranteed depths. Clay, salt water, saturation and modern road material can absorb the signal before it reaches the archaeology.
V · Four landscapes, four outcomes
The method succeeds only when the ground allows it
These cases show why archaeological survey is not a contest between technologies. Each method contributes a different layer—and sometimes the most important result is understanding why no result appeared.
Camp Hill · Nottinghamshire
One method saw the road; another saw what the road surface concealed
LiDAR revealed the Roman road where relief survived. Multispectral drone data exposed additional archaeological features that extensive agriculture had left with too little topographic expression for LiDAR. 4
What it teaches
Different sensors can reveal different parts of the same site.
Roman York · Eboracum
Success and failure in the same city
On open grass, a shielded 250 MHz system detected Roman barracks, streets and structures at depths of roughly 1.5–2.5 metres. Along the modern city streets, known archaeology remained invisible below about one metre because road makeup, clay and sediments attenuated the signal. 5
What it teaches
More sophisticated equipment cannot overturn the physics of difficult ground.
Silchester · Calleva Atrebatum
A Roman street plan beneath open ground
With much of the town free from later urban development, GPR and other geophysical methods revealed the regular street network and associated buildings in remarkable detail. 6
What it teaches
Good preservation and suitable ground can turn a field into a buried city map.
Binchester · Vinovium
A route assembled from several forms of evidence
Geophysics showed Dere Street through the fort. Aerial evidence suggested a route beyond it, and LiDAR helped extend the probable line into the surrounding landscape. 7
What it teaches
The strongest reconstruction is often a conversation between methods.
VI · Where the methods fall silent
Every technique has a landscape in which it struggles
☼
Multispectral imaging
Needs a responsive crop and the right season
- No arable crop: pasture and woodland provide no useful cropmark.
- Wet conditions: moisture stress may never become strong enough to reveal the feature.
- Clay soil: retained moisture can reduce differences in growth.
- Urban ground: hard surfaces prevent cropmarks entirely.
- False patterns: drains, pipelines, removed boundaries and farming can resemble archaeology.
- No depth: the image cannot show how deeply the road is buried.
◌
Ground-penetrating radar
Needs the signal to travel and the materials to contrast
- Clay: electrical conductivity can rapidly absorb the radar energy.
- Salt water: saline groundwater may reduce penetration almost to zero.
- Waterlogging: saturated ground can weaken useful reflections.
- Similar materials: a lightly built earth road may not contrast with its surroundings.
- Large landscapes: detailed survey is slower and more expensive than broad reconnaissance.
- No date: a buried road-like structure is not automatically Roman.
VII · The archaeological toolkit
Choose the method to match the trace
No single technique replaces the others. The investigation becomes stronger as different forms of evidence converge.
Method
What it detects
Best landscape
What it cannot prove
LiDAR
Surviving banks, ditches, terraces and cuttings
Woodland, upland, pasture and preserved relief
Date or buried structure
Multispectral imaging
Crop stress and vegetation response
Dry, well-drained arable fields
Depth, construction or Roman date
Magnetometry
Magnetic differences, especially filled ditches
Rapid survey of broad open areas
Stone surfaces where magnetic contrast is weak
GPR
Depth, road layers, ditches and buried structures
Targeted rural, urban and hard-surface investigation
Date where no datable material is recovered
Maps and documents
Lost alignments, former boundaries and inherited routes
Any landscape with useful historical records
Whether a mapped line originated in the Roman period
Excavation
Construction sequence, artefacts and datable contexts
Targeted locations selected from earlier evidence
The complete long-distance route from one small trench
1Find a possible corridorLiDAR, aerial evidence, maps or modelling
→
2Test the anomalyMultispectral, magnetometry or targeted GPR
→
3Understand the structureDepth slices, profiles and combined interpretation
→
4AuthenticateTargeted excavation and dating evidence
VIII · Go deeper
Continue only as far as your curiosity takes you
The investigation above contains the main explanation. These links are optional next steps for visitors who want the full technical report, the preceding LiDAR investigation or the original case-study research.
Companion report · Full technical study
Multispectral Imaging and Ground-Penetrating Radar in the Search for Roman Roads
Equipment, wavelength bands, antenna frequencies, vegetation indices, British case studies, limitations and the full bibliography behind this investigation.
Read the report → Previous investigationGhost Roads
Return to the first part of the story: what LiDAR reveals on the surface and why a visible line is not proof.
Explore Ghost Roads → Accessible official guideHow cropmarks form
Historic England explains how buried walls, ditches and compacted surfaces alter crop growth.
Open the guide → Accessible official guideHow ground-penetrating radar works
Wessex Archaeology explains radar pulses, reflections, survey grids and three-dimensional data.
Open the guide → Specialist case studyRoman York beneath the streets
Read the 2025 city-centre GPR report, including the important negative results and signal attenuation.
Open the report → Roman Storyworld archiveRoman roads and landscape research
Continue into the wider archive of route studies, field observations, maps and Roman Britain research.
Browse the archive →IX · Sources and further reading
Where the evidence comes from
The main claims on this page are grounded in archaeological guidance, published research and specialist survey reports. The list is provided for transparency; it is not required reading.
Historic England — The Formation of Cropmarks
Official explanation of how buried roads, walls and ditches alter crop height and colour.
Open source →Historic England — Satellite, Multi- and Hyper-Spectral Imagery
Guidance on archaeological uses of wavelength bands beyond ordinary photography.
Open source →Wessex Archaeology — Ground-Penetrating Radar
Practical explanation of radar transmission, reflections, depth and array survey.
Open source →Brooke and Clutterbuck (2020)
Multisensor drone survey at Camp Hill and Shelford Priory, demonstrating that no single technique revealed every feature.
Open paper →Roman York Beneath the Streets reports
Open-area successes, city-street attenuation and the effect of ground conditions on GPR performance.
Open Report 3 →Linford (2004) and the Silchester Mapping Project
Foundational GPR work over Roman remains and the geophysical mapping of Calleva Atrebatum.
Open paper record →Roman Roads Research Association Newsletter 7 (2018)
Geophysics at Binchester and examples of the multi-method investigation of Roman roads.
Open newsletter →Verdonck, Launaro, Vermeulen and Millett (2020)
Large-area multi-channel GPR survey of the Roman town at Falerii Novi.
Open paper →Roman Storyworld — Companion technical report
The full analysis of multispectral imaging, GPR equipment, case studies, limitations and the silver-bullet question.
Read the report →What to take away
The hidden road may survive as shape, stress, reflection—or only an unanswered question.
LiDAR sees relief. Crops respond to buried ground. Radar measures contrasts below the surface. Maps preserve inherited lines, and excavation supplies the evidence needed to establish date.
The skill lies not in finding one perfect machine, but in recognising which trace remains and choosing the right way to see it.