The Inter-City Villa
Lower Basildon, Berkshire
25 February 2001

Using aerial survey in identifying and recording archaeological features

Viewers of the Lower Basildon programme will have seen Bernard Thomason and colleagues plotting aerial photographs and maps to identify archaeological features on the ground. Here Simon Crutchley, Senior Investigator, Aerial Survey, English Heritage, describes the techniques they used.

Aerial photographs have been used for the identification of archaeological features since the early 1920s, when O G S Crawford and Alexander Keiller carried out a series of flights in the south of England leading to the publication of Wessex from the Air in 1928. Crawford then went on to map features seen on aerial photographs onto the Ordnance Survey map base.

He began by simply sketching the rough size, shape and position of the relevant features. This procedure was refined over the years with various manual plotting techniques. All of these methods – and, indeed, their modern counterparts – worked on the principle of producing a scaled plan view of a site by comparing the position of features visible on the photograph with their known position on the ground, taken from a map. These so called 'control points' tend to be features visible on maps, such as the corners of walls, posts, angles in hedges or anything for which the ground position can be obtained from a map or other source.

In recent years the speed and accuracy of plotting has increased considerably with the advent of computer programes designed to do the complex mathematical calculations required. There are now programes available that take a scanned image and, using control information supplied from a base map or other source, allow the production of a rectified and geo-referenced image (program details are given below). Depending on the initial scale of the photograph and the angle at which it was taken, the scale of the base map providing the control information and the relief of the land, sites can be plotted to within a metre or two of their ground position. It only takes one control point to have moved, however – such as a farmer moving a hedge or the Highways Agency straightening a road – for the ground accuracy to be effected. The same is true if any of the control points are at significantly different heights.

For the Lower Basildon mapping, because Aerial Survey did not have very long to produce the plot, we were unable to gain access to any vertical photographs which, when viewed in stereo, would have shown the drop in the south west corner of the field. This meant that our plotting of the site was less accurate than we would normally accept. Our normal levels of accuracy for detailed plotting are well within the tolerances of excavation and we regularly provide plots to archaeological contractors to enable them to place their trenches over features, without the need for any other form of remote sensing, such as geophysics. Once the image has been rectified it is then possible to trace over and highlight any archaeological features visible.

As with geophysics and other techniques, the technical side of plotting is only a small part of the work of aerial surveyors. The most important part of aerial survey is the recognition of archaeological features and their interpretation. Apart from the need to be able to filter out all those features appearing on photographs that are not of archaeological origin, the aerial surveyor is also able to give an interpretation of the presumed date and function of a given feature. This is based on their previous knowledge and comparison with known features.

For example, although it was not mentioned on the programme, when the Aerial Survey section at English Heritage produced the plot of features for the site at Lower Basildon, we suggested that the sub-rectangular feature in the centre of the field was quite possibly a Neolithic mortuary enclosure. This was based on comparison with other sites recorded as part of previous work in the Thames Valley carried out for the National Mapping Programme. This is a survey being carried out by the Aerial Survey section of English Heritage, the aim of which is to enhance our understanding about past human settlement by providing primary information and syntheses for all archaeological sites and landscapes (visible on aerial photographs) from the Neolithic period to the 21st century.

In practical terms, the purpose of the NMP is to map, document and classify – at a common scale and to a common standard – all archaeological sites and landscapes recorded in England on aerial photographs. This relatively small-scale mapping (at scale 1:10,000) has the advantage of being able to put sites into their context. Using aerial photographs we are able to map large swathes of land (such as the whole of the Thames Valley, Salisbury Plain or the Yorkshire Wolds) and interpret not just single sites but entire landscapes. We are able to relate settlements to one another and their associated field systems, and show the relationships over time of different ritual features, such as Bronze-Age round barrows clustering around Neolithic long barrows. The NMP has already mapped approximately one quarter of England and the results (both in the form of transcriptions and the interpretations of the monuments) are available for consultation at the National Monuments Record Centre, Swindon.

Aerial survey will never remove the need to carry out excavations as it is impossible to be absolutely certain of the date of many features based purely on their appearance from the air. But it is an important tool, which when used properly can provide a lot of information about a site, save time in the location of excavation trenches and put the site in a context unavailable from any other source.

Technical information

The main program currently used by Aerial Survey, English Heritage, is AERIAL, produced by John Haigh, of the Department of Mathematics at Bradford University. An alternative program, AirPhoto, has been developed by Irwin Scollar, at the University of Cologne. In general terms, both work on the principle of comparing each control point as taken from a photograph with each control point for which the ground co-ordinates are known. Then a series of complex mathematical formulae are used to ascertain the true ground position of every other point on the image.

In areas of exaggerated relief, it is possible to use a Digital Terrain Model (DTM), from whatever source, to further enhance the accuracy and remove any anomalies caused by height differences. With control information from 1:2500 scale maps an accuracy of +/- 2.5 metres is expected, though it is often considerably better. As the map scale goes up, so the accuracy decreases so that for 1:10,000 mapping the accuracy is in the range +/- 5–15 metres. While this is quite adequate for landscape assessment and general planning control, we obviously recommend a more detailed survey (1:2500 or larger) if archaeological investigation such as excavation is to be carried out.

Once the rectified image has been created it is possible to import it as a geo-referenced file into many applications, such as various GIS (Geographical Information Systems) packages, or the normal Aerial Survey working application, AutoCAD. Once imported, the archaeological features are then highlighted and can be printed out either with or without the photographic image.

Further information

For further information on the work of the Aerial Survey section at English Heritage visit the English Heritage website at www.english-heritage.org.uk. For further information on aerial survey in general or the specific techniques applied visit the website of AARG (Aerial Archaeology Research Group) at RS6000.univie.ac.at/AARG/. The site contains information on the group as well as brief introductions to the history of Aerial Archaeology and links to other sites dealing with international aspects of aerial survey, as well as the technical aspects of some of the programes and applications used.

Back to Basildon

Back to the Time Team Past programmes page

Back to the 2001 series page