Ground Penetrating Radar (GPR) has proved successful in detecting cavities which is of importance in the fields of civil engineering, building assessment and archaeology. The pulse emitted by the GPR antenna is reflected comparatively strongly by the large dielectric difference between the surrounding medium and the air in the cavity. Additionally the geometry of the problem usually allows for multiple reflections (multiple bounce).
Using a broadband system like GPR enables a large amount of information to be gathered simultaneously. Using signal analysis models developed for atmospheric radar, more than just location information can be extracted from the data. By using a method of spectrum analysis based on exponentially damped sinusoids - Prony's method - resonance's of the targets can be detected. This is especially valid in the late-time, after the forced response (first reflection from any boundary).
One of the first applications of time slice analysis of ground penetrating radar (GPR) in archaeological prospection was reported by Nishimura and Kamei (1990). Since this early application several advanced methods for creating topographically corrected time slices and general horizon slicing has also recently been presented (Goodman and Nishimura, 1994). The advancement of static corrections in the geophysical processing of raw radar records has significantly improved the quality and utility of ground penetrating radar datasets collected at archaeological sites with topographic relief. Several examples of topographically corrected time slice datasets from Kofun Period (300-700 A.D.) stone chambers in Japan and Mississippian Indian house mound sites (1450 A.D.) in the United States are presented.
Horizon slicing (Telford et al., 1977), which is the most general form of slicing a user defined surface through a 3-D dataset, is used to image structures below subsurface interfaces that have unequal overburden thicknesses. The apparent reflection slopes of (level) built dwellings on raw radargram caused by the variable overburden, can be effectively removed during the horizon slice correction. One example of a (customised) horizon slice applied in the search of a Jomon Period (1000 B.C.) habitation site in Guma prefecture, Japan was recently analysed (Goodman, 1995). The results of horizon slices made below a variable pumice/ash layer that violently buried the site, are used to find traces of ancient dwelling foundations.
In recent years, interest in the use of acoustic methods to image the shallow subsurface has increased greatly: for medium- and large-scale applications considerable effort has been put into borehole methods, such as acoustic diffraction tomography, and there have even been attempts to employ surface-based methods for the detection of larger-scale archaeological features.
Acoustic approaches do hold out significant potential advantages over other techniques: to a first approximation, the propagation of sound waves through the ground can be likened to that of the electromagnetic waves of ground-penetrating radar (a fact which has led to attempts to use established seismic data processing tools for GPR), but the speed of propagation of the acoustic waves is far smaller, requiring much less expensive control and detection electronics, and the circumstances where radar is most ineffective (wet soil) are those where acoustical techniques might be expected to perform best.
A major hurdle with acoustic approaches is transmitting sufficient energy into the ground: as with GPR, any ingoing signal will largely be reflected upwards from the ground surface - common approaches to overcome this problem involve the use of heavy hammers or weights and metal plates, or even shotgun cartridges.
A preliminary theoretical study has been carried out to establish whether there is potential for a more accurate approach using a more sophisticated electronically-controlled system with a rather better-defined input signal. A detailed discussion is given of the potential merits and demerits of a high-resolution acoustic system, specifically intended for shallow detection, including the difficulties of field use, its archaeological potential in terms of the likely size and nature of detectable targets, and the degree of complexity of the recorded field data.
Archaeological prospection borrows its techniques from geological geophysics and adapts them. Over the years many different methods have been tested and a handful has proved to be extremely transferable. It is only recently that seismic techniques, primarily seismic refraction, have been tested in archaeological applications. This has arisen from the growing interest in obtaining depth information and the success and failures of ground probing radar. Seismic refraction has the advantages of simple, albeit slow, data acquisition and relatively straight forward interpretation. It is suitable on sites with a clayey soil and provides true depth information giving a profile of buried near surface layers.
Over the past few years Geophysical Surveys of Bradford has carried out several seismic refraction surveys on various sites. This paper aims to discuss the results obtained so far looking at the advantages and disadvantages of the technique. The possible limitations of the method with respect to the surveys already undertaken will be discussed together with possible problems that may be encountered at other sites and how these may be over come.
Until recently, investigation of the third dimension of archaeological structures relied upon the destructive technique of excavation. The research detailed in this paper will illustrate the credibility of producing realistic estimates of depth using a Twin-Probe based method and an Electrical Imaging System. Results will be analysed from Lambeth Palace gardens, London and Wroxeter Roman Town. The former was instigated by Channel 4 'Time Team' and based on an excavation carried out by Davis in the 1930's who suggested the presence of a Roman Road within the grounds of Lambeth Palace. A Twin Probe resistivity technique (0.5m spacing on a 1m grid) was employed to provide a plan of buried remains. This allows a plan of building features, medieval ditches and recent garden features to be developed. In this project, electrical imaging was employed to add the third dimension to buried features. Selected segments along the electrical imaging sections were later excavated. The twin probe results, electrical imaging and excavated sections will be shown and together shed light on the history of the site. Investigations at Wroxeter involved using gradiometry to locate buildings and roads which were targeted using the Campus and the modified twin probe imaging system. The aim was to compare the results of the two techniques.
At present, many resistivity measurements are made with one probe configuration, and hence one investigation depth, with the aim of covering as large an area as possible, and then separate depth investigations, if required, are applied to selected section lines, usually with a line of stationary multiplexed electrodes. A new, portable multiplexing system, using a set of mobile electrodes, is described which offers the prospect of simultaneous area and depth investigation. Increased data sampling over the whole of the area, rather than just selected lines, offers scope for better interpretation of the site as a whole. The multiplex system may also be used to greatly speed up conventional Twin area measurements or provide more detailed area coverage with virtually no increase in survey time, by using parallel arrays. The system can be used to configure most common probe arrangements and will automatically capture single or multiple readings at a station, at up to 4 readings per second.
The system comprises an RM15 resistance meter, PA5 probe array and the new MPX15 multiplexer module. The PA5 probe array frame (based on a standard 0.5m Twin frame with additional wings) can be configured for a length of between 0.5m and 2m. The PA5 length has holes positioned on a 0.25m pitch, allowing 2 to 6 probes to be fixed in position. The RM15 resistance meter and MPX15 multiplexer are also mounted on the frame and determine which probes are utilised for each measurement - an additional pair of remote probes can also be connected for Twin, Pole-Pole and Gradient measurements. The frame is moved by the user, from station to station, whilst the multiplex system automatically logs data and steps through the reading sequences as soon as the probe are inserted into the ground. The MPX15 is controlled by the RM15 and uses a matrix of solid state relays, rather than conventional reed relays, which results in a system of higher reliability, lower power consumption and reduced weight and size, allowing it to be easily mounted on the frame.
A number of pre-programmed configurations and sequences are provided, or the user may configure and store their own measurement sequences and configurations. Pre-programmed configurations include all the common arrays, parallel Twin arrays (2, 3 or 4) for increased speed or increased area resolution, and multiple Twin (2 or 3) for depth investigation. The 8 user defined programs, with up to 8 sequences per program, could be used, for example, to make simultaneous measurements at a station with 0.25m, 0.5m, 0.75m, 1.0m, 1.25m, and 1.5m Twin arrays, along with 0.5m Wenner and Double-Dipole arrays. It is also possible to make up to 8 Twin measurements (0.25m to 2m) with the 6 probes, providing software is used to realign the array centres if pseudo-sections are to be generated. The user has complete control over the programmed measurement sequence, which is permanently stored in the RM15 for future use.
A preliminary study is presented of the possibilities and performance of this system, including the results of tank experiments and field trials. Geoplot 2.1 software is used to (a) produce stacked pseudo-sections across the whole of the survey areas, (b) normalise these sections for improved presentation, (c) present the area data sets associated with each probe configuration side by side (normalised for easy comparison) and (d) subtract data sets from one another for geology correction.
Data visualisation has several stages. First there is the rapid evaluation in the field. Then there is the laborious process of interpreting, combining and processing the data and the last stage is making the data nice, smart and otherwise visual for the contractor.
Every step needs its own programs and ways of looking at the data and the visualisation process.
During this presentation several programs used by Stichting RAAP will be discussed with their pro's and con's.
The Insite Program has been developed to enable the downloading, processing and visualisation of archaeogeophysical and topographic data under Microsoft Windows. InSite controls the task of printing on a range of hardcopy devices.
InSite uses a layer model to manage and display site features, grid maps and images together with their geophysical and archaeological interpretations. Up to 3 geophysical or topographic data sets can be manipulated simultaneously and be combined using logical bitmap or arithmetic operators to explore reduction of artefacts specific to magnetometer data such as striping and staggering on zigzag traverses, instrument drift and periodic defects due to instrument oscillation or operator magnetisation.
The program has been evaluated during a project, sponsored by English Heritage, to map the medieval Friary and Civil War defences on Hartlepool Headland by geomagnetic and resistivity survey. High concentrations of ferrous litter and pools of surface water produced data artefacts which were reduced by appropriate processing and a number of subsoil features of archaeological interest detected.
This paper discusses the interpretation of archaeological gradiometer data using a system called Integra developed at Bradford University in collaboration with Geophysical Surveys of Bradford. The system is a Windows based package, offering all the traditional processing capabilities, as well as proprietary algorithms designed specifically to aid the analysis of gradiometer data.
First a feature location algorithm, designed around a gradient filter, detects magnetic anomalies within a data set. This information, as well as being a visual aid to interpretation, is then used to calculate 2-D profiles across each anomaly. These profiles are then presented to a neural network, trained to recognise magnetic anomalies. The output of the network is the top face depth and width of the anomaly's source. This provides a map of the anomalies on a site, their depths of burial and top face widths.
The design of the system and the specialist algorithms are discussed briefly, followed by a case study demonstrating the application of this approach. The performance of the location algorithm is compared with current approaches and the calculated depths and widths are assessed by comparison with excavation results.
Although a range of geophysical tools are available for probing the subsurface, the equipment and interpretative techniques available have usually first been developed for larger-scale applications such as mineral exploration. For many large-scale applications mere detection of an anomaly is sufficient in that it is intended to guide later borehole exploration - the extraction of detailed information, although desirable, is a secondary concern.
Archaeological prospection imposes a rather different set of requirements and problems: not only is a much higher resolution required, but the features to be detected are much closer to the survey tool, the effect of the surface topography is different, the sources of spurious noise (unknown moisture levels, non-archaeological soil variations) are different, and the survey/sampling strategies are different. The particular difficulties associated with shallow prospection are the driving force behind the development of new surveying equipment such as multiprobe resistive tomography and continuously profiling probe arrays: the aim of such equipment is to gather both wider ranges and larger quantities of field data.
The clear objective is that the increased information should allow not only the detection of a typical anomaly, as in standard magnetic or resistivity profiling, but should also provide "diagnostic" information, such as size, depth or the degree of contrast to the natural background. The development of such new techniques brings the promise of more useful geophysical surveys, but this will only be so if the recorded data set can be interpreted in a quantitative, reliable and physically realistic fashion.
Here, some of the difficulties of handling the increased quantities and types of field data are outlined and examples are given which show that the increased quantity and complexity of the field data is not along sufficient to achieve the objectives. It is also necessary to incorporate "non-experimental" information about the intrinsic geophysics of the particular survey technique, equipment characteristics, physical constraints and/or assumptions relating to physical continuity and smoothness: the inclusion of such non-experimental information reduces the intrinsic ill-conditioning of the geophysical data inversion problem. The implications for archaeological geophysical prospection are discussed, in terms of cost, complexity, utility and survey strategy.
Specialist interpretation of aerial photographs and accurate mapping of those features identified will provide a sound framework on which to base other forms of archaeological survey. This has become routine in many developer-funded assessments but has rarely happened on an extensive scale in Britain. It is proposed that there are historical reasons for this as well as there being present-day differences in interests of the principal funding bodies.
Recent survey of the East Anglian Fenland has allowed results of a long-term programme of field walking to be combined with those from interpretation and mapping of aerial photographs. Examples will identify the efficacy and limitations of each form of survey and demonstrate the considerable increase in archaeological information that is gained by integrating the two. Field survey and air photo interpretation in the Fenland were carried out in succession rather than in parallel and can readily highlight how each method can assist the other. This leads on to suggest ways in which our designs for archaeological surveys can be improved.
Most techniques for non-invasive archaeological prospecting yield data amenable to processing and integration in a GIS, in that they can be represented as two-dimensional raster map layers or images. Examples of such data sets are thematic images captured by satellite and airborne scanners, vertical and oblique aerial photographs, and earth resistivity, conductivity, and magnetometry measurements.
The Wroxeter Hinterland Project (WHP) attempts to make full use of this capacity by applying GIS technology to integrate such non-invasive data with data gathered using geochemical sampling, archaeological field walking and excavation, (micro-) topographic survey, and digital maps of landscape properties such as soil types, elevations, and land use.
Ground-based remote sensing techniques comprise a range of specialised, scientific photo-imaging and computer enhancement methods capable of revealing invisible, or poorly perceived, details from man-made structures and artefacts, particularly historic buildings and their associated fixtures and fittings. Several classes of information may be sought using these tools, ranging from geological differences in building fabric, through features hidden below paint and plaster, to indecipherable inscriptions. The methods are entirely non-destructive and non-intrusive and may be employed in situations where physical examination of historic fabric is not permissible or practicable. It is possible to image relatively large areas quickly, making the procedures of great value in situations where emergency recording is needed. Many of the techniques are also applicable to the examination of archaeological sites both before and during excavation, and have proved particularly useful in the enhancement of poorly defined features in soil. Recent developments have focused on the use of computer image enhancement methods to extract further information from the imagery derived in the field.
Magnetic prospection by a high-resolution caesium magnetometer (1 x 10^-11 Tesla) and aerial photographs were used at the Bavarian State Conservation Office in order to detect and to describe precisely the buried heritage. Digital image processing of aerial photographs and of the magnetic data enables us to make detailed plans of the archaeological sites. Case histories of magnetic and aerial prospection for different soils will be shown. At these sites, archaeological excavations were performed which allowed sampling of the anomaly and surrounding undisturbed soil. Magnetic characterisation was done by measuring the natural remanent magnetisation NRM, the susceptibility k, the anhysteretic remanent magnetisation ARM, the saturation remanent magnetisation SIRM and the remanence coercivity (BO) CR on undisturbed soil samples. The demagnetisation behaviours of the NRM, the ARM, and the IRM were determined.
Generally, an enhancement of ferrimagnetic minerals is responsible for the formation of magnetic anomalies in soils at archaeological sites. Specifically, magnetite and maghemite were identified by Curie-temperature analysis, Verwey transition and by X-ray diffraction.
A Neolithic structure serves to show the influence of the susceptibility and the NRM on the intensity of magnetic anomaly. Undisturbed areas of this structure yielded higher intensities of the magnetic signal, whereas parts of a ditch, which have been excavated and refilled in 1914, showed a lower signal due to the loss of the NRM during excavation. Generally, undisturbed structures seem to have stable directions of the NRM due to alternating-field demagnetisation. Königsberger Q factors (i.e. the ration of the NRM to the induced magnetisation) range from 0.8 - 1.2 and confirm these observations.
For a geophysical assessment of buried archaeological sites different geomagnetic methods are available which can be combined to form even more powerful tools. For example, results of magnetometry surveys can be more readily understood if the magnetic properties of the soil are known. Investigating these properties in their own right can contribute considerably to the understanding of a site, especially if their vertical distribution is studied. The stratification of soil properties is closely related to human occupation and is well suited to identify different phases. Core samples may be taken for investigation. A device is introduced which allows the continuous recording of the magnetic susceptibility response along such cores. Data obtained with this device are subjected to mathematical analysis (maximum likelihood estimation) in order to obtain the true susceptibility at any position along the core. Details of the device and the mathematical procedure will be presented together with the results from archaeological sites.
Magnetic Susceptibility is being more widely used in the evaluation of sites, but how can areas of archaeology enhanced magnetisation be separated from natural variations in MS due to the underlying geology? This paper discusses the problem, and by using some recent field survey results looks at how regression analysis of topsoil and subsoil susceptibility data, combined with determinations of percentage fractional conversion, can help in sorting out the archaeology.
In the face of such speedier and higher resolution techniques as gradiometry, resistance or aerial photography does magnetic susceptibility have any role to play in archaeology? Likewise, P analysis is slow and produces a low resolution picture of a site. With alternative means of site prospection and delimitation, why bother with MS and P analysis at all?
Whilst the efficacy of these two techniques, especially P analysis, has been overstated in the past, they do have several advantages which warrant their inclusion in the usual battery of archaeological site prospection and interpretation techniques.
Mass specific magnetic susceptibility () and total P analysis on soil augered samples are depth specific to within centimetres. Hence, the adulterating and interfering influence of, for instance, a colluvial/alluvial covering or a disturbed Ap horizon can be overcome. Also, this depth specificity allows for some disentangling of the various 'stratigraphies' which comprise the composite picture of a site as visible to surface techniques.
Furthermore, and, especially, total P analysis have an important micro- and meso-scale 'forensic' function which assists in archaeological explanation and interpretation. Analysis on such a scale overcomes the problem of 'low resolution' results.
There are many other considerations and situations which can militate against gradiometry, for instance, but not total P and analysis and, hence, they are often a useful companion or alternative to other techniques.
Total P and analysis have a valuable contribution to make to archaeological investigation so long as the practical and theoretical constraints borne of their medium of application (the soil) are kept in mind. This paper will discuss and illustrate these points.
Since 1988 a large-area geophysical prospection of the plateau region south of the well known settlement mount of Troia (Troy) is being performed as part of the renewed archaeological investigations. Its aims are to investigate the so far unknown settlement pattern of the Hellenistic and Roman city of Ilion (period Troia IX) and to detect architectural remains and defence structures of a lower settlement of the Bronze Age period Troia VI. Such structures had been postulated to exist on the same plateau within a distance of about 400m from the fortified citadel. They would be buried underneath the Roman city down to a depth of about 3m below the present surface.
Magnetic prospection of stone architecture turned out to be feasible due to a high amount of magnetite in the cultural debris of the area thus providing a good susceptibility contrast. Initially the measurements were performed with a fluxgate gradiometer, which appeared to be adequate for the upper strata. The deeper lying Troia VI structures, however, required the application of a more sensitive, optically pumped caesium magnetometer. The support with this instrument was provided by H. Becker and J. Faßbinder, both at the Bayerisches Landesamt fur Denkmalpflege, Munich, Germany.
The major results of the combined measurements at 0.5m intervals on more than 20 hectares will be discussed. They provide a clear image of the orthogonal street plan and even of individual large buildings of the Roman city, and allow the identification of a defence structure around the lower city of Bronze Age Troia VI. Examples of correspondingly planned excavation soundings are given.
For 10 years the "archaeometry working group" of the institute of geophysics (Christian-Albrechts-Universit, Kiel) has carried out high resolution geophysical measurements for archaeological prospection at home and abroad. Generally, the need for fast high resolution geophysical methods and improved interpretation techniques as well as the appropriate graphical presentation has risen rapidly to assist in solving the problems of archaeological research. There are financial restrictions concerning the excavation of large areas. Non destructive methods are used to get more underground information in advance which should be linked with the details of the surrounding landscape and the ancient environment. The different applied methods are georadar, magnetics, resistivity and electromagnetics. Their combined application for determining geometry, material as well as lateral and vertical extension of the actual archaeological target will be demonstrated. Therefore, examples from the latest field campaigns will be shown (1993 and 1995). The sites date from different periods. Moreover, the findings are located in varying pedological and geological surroundings. The wooden remains of a Medieval refuge in a peaty soil are prospected with the georadar (SIR 10, 120 MHz, 500 MHz, planar interpretation technique) and electromagnetics (EM31) in the northern part of Germany. (Schleswig-Holstein). Celtic and Roman house and temple grounds are detected by magnetics (array of five parallel Fluxgate magnetometers) and georadar measurements in the western part of Germany (Rhineland Palatinate). Finally, the remains of a Hittite settlement are found using magnetics, resistivity (ECS960) (developed at the institute of geophysics, Kiel) and georadar in the eastern part of Turkey.
The Chang Jiang (River Yangtse) supported an ancient civilisation which depended upon its waters, and has received the attention of archaeologists from many countries.
In December 1994, construction of the biggest dam of this century was started on the Chang Jiang. As a result, in 1997 about 632 km^2 of land will be covered by water. Archaeologists have established that, overall, about 828 valuable sites will disappear. Important among the many sites in San Xia are Ba Gou and Chu Gou.
In order to record these cultural assets and to protect their contents, the Chinese government decided that research and excavations should be completed within 7 years. There is little time to spare, and they regard it as important that scientific methods should be used. In April 1994 San Xia was excavated by a combined Chinese and Japanese expedition.
The lecture will describe the geophysical location of ancient graves in San Xia, using radar, soil resistivity and magnetometry. It will also summarise the current state of development of archaeogeophysical survey in China.
The use of geophysical techniques in Irish archaeology has, until very recently, been a neglected area of study and application. Although the first ever application of geophysics in Irish archaeology dates back to 1952, there are few records of further geophysical studies or applications during the 1960's - 70's. Indeed, the application of geophysical exploration to archaeological sites in Ireland did not really begin until the early 1980's when Ronnie Doggart, a graduate student of Queen's University Belfast, carried out several magnetic surveys on archaeological sites, both for research and rescue purposes.
Although Doggart departed from Irish archaeology in the mid 1980's, his work contributed to a growing awareness of the potential of geophysical methods in site discovery and interpretation. At that time the Dept. of Archaeology at University College Cork decided to introduce geophysical prospecting into its teaching and research programme. An electrical resistivity meter, a fluxgate gradiometer and magnetic susceptibility equipment were purchased, but the department was not in a position, at that time, to employ qualified personnel to use this equipment. This led directly to my interest in the topic of geophysics and in 1987 I commenced an MA thesis on the use and applications of the techniques of electrical resistivity in Irish Archaeology. However, it was not until recently that I could prove or disprove my interpretations of the surveys undertaken as part of my research, as it has only been in the last year that excavation results have been formally written up.
The true potential of geophysical investigation has recently been highlighted by the work of the Discovery Programme. This government sponsored programme has, initially, taken the Bronze Age, as its research area / era, and has used outside consultants from Britain to undertake the geophysical work. The results, to date, have been very successful, and there is now a clear long-term need to develop some home-based expertise. The development of geophysics in Irish archaeology would be greatly encouraged by training initiatives in Irish Universities, such as the development of an M.Sc. programme in archaeological site science; there is a pressing need to develop a specialised remote sensing unit, attached to the National Monuments / Archaeological survey sections of the Office of Public Works, while within the private sector, the needs of a growing number of archaeological consultancies are bound to stimulate demand for geophysical expertise in the coming years. It is important that the results of future geophysical investigations be published, to further our knowledge of this important scientific application in Irish archaeology.
This paper will investigate the results of an integrated survey and research excavation of a multiperiod settlement mound at Scatness, Shetland, to be carried out in 1995. This settlement mound contains a broch (truncated by a modern road) and 3 metres or more of stratified archaeological deposits. This paper will explore the combined use of three dimensional topographical survey with magnetic and earth resistance surveys using the twin probe configuration (at both 0.5m and 1m separation of the mobile probes). The geophysical survey of multiperiod 'tell-like' settlement mounds found in the Northern Isles presents a number of problems to the archaeologist in gathering pre-excavation data. The mound is not single phase but represents a complicated stratigraphic sequence. Experience has shown on other settlement mounds the value of integrating both magnetic and earth resistance data sets in interpreting the underlying deposits. This, combined with a detailed survey of the topography at the Scatness site and the resolution at greater depth provided by the increased probe separation, will provide greater detail of the buried archaeology and will allow the targeting of excavation areas to address specific questions on the development of the settlement mound.
This paper will describe the results of high-resolution geomagnetic, resistivity and topographic surveys of land within and around the fort of Bremenium in Northumberland. The findings of the research include:
The paper will also describe the use of electrical resistivity tomography for obtaining vertical sections through the annexe defences.
The hillforts of central southern England are impressive landscape features requiring active conservation and management. Geophysical survey can play an important role in the response to these needs. Many hillforts are in public ownership and accessible to visitors, and require careful management and the provision of informed interpretation to the public. Countryside Stewardship schemes are increasing the number of hillforts in this category. Excavation can inform the management and interpretation process, but large scale archaeological interventions are expensive and cause unnecessary damage. Only a minority of hillfort interiors have been extensively excavated in the past and the internal layout of hillforts is still only poorly understood, while surface evidence for their utilisation is often absent. Geophysical survey, with an emphasis on magnetometry, can provide a cost-effective solution to the problems of interpreting hillforts where past excavation evidence is lacking and where future excavation is not an option. This paper will review recent geophysical surveys carried out by the Ancient Monuments Laboratory and links their results with the management and study of hillforts and their environs. It will also preview future research that is planned to develop further the application of geophysical techniques to the mapping of hillfort interiors.
Northamptonshire Archaeology over the last few years has developed the use of scan/reconnaissance surveying technique as part of their services to carrying out evaluations. These are a common requirement of the archaeological assessments and evaluations which have been sought since November 1990 under the recommendations of PPG16
The surveys are carried out using FM18/36 Geoscan Fluxgate Gradiometers. The instrument continuously monitors the geomagnetic field as the operator walks across the area of survey. The surveys are undertaken at a rapid walking pace and where significant anomalies are located, each are marked on the ground. These are subsequently mapped onto an OS base map.
Reconnaissance surveys are undertaken at varying intervals along parallel transects. All anomalies that are mapped are analysed and selected for detailed survey. Ensuing trial excavation provides a useful test of specific anomalies and assigns a context to them.
A number of examples will be drawn upon in this paper to illustrate the successes and failures of the techniques.
Whilst time and effort have been expended in developing methodology and data processing techniques little formal attention has been applied to making accurate predictions from the data. Most work has centred around theoretical modelling and this is an important aspect of developing precision and accuracy of interpretation. However, it is widely accepted that perfect, theoretical conditions are a far cry from those encountered in real-life situations. Furthermore, the philosophical approach to the interpretation of geophysical surveys has concentrated on achieving a clear understanding of buried archaeology without turning a sod, thus is the nature of non-destructive archaeological techniques. The thought that perhaps the geophysical results are providing much information of intrinsic archaeological vale, rather than just mapping information, is greatly neglected.
This paper will start by looking at the interpretation of a number of surveys and the assumptive nature of most geophysical interpretation. It will take three case studies of sites that have subsequently been excavated and highlight the difference between a geophysical interpretation and the archaeological reality. Further case studies will be presented demonstrating how prior knowledge of the geophysical results and the use of geophysical methods during both excavation and post excavation enabled information to be gained about sites which might otherwise have been lost. The presentation will conclude by proposing ways in which geophysical methods can be included on the entire menu of techniques available during archaeological projects rather than as simply an entree.
Several hundred gradiometer surveys are undertaken each year in Great Britain by a number of groups, both independent and publicly funded.
Geophysical Surveys of Bradford has completed over one hundred of these surveys per year for a number of years, and until now, no overall assessment of their 'success' has been undertaken.
The author has, since 1992, been undertaking part time research in an attempt to evaluate the role of gradiometry within archaeological investigations. Not only has the technical success of gradiometry, in terms of whether results were subsequently confirmed by further investigations, been studied, but also the success in terms of the different clients' expectations of the survey.
The research has proceeded by means of both a questionnaire sent out to all clients who commissioned a survey during a given time period, and a detailed analysis of a number of gradiometry surveys where excavation reports are available. The latter included the use of magnetic susceptibility measurements as a tool to investigate reasons for falsely identifying possible archaeological features and missing others.
It was hoped that the study would illuminate the role of gradiometry, and in doing so, establish criteria for defining a successful survey.
Back to Archaeological Prospection 1995