ABINGDON ARCHAEOLOGICAL GEOPHYSICS
ARCHAEOLOGICAL GEOPHYSICS - GENERAL
Why do it?
Because it is:
Non- destructive, hence preserving sites for the future.
Fast and can be fitted in with agricultural operations.
Inexpensive compared with excavating the site.
Discovers previously unknown sites and remains.
Enables excavation to be targeted and minimised. Even if you have air photographs of features, these are difficult to plot on the ground and geophysics can locate these accurately.
Enables developments to be designed to avoid areas of archaeology thus protect the archaeology and avoid the expense of excavating it.
A combination of geophysical and other methods, (such as trial trenching to locate small postholes invisible to geophysics), can often give more information than using only one method.
How accurate is it?
This is discussed below in the description of the main methods and depends on:
The depth, size and type of remains.
The method used
But - If geophysics doesn't find anything it does not necessarily mean that there is nothing there.
What are the main methods and what are their strong and weak points?
This is good for detecting features in deposits up to 1 metre deep although for some sites, such as pottery kilns, it can detect them up to three metres deep. It is good at locating ditches and burnt areas but less good at finding stone buildings unless the stone has significantly different magnetic properties to those of the surrounding soil. It is not good at locating graves as the graves tend to be backfilled with the same soil from which they were dug and do not gradually silt up. Similarly almost all post holes are too small or have the wrong fills to be detected by this method.
It is fairly rapid and approximately 15 30 x 30 metre grids can be done in a day on level ground with no obstructions.
Whilst the soil needs to have a certain amount of iron in it to be enhanced by human activity and detected by the magnetometer, it is less useful where there is a lot of metallic iron in the soil-such as old rubbish tips, or in the vicinity (within say 5 metres) of metal fences such as those containing barbed wire. Our experience in Sweden was that this method did not show ditches which were clear to resistivity, despite the soil having magnetic susceptibility levels which should have made for good magnetometry results. This is still a bit of a mystery - it may well be that the occasional occupation of the area was so ephemeral as to not leave any ditches or other remains which could have survived ploughing, or that the acidic conditions of the coniferous forest has moved the iron within the soil.
Previously flooded areas can have naturally occurring iron minerals which obscure the archaeology and D.G.Weston has shown how waterlogging may cause iron to be reduced from its more magnetic forms to less magnetic ones in Archaeological Prospection 2002 (207-215). Similarly in Archaeological Prospection 2006 (207-227) D Maki, J A Homberg and SD Brosowke explain how hearths which are fired at high temperatures for long periods can become less detectable to magnetometry.
Some geological areas are better than others for magnetometry. In general ironstone areas are good and oolitic limestones less good - although this varies as the soils on these geologies can be be quite good. In general it is probably best to at least do a trial area and see what the results are like. At Bollown on the Isle of Man, for example, Tim Darvill got good magnetometry results on glacial till soil which had been expected to be poorly responsive.
In Egypt it has been found to be able to detect mud brick and, even better, as the sand to clay proportion in the brick varied at different periods so it can give an indication of the date of the remains. We have detected mud brick structures in Iran. K P Games considers that this may be caused by bashing the clay into moulds which may slightly enhance its magnetisation.
It is used for investigating areas where archaeology is expected or where the presence of archaeology would cause significant cost and time problems for a development.
In the past people have used magnetometry scanning to detect features. This involves walking with the magnetometer set so it can give readings or a sound to indicate high readings. Canes are then put in the ground where the high readings are located and then one tries to make some sense of the patterns of canes. Nowadays automatic data logging enables area surveys which can detect smaller and fainter anomalies. This has made scanning virtually obsolete apart from, for example, locating iron working in wooded areas where area surveys are impossible.
This is good for finding ditches and walls although the results can vary greatly depending on whether the soil is too damp or too dry. It is approximately one quarter the speed of magnetometry, if not slower.
As it works by finding the difference in the electrical resistance of the soil and, as this is related to its dampness, it does not require magnetic properties to be present in the soil. It can therefore find features which are invisible to magnetometry. One problem is that results vary according to the dampness of the soil so features may be visible at some times of the year and less visible at other times.
Sample spacing is important as a 1m sample interval should find a 3m wide barrow ditch but is unlikely to locate a wall 25 cm wide and post holes will elude all but unrealistically small spacings.
This method is good for up to about three metres deep although it is at its best in the top metre or so - fortunately most of the archaeology is often in this area. The depth of readings depends on soil conditions and dampness. In general a twin probe array, (which is most commonly used), with a 0.5 metre mobile probe separation will read at 0.5 to 0.75 metres deep and a 1 metre separation will read to 1 to 1.5 metres deep.
A configuration of resistivity apparatus can be used which gives pseudo- sections which can indicate the depth and stratigraphic relationship of features and can read far deeper than the usual area surveys.
Resistivity is usually used where other methods indicate that there may be features worth investigating.
This method sees the amount to which the soil can be temporarily magnetised by exposing it to a magnetic field. Generally speaking, the more the soil can be magnetised the greater the chance that human occupation has taken place in that location.
This is good for shallow surface sites and for rapid evaluation to see where the areas of archaeological potential are. It is best for thin spreads under 10 centimetres deep but can detect sites at greater depths as archaeological material is often brought to the surface by worm action. It is less good for soils with low iron content. It is mainly used to identify areas for further investigation.
One problem is that fly ash from coal fired power stations and similar gives high magnetic susceptibility readings so one needs to be careful to avoid giving an archaeological interpretation to such spreads.
This method is also used to locate shallow features which are invisible to all other methods, such as where Iron Age houses have been. To do this requires having readings at close intervals and this is very time consuming.
Ground Penetrating Radar
This is used in the UK mainly on urban sites were tarmac and other hard surfaces prevent other methods from being used although it is increasingly being used in other areas. It can indicate the depth and stratigraphic relationship of features and can locate walls, ditches etc. The method does not work very well on clay soils in damp conditions as salts affect its usefulness. The data processing is fairly complex to convert the readings to something which is archaeologically useful. On the Continent this method tends to be used in the field where resistivity would be preferred in the UK.
These, such as soil conductivity meters, metal detectors and micro gravity, exist but will not be discussed here. Sonar methods for under water appear to be very good but flooding sites to get better geophysical results may be a bit impractical.
Seeing beneath the Soil, Anthony Clark, Batsford, London 2001
Geophysical Survey in Archaeological Field Evaluation, A. David, English Heritage Research and Professional Guideline no1 1995
Revealing the Buried Past, Chris Gaffney and John Gater, Tempus, Stroud 2004
Archaeological Prospecting and Remote Sensing, I Scollar, Cambridge University Press, Cambridge 1990
Magnetometry for Archaeologists, Arnold Aspinall, Chris Gaffney and Armin Schmidt AltaMira press 2008
Evaluation of Archaeological decision making processes and sampling strategies. G. Hey and M. Lacey available at:- http://www.planarch.org/downloads/library/eval_of_arch_decision-making.pdf