High resolution magnetic surveying

The de facto standard for magnetic surveying is currently at 25cm samples in the walking direction using 1 metre traverses. This gives a reasonable balance between anomaly detection and the time taken to collect the survey data. However, at specific times during the last 16 years (see table below for details) the LRC has had the opportunity to test higher resolution surveying over areas of known archaeological potential.

The first tests were carried out in 1991 (see image below for location) and early in 1992 by Tim Carew, over 4 areas covering a total of 0.465 hectares. All of these surveys were conducted with an FM36 after the removal of the topsoil, but remaining above the archaeological strata, with the features sealed by a thin layer of hillwash and thus generally remaining invisible on the ground. 

In the image above the same area surveyed at a normal resolution (25cm by 100cm) prior to the removal of the topsoil (left, survey by the AML team) can be compared with the high resolution survey (25cm by 25cm) carried out by the LRC after topsoil stripping.  For further details on the 1995 surveys see Lyall, J. & Powlesland, D. J. (1996). 

The technique of using high resolution surveying has been used twice more over different areas of the project zone, once after the removal of topsoil, and once on the current ground surface. In both cases, enhanced detail over the usual survey resolution was present. 

The second test area was in a different geological zone, where deep bands of sand and chalk gravel are overlain in parts by windblown sand. It was located just to the west of the area where Arnold Aspinall had carried out the initial unsuccessful geophysical tests. The effect of more magnetically susceptible ploughsoil masking underlying features is graphically demonstrated below, showing the results of two gradiometer surveys conducted at Site 010, one carried out before the removal of the topsoil, and one shortly afterwards.

The two site 010 surveys are not directly comparable, as the initial survey had a spatial resolution of 25cm north-south and 100 cm east-west, while the second survey had a higher spatial resolution of 25cm in both directions. Nevertheless, the difference in anomaly detection is quite remarkable, especially when considering that the overburden of ploughsoil removed from the western half of the second surveyed area was no more than 30-35cm in depth. During this second high resolution survey, it was noticed that the difference in magnetic contrast between the filled archaeological features and the parent sand and chalk gravel subsoil was slight; no more than 1 or 2 nT.    

The third trial of high resolution surveying was carried out just to the south-east of the earthworks associated with the deserted medieval village, located to the south of the present village of East Heslerton. The original survey results were good, showing a curvilinear enclosure, a potential Grubenhaus and a circular anomaly just to the south-east of the enclosure, but it was felt that further enhancement of detail could be achieved if a closer traverse interval was used. A small test area (30m by 60m) was then surveyed.

The results of the high resolution survey (25cm by 25cm) were generally excellent, defining the enclosure, as well as confirming the ring ditch feature to the south-east. The potential Grubenhaus looks even more convincing in this dataset. Further anomalies not easily defined in the first dataset were obvious in the second. Although the ploughsoil had not been stripped away prior to the high resolution survey, the location of the surveyed area was on the brow of a small chalky knoll, and it is likely that the ploughsoil depth here is extremely shallow. The implication is that in areas where magnetic techniques are known to give good results and thin layers of topsoil are present, the use of close traverse surveys should give an improved definition of anomalies.

The first tests were carried out in 1991 (see image below for location) and early in 1992 by Tim Carew, over 4 areas covering a total of 0.465 hectares. All of these surveys were conducted with an FM36 after the removal of the topsoil, but remaining above the archaeological strata, with the features sealed by a thin layer of hillwash and thus generally remaining invisible on the ground. 

In the image above the same area surveyed at a normal resolution (25cm by 100cm) prior to the removal of the topsoil (left, survey by the AML team) can be compared with the high resolution survey (25cm by 25cm) carried out by the LRC after topsoil stripping.  For further details on the 1995 surveys see Lyall, J. & Powlesland, D. J. (1996). 

The technique of using high resolution surveying has been used twice more over different areas of the project zone, once after the removal of topsoil, and once on the current ground surface. In both cases, enhanced detail over the usual survey resolution was present. 

The second test area was in a different geological zone, where deep bands of sand and chalk gravel are overlain in parts by windblown sand. It was located just to the west of the area where Arnold Aspinall had carried out the initial unsuccessful geophysical tests. The effect of more magnetically susceptible ploughsoil masking underlying features is graphically demonstrated below, showing the results of two gradiometer surveys conducted at Site 010, one carried out before the removal of the topsoil, and one shortly afterwards.

The two site 010 surveys are not directly comparable, as the initial survey had a spatial resolution of 25cm north-south and 100 cm east-west, while the second survey had a higher spatial resolution of 25cm in both directions. Nevertheless, the difference in anomaly detection is quite remarkable, especially when considering that the overburden of ploughsoil removed from the western half of the second surveyed area was no more than 30-35cm in depth. During this second high resolution survey, it was noticed that the difference in magnetic contrast between the filled archaeological features and the parent sand and chalk gravel subsoil was slight; no more than 1 or 2 nT.    

The third trial of high resolution surveying was carried out just to the south-east of the earthworks associated with the deserted medieval village, located to the south of the present village of East Heslerton. The original survey results were good, showing a curvilinear enclosure, a potential Grubenhaus and a circular anomaly just to the south-east of the enclosure, but it was felt that further enhancement of detail could be achieved if a closer traverse interval was used. A small test area (30m by 60m) was then surveyed.

The results of the high resolution survey (25cm by 25cm) were generally excellent, defining the enclosure, as well as confirming the ring ditch feature to the south-east. The potential Grubenhaus looks even more convincing in this dataset. Further anomalies not easily defined in the first dataset were obvious in the second. Although the ploughsoil had not been stripped away prior to the high resolution survey, the location of the surveyed area was on the brow of a small chalky knoll, and it is likely that the ploughsoil depth here is extremely shallow. The implication is that in areas where magnetic techniques are known to give good results and thin layers of topsoil are present, the use of close traverse surveys should give an improved definition of anomalies.

Although the results from the high resolution survey have already been discussed (Lyall, J. & Powlesland, D. J. (1996), there are a number of issues which were not addressed in that paper. One benefit of conducting high resolution surveys is the ability to sub-sample the data down to lower resolutions in order to test the returns from a number of different resolution settings.

The figure above gives the 9 possible combinations of sub-samples which can be obtained from the 25 by 25 cm resolution data, as well as the number of points in a 30 metre square grid for each resolution. The standard survey resolution is shaded. During high resolution surveys, it is even more vital than normal to ensure that the machine is carried absolutely upright during the survey, as even a slight tilt of the machine would negate the attempt to achieve a true 25 cm resolution, and create striping problems in the data. For this reason it is recommended that this form of survey only be carried out by experienced operators. (Even so, note the slight zig-zag error visible where we crossed the ditch in the full resolution 25 by 25cm data).

In order to test the returns from the various resolutions, an area where anomalies of less than 50cm in diameter were detected was chosen (Area 012AE). The location was on a raised chalky knoll, providing excellent magnetic contrast between the chalk and the earth-filled features. One downside was that because of the location, the archaeological features were all plough-truncated to some degree. The results from the nine possible combinations illustrated in above . Note that they also demonstrate the orientation of survey effect, noticeably changing the response of linear anomalies depending on whether they were intersected at right angles or were aligned along the survey orientation.

Because the English Heritage (AML) personnel had previously carried out a normal resolution survey prior to the removal of the ploughsoil, it was also possible to compare the results of a normal resolution (i.e. 1m in the east-west axis and 25cm in the north-south axis) survey, both before and after the topsoil removal.

As can be seen in the figure above , the definition of the anomalies is generally better, with the linear east-west plough marks no longer visible. What proved to be a pit (just to the south-east of the eastern end of the enclosure ditch) is much more clearly defined. Although the linear slot of the post-in-trench structure is just visible in the earlier survey, it would not have been possible to interpret the anomaly as such from the normal resolution survey. There are a number of anomalies visible to the south of the post-in-trench structure, and as such would have been flagged as an area of potential archaeological interest, but they are very slight positive anomalies, and certainly could not have been interpreted as the remains of an Anglian posthole structure on this evidence alone. As a result of these findings, it is concluded that in order to detect and allow some scope for the interpretation of small anomalies, the combination of high resolution surveying with the removal of the ploughsoil is vital. As always, even using high resolution surveying does not detect all of the archaeological features, clearly demonstrated by the excavated plan in the diagram above

Back to Main Magnetometry Index