The concern that, in Norway, magnetic ‘noise’ would drown signals from prehistoric remains has proved to be seldom the case, even though geological and soil conditions here are often significantly less favourable than in large parts of, for example, the UK, Denmark and Germany.

Several years experience of investigating actual and potential prehistoric sites in many parts of Norway from Tromsø and east Finnmark in the north to Stavanger in the south, as well as in the Faeroes, has shown that even though magnetic ‘noise’ from bedrock and superficial deposits is a problem in some places, this generally only applies to a small part of an area being investigated. In fact, enhanced magnetic signals from stones often prove advantageous because accumulations of stones, like cairns, flat graves, foundation walls and post packing, are more readily displayed by a gradiometer survey. This experience stands in marked contrast to claims in some British literature that such features as burial cairns and stone foundation walls are not normally revealed by magnetometry. This is far from true when many of the stones with which they are constructed are more magnetic than the surrounding soil, such as is the case in much of Scandinavia, and for that matter Scotland.
 

• a fluxgate gradiometer (Geoscan FM256) with memory space for 256 000 measurements closely spaced measurements (down to 0.0625 m) can be made
• additional equipment permitting measurements 15 cm from the ground surface (instead of the standard 30 cm), for use where conditions are particularly favourable
• computer-processing hardware and software.

Area surveying

Area surveying is most conveniently and efficiently performed on meadows, pastures and reaped cornfields. However, fairly even, rough grazing and open woodland can also be mapped, but this takes longer and the results are often poorer and more difficult to interpret. Ploughed fields and similarly disturbed ground should not be surveyed before the particles in the soil have had time to recover their natural magnetic alignment (several weeks). The work does not require an assistant, but help saves time and enhances precision on difficult ground, particularly among bushes and trees.

Area surveying is performed manually or ‘automatically’ (using a pre-set walking speed) in grids measuring 20 x 20 m (occasionally 10 x 10 m or 30 x 30 m) that have been laid out beforehand. The grids must be very precisely laid out (± 5 cm and N-S/E-W), and the traverses (normally S-N) are marked with non-magnetic pegs. Many grids can be surveyed in a continuous area, but these must be of uniform size and have identical spacing of traverses and measuring points. 

Manual surveying generally takes place using a line spacing of 0.5 m or 1 m with 0.5 m or 0.25 m between measuring points. A 20 x 20 m grid with traverses spaced 1 m apart and measurements taken at 0.5 m intervals gives 800 measurements (0.25 m intervals give 1600 measurements); both alternatives take 20-25 min. to map. Grids with denser traverse spacing than 1 m (0.5 or 0.25 m) take longer to lay out and twice or four times as long to survey.

‘Automatic’ surveying is less flexible, requires easily-walked, obstacle-free terrain, but is essential when closely spaced measurements are needed to seek small sources of magnetic anomalies (e.g. relatively small postholes). The time required chiefly depends on the traverse spacing, and a 20 x 20 m grid with a spacing of 0.25 m will take nearly 1.5 hrs to survey, irrespective of the measurement spacing. We do not perform ’zig-zag’ surveys, which may save time but are regarded as less precise.

Scanning

Scanning entails more or less unsystematic walking over an area, during which the shifts in the magnetic signals can be registered by the operator, but cannot be preserved in the gradiometer memory for subsequent processing. Scanning is most useful for following up anomalies recorded during ordinary mapping, where the terrain or vegetation do not permit ordinary mapping, or for reconnoitring an area prior to deciding whether it should be mapped.