Archaeologia Islandica - 01.01.2002, Qupperneq 17

Geophysical Techniques
allowed, this grid was surveyed with both
magnetometer and earth resistance tech-
niques.
Earth resistance surveys
Earth resistance surveys were conducted
using a Geoscan RM15 resistance meter
using the twin probe configuration with a
mobile probe separation of either 0.5m or
l.Om. For surveys with a mobile probe
separation of 0.5m, readings of apparent
resistivity were collected at 0.5m inter-
vals along traverses spaced 0.5m apart.
For the grids surveyed with a l.Om
mobile probe separation, readings were
recorded at 0.5m intervals along travers-
es 1 .Om apart. These high resolutions
were chosen so as to gain the maximum
information within time constraints.
Fluxgate gradiometer surveys
The magnetometer employed for this
study was a Geoscan FM36 fluxgate gra-
diometer. Readings were recorded at the
high resolution of 0.25m intervals along
traverses spaced 0.5m apart, to provide a
high level of detail. This was made pos-
sible by the use of a sample trigger con-
nected to the instrument, which automat-
ically records a pre-set number of meas-
urements at regular intervals.
It was decided to collect data within a
grid by walking in the so-called 'zigzag'
fashion, as opposed to 'parallel'. In this
way the direction of travel alternates
between adjacent traverses to maximise
survey speed, allowing a greater area to
be covered. However, the magnetometer
is always kept facing in the same direc-
tion, regardless of the direction of travel,
to minimise heading error (Horsley,
1998). It is acknowledged that collecting
measurements in this way can produce
defects in the data. 'Bunching' within the
data is caused by differences in the han-
dling and orientation of the instrument in
altemate traverses causing a pronounced
striped effect parallel to the traverses
(David 1995, 19).
Data processing
Geophysical data often benefits from the
application of pre-treatment and process-
ing techniques, specifically applied to
remove background noise or to correct
for collection defects. For this purpose a
dedicated computer program is often
used to process a survey data set and aid
in its interpretation. The package
employed for processing and display of
survey data for this assessment was the
dedicated Geoplot software, of which
both Geoplot 2.02 (© 1994 Geoscan
Research) and Geoplot 3.0 (Beta version
GPW3045R © 1999 Geoscan Research)
were used.
The Surveys
Gásir, Eyjafjörður
Gásir, near Akureyri on the west coast of
Eyjafjörður, was an important trading
centre in medieval lceland, dating to
before 1400 according to documentary
evidence (Friðriksson 1994, 95). Visitors
to the site today are presented with a
great number of earthworks - the remains
of trading booths that stretch along a nat-
ural harbour. Another circular earthwork
to the west, at a distance from the booths,
may indicate the site of a church and
churchyard, although the exact nature of
15