Jökull - 01.01.2015, Page 66

Sigurðardóttir et al.
Figure 5. An example of the application of the Peters
half slope method for an anomaly in profile 4 (see
Figure 4 for location). Maximum depth is estimated
as the distance along the profile between the points
where the slope is half the maximum slope divided
by 1.6 (mean proportionality constant). In this exam-
ple the half slope distance is 14 m, giving a maximum
depth of 7 m when sensor height (2 m) has been sub-
tracted. – Peters reglu beitt á segulfrávik í sniði 4.
Fundin er fjarlægð eftir sniði milli punkta þar sem
hallatala sviðstyrksins er helmingur mesta halla deilt
með 1.6 (hlutfallsfasti). Hér er fjarlægðin 14 m sem
gefur mesta mögulega dýpi niður á upptök um 7 m
þegar hæð nemans (2 m) hefur verið dregin frá.
RESULTS
All profiles could be separated into 3 main groups ac-
cording to their magnetic interpretation. First, profiles
where a lava edge was detected, profiles where a lava
edge was not detected and profiles where a clear seg-
mentation cannot be identified. This last group indi-
cates that more than one lava flow may be buried be-
neath sediments in the north-eastern part of the survey
area but they cannot be identified due to a lack of con-
nection with known lavas. For further information of
the magnetic interpretation see Sigurðardóttir (2014).
Lava edge detected
A buried lava edge beneath Mýrdalssandur was de-
tected on profiles 4, 5, 8 and 11 (Figure 6), expressed
by significant anomalies with amplitudes of a few
hundred nT to the east of the assumed lava edge while
they are largely absent on the western side. The lava
edge is thought to be that of the Eldgjá lava flow.
Depth estimations based on the Peters method show
a significant depth decrease from a source at ∼100 m
depth to a shallow source at ∼10 m depth. The mag-
netic source becomes shallower to the east and north-
east towards the visible lava flow on the surface. Lava
ledges are also observed where the depth decreases
significantly as the amplitude of anomalies increases
and anomaly spatial frequency decreases to the east
of the buried lava edge. The easternmost parts of the
profiles are measuring the lava flow on the surface.
Profiles 4, 5 and 8 go from Hafursey and the edge of
Kötlujökull outlet glacier across Mýrdalssandur (see
Figure 4 for location) and over the visible lava flow
on the surface in the east. The western part of these
profiles have smooth and wide anomalies indicating
absence of significant sources of magnetic anomalies
at shallow depth. According to the Peters method
the anomalies observed in the westernmost part of the
study area may originate from a magnetic source at
more than 100 m depth. This could be a result of sed-
iment thickness close to 100 meters. An alternative
explanation would be that the depth to the bedrock
is significantly less but that it is exclusively made up
of hyaloclastite tuff which is known to have very low
remanent magnetization (Kristjánsson and Jónsson,
2007; Gudmundsson and Milsom, 1997). A combi-
nation of both is also possible.
No lava edge detected
Apparently, several of the profiles do not reach long
enough to the west or southwest to detect the true
edge of the Eldgjá lava. These are profiles 1, 2, 3,
6, 7, 9, 10 and 12 (Figure 6). They indicate relatively
shallow magnetic sources (2–10 m) and are therefore
considered to be underlain for their whole length by
the lava flow. However, several of these profiles (1,
2, 3, 6, 7, 10 and 12, see Figure 6) show evidence
of a buried lava ledge where the lava flow is thinning
towards the west, marked by an increase in anomaly
amplitude and decrease in spatial frequency.
66 JÖKULL No. 65, 2015