Jökull - 01.01.2010, Síða 124

P. Einarsson
In a more general way, however, it is obvious that
the size of push-ups depends on the magnitude of the
earthquake. The largest push-up formed in the 2000
earthquakes (magnitude 6.5 Mw) was of the order of
tens of centimeters. The push-ups formed in the 1912
and 1630 earthquakes are of the order of 2–4 m high.
The 1912 event had a magnitude of 7.0 (MS).
Sinkholes form where soil covers an open fracture
in the underlying bedrock. They are funnel-shaped de-
pressions, of the order of meters to ten meters across
and tens of centimeters to a few meters deep. The
size of the sinkholes is mainly dependent on the thick-
ness of the soil cover, only mildly dependent on the
width of the underlying fracture. The hole or depres-
sion forms when the soil seeps into the fracture, either
by gravity or helped by water. The water is probably
mostly precipitation seeping through the soil and into
the fracture but fluctuating ground water level in the
fracture may also play a part. The holes may be cir-
cular in shape but frequently they are elliptical with
the major axis aligned with the underlying fracture.
A fracture may be expressed at the surface by a lin-
ear row of sinkholes. Sometimes the sinkholes merge
into a continuous depression above the fracture (Fig-
ure 5). In some of them the vegetation is unruptured,
in others there may be a breach in the vegetation at the
bottom, maintained by flowing or standing water.
Even though the sinkholes are the consequence of
surface fault rupturing they do not necessarily acquire
their present form at the time of fault movement. A
sinkhole will grow in size as long as the underlying
fracture can accommodate material. A good example
was provided during our studies in the district Skeið
where fractures formed in one of the 1896 earthquakes
were seen on older aerial photographs. Farmers had
then filled in the sinkholes because of the inconve-
nience of having them in the fields. Animals would
fall into them and tractors would get stuck in them. So
at the time of our initial study, around 1980, we could
not find them. Twenty years later, when we began our
GPS-guided mapping of fractures in this part of the
zone, all the sinkholes had reappeared and could be
mapped.
THE FAULTS, SPECIFIC EXAMPLES
The mapping project has revealed numerous fracture
arrays, here shown in Figure 1. The arrays have been
grouped into systems that can be taken to represent the
surface expressions of underlying strike-slip faults.
More than 30 such faults may be defined from the
fracture data at hand, arranged side by side within
a zone extending from the Hengill triple junction in
the west to the eastern end where the zone merges
with the Eastern volcanic zone near Hekla. Many
more faults may be defined at depth from the rela-
tive location of microearthquakes, e.g. by Hjaltadótt-
ir et al. (2005a,b), Hjaltadóttir and Vogfjörð (2005),
and Hjaltadóttir (2009). Furthermore, Bjarnason and
Einarsson (1991) found that the 1987 Vatnafjöll earth-
quake (Mw= 5.9) originated on a N-S fault at the east-
ern end of the zone, and Soosalu and Einarsson (1997,
2005) identified two N-S seismic lineaments underly-
ing the Hekla volcano, one of which joins with the
Vatnafjöll fault. The separation between neighbour-
ing faults is in the range 1–5 km.
Several of the surface faults can be paired with
known historical earthquakes (e.g. Einarsson et al.,
1981). Recent revision of research into this was
published by Roth (2004) and Richwalski and Roth
(2008). Most of the fractures are of unknown age, ex-
cept that they are exposed in Holocene surface forma-
tions and were therefore active in the Holocene. Frac-
ture maps of two areas are presented here as exam-
ples of the level of detail in the mapping. The fracture
systems in these maps are not associated with known
historical earthquakes.
The Hestfjall fault
The Hestfjall fault is marked by a relatively continu-
ous, 4 km long chain of fissures, sinkholes and push-
ups exposed in an interglacial lava shield (Figure 6).
The system passes slightly west of the apex crater of
the shield but otherwise it appears to be unrelated to
the structure or the existence of the shield. Several ob-
servations of this system are important for the general
understanding of the structure of the fracture systems.
1) The system has all the structural characteris-
tics of SISZ fracture systems, including sinkholes, en
echelon arrangements, and push-ups. These charac-
124 JÖKULL No. 60