Content uploaded by Amy Clifton
Author content
All content in this area was uploaded by Amy Clifton on Jan 21, 2024
Content may be subject to copyright.
Structural architecture of a highly oblique divergent
plate boundary segment
Amy E. Clifton
a,
⁎, Simon A. Kattenhorn
b
a
Nordic Volcanological Center, Institute of Earth Sciences, University of Iceland, Reykjavík, Iceland
b
Department of Geological Sciences, University of Idaho, P.O. Box 443022, Moscow, Idaho 83844-3022, USA
Received 28 July 2005; received in revised form 3 March 2006; accepted 20 March 2006
Available online 6 May 2006
Abstract
The Reykjanes Peninsula in southwest Iceland is a highly oblique spreading segment of the Mid-Atlantic Ridge oriented about
30° from the direction of absolute plate motion. We present a complete and spatially accurate map of fractures for the Reykjanes
Peninsula with a level of detail previously unattained. Our map reveals a variability in the pattern of normal, oblique- and strike-
slip faults and open fractures which reflects both temporal and spatial strain partitioning within the plate boundary zone. Fracture
density varies across the length and width of the peninsula, with density maxima at the ends and at the northern margin of the zone
of volcanic activity. Fractures with similar strike cluster into distinct structural domains which can be related to patterns of faulting
predicted for oblique extension and to their spatial distribution with respect to volcanic fissure swarms. Additional structural
complexity on the Reykjanes Peninsula can be reconciled with magmatic periodicity and associated temporal strain partitioning
implied by GPS data, as well as locally perturbed stress fields. Individual faults show variable slip histories, indicating that they
may be active during both magmatic and amagmatic periods associated with different strain fields.
© 2006 Elsevier B.V. All rights reserved.
Keywords: Iceland; Oblique spreading; Strain partitioning; Tectono-magmatic cycle
1. Introduction
Although the quality of bathymetric data is contin-
ually improving, a dearth of high-resolution fracture
data from mid-ocean ridges (MORs) is still an obstacle
to unraveling the complex interplay between faulting
and magmatism in the ridge environment. Exposure of
the neo-volcanic zones of Iceland above sea level
provides a unique opportunity to study mid-ocean ridge
systems. The Mid-Atlantic Ridge (MAR) bends its way
across Iceland in response to the presence of a hot spot
currently situated under the Vatnajökull glacier (Fig. 1).
According to the classification of Macdonald (1998),
the 50–100 km long neo-volcanic zones in Iceland are
second-order ridge segments separated from each other
by either segment overlap or by a change in strike
direction. The Reykjanes Peninsula (RP) ridge segment
is distinguished from the Reykjanes Ridge (RR) at its
western end by a change in strike direction. Its eastern
end is a triple junction, where it connects with the
Western Volcanic Zone (WVZ) to the north via a change
in strike direction, and with the South Iceland Seismic
Zone (SISZ) transform to the east. As the RR segment of
the MAR comes onshore at the Reykjanes peninsula, it
Tectonophysics 419 (2006) 27–40
www.elsevier.com/locate/tecto
⁎Corresponding author.
E-mail address: amy@hi.is (A.E. Clifton).
0040-1951/$ - see front matter © 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.tecto.2006.03.016
bends gradually to the east, until it ultimately becomes
highly oblique (approximately 30°) to the NUVEL-1A
(DeMets et al., 1994) direction of plate motion (Fig. 1).
Consequently, the RP accommodates both left-lateral
shear and extensional strain.
Over 70% of all spreading centers in the world's mid-
ocean ridge systems are oriented obliquely to the
direction of absolute plate motion (Woodcock, 1986).
Oblique spreading centers have been simply character-
ized geometrically in terms of the acute angle, α, be-
tween the plate boundary and the direction of absolute
plate motion. A general pattern of fracture orientations
can be hypothesized based on both analytical strain-field
solutions (Withjack and Jamison, 1986) and to a better
degree using analog models (e.g., Withjack and
Jamison, 1986; Clifton et al., 2000; Clifton and
Schlische, 2001). In clay models where the angle, α,
was 30° in a left-lateral sense (roughly equivalent to the
Fig. 1. A) Tectonic map of Iceland showing ridge segmentation. Dark grey areas denote neo-volcanic zones and ovals denote volcanic systems
(Einarsson and Sæmundsson, 1987). RR= Reykjanes Ridge, RP = Reykjanes Peninsula, WVZ= Western Volcanic Zone, SISZ=South Iceland
Seismic Zone, EVZ =Eastern Volcanic Zone, NVZ =Northern Volcanic Zone. Arrows show the direction of plate motion according to the NUVEL 1A
model (DeMets et al., 1994). Areas in white indicate glaciers. Va= Vatnajökull glacier. Volcanic systems on the Reykjanes Peninsula (Sæmundsson,
1979) are labeled R= Reykjanes, Kr = Krísuvík, Br =Brennisteinsfjöll, H =Hengill. B) Outlines of volcanic fissure swarms on the Reykjanes
Peninsula (redrawn from Jakobsson et al., 1978). Th = Lake Thingvallavatn.
28 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
RP), three sets of faults formed: rift-perpendicular right-
lateral oblique-slip faults, rift-subparallel left-lateral
oblique-slip faults and normal faults striking approxi-
mately 20° counterclockwise to the rift trend (Fig. 2).
Similarities to this pattern on the RP are described by
Clifton and Schlische (2003) and discrepancies are
ascribed in general to magmatic activity. However,
details of the fracture patterns (tension fractures,
eruptive fissures, normal, oblique-slip, and strike-slip
faults) on the RP revealed in the maps presented here
show additional complexity. Many characteristics of the
fracture pattern on the RP clearly indicate that factors
other than spreading obliquity must come into play to
explain the great range of feature types, feature
orientations, structural domains, and variable motion
histories observed. This structural complexity suggests
temporal heterogeneity in tectono-magmatic activity
across the peninsula and a spatial partitioning of strain
within the active plate boundary zone.
GPS measurements between 1993 and 1998
(Hreinsdóttir et al., 2001) show primarily left-lateral
transcurrent motion parallel to the currently active plate
boundary, at a rate of 16.8±0.9 mm/yr. The expected
8.5± 0.9 mm/yr component of extension is not ob-
served. The authors suggest that the “missing”rifting
component is accommodated along normal faults and
eruptive fissures during so-called “rifting episodes”or
magmatic events. The last such event on-shore was in
the 13th century. A periodicity of 1000 years has been
suggested for rifting events on the RP (H. Jóhannesson,
2005, personal communication). The goal of this paper
is to demonstrate that the structural fabric of the RP is
neither homogeneous nor predictable based on obliquity
alone. We demonstrate spatial variability in the fracture
pattern which reflects strain partitioning within the plate
boundary zone. We also show that structural complexity
can be reconciled with magmatic periodicity on the RP
and associated temporal strain partitioning implied by
Fig. 2. A) Fault trace map of clay model surface from Clifton and Schlische (2003) for α= 30° after 2.5 cm of displacement. Trace map is rotated 15°
clockwise to resemble the RP. Arrows show the direction of the moving wall. Light grey dashed lines mark approximate locations of the edge of the
latex sheet. B) Detail of modeling apparatus surface without clay layer, showing that αis defined as the acute angle between the rift axis and the
displacement direction (Withjack and Jamison, 1986).
29A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
GPS data. Individual faults show variable slip histories,
indicating that they may be active in both magmatic and
amagmatic periods associated with different strain
fields. We present a complete and spatially accurate
map of fractures for the entire RP with a level of detail
previously unattained. The map illustrates the kinematic
history of faulting in SW Iceland during the post-glacial
period (since about 12 ka) and allows a better
understanding of the role faulting plays in the plate
boundary dynamics of an oblique, slow-spreading ridge
segment.
2. Reykjanes peninsula background
The RP has been an active spreading center since 6–
7Ma(Sæmundsson, 1979). The oldest rocks at the
surface are Tertiary lavas, Pleistocene hyaloclastites
(formed in sub-glacial eruptions) and interglacial
basaltic lava flows. In the axial rift zone these are
mostly covered by basaltic lavas of Holocene age
(∼12 ka–1240 AD), the products of early post-glacial
shield eruptions and episodic fissure eruptions. Eruptive
fissures have been grouped into distinct swarms
arranged in a right-stepping en echelon pattern and are
spaced approximately 5 km apart, with an average strike
of N40°E (Fig. 1B). These have been described in the
literature as comprising either five “volcanic fissure
swarms”(Jakobsson et al., 1978) or four “volcanic
systems”(Sæmundsson, 1979), each with their own
magma supply, a center of maximum volcanic produc-
tion, and a high temperature geothermal system. We
adopt the nomenclature of Jakobsson et al. (1978)
shown in Fig. 1B. Clusters of fractures comprised of
shear fractures (normal and oblique-slip faults) and
tension fractures (mainly gaping fissures with no shear
displacement) are closely associated with the volcanic
fissure swarms. The overall trend of the en echelon
pattern of fissure swarms has a strike range between
060° and 070° as it steps gradually northeastward from
where the RR comes on shore to where the RP connects
with the WVZ. A significant number of strike-slip faults
are also present (Sigurdsson, 1985; Erlendsson and
Einarsson, 1996; Eyolfsson, 1998), but their relationship
to the fissure swarms is unclear, as they lie in a zone
trending nearly east–west across the center of the
peninsula.
3. Methods
The current work expands greatly on previous
mapping by Jónsson (1978; scale 1:25,000) and the
most recent published geologic map of the peninsula
(Saemundsson and Einarsson, 1980: scale 1:250,000) as
well as proprietary data made available by the energy
companies Hitaveita Suðurnesja and Orkuveita
Reykjavíkur.
The fault map was made using digital 3-band (real
color), orthorectified aerial photographs and a digital
elevation model (DEM) from the Icelandic company,
Loftmyndir hf. Resolution of the photos is 0.5 m per
pixel. The DEM has a grid spacing of 20 m and vertical
accuracy of 5 m. Mapping of over 6700 fractures was
achieved by tracing vector line coverages overlain on
the color images using GIS software (Erdas Imagine
V.8.6 and ArcGIS 9.0). In order to facilitate interpreta-
tion, photos were also draped over the DEM and viewed
using the 3-d viewing module of the Imagine software.
Attached to each line is a set of attributes, including line-
length and strike calculated automatically within the
GIS, which were used to identify distinct structural
domains. Real-time differential GPS data was collected
along a number of faults in order to ground-truth the
spatial accuracy of the air photos and for finer detail
mapping.
4. Structural architecture
The map in Fig. 3 shows that fractures are unevenly
distributed across the length and width of the peninsula.
Fracture density is highest at the eastern, western, and
northern parts of the peninsula and lower in the middle.
In some locations the most recent lava flows have
covered portions of fault systems. For example, there is
a clear gap in the distribution of fractures northeast of
the Reykjanes fissure swarm, where fractures are
covered by lava flows from the 13th century. However,
not all gaps in the fracture distribution are necessarily
the result of being covered by young lava flows, and
some of the youngest lava flows are cut by fractures
(Fig. 4).
The rose diagram in Fig. 3 shows that fractures
striking between 041° and 060° are most abundant and
those striking between 061° and 080° are next most
common. However, fractures striking between 021° and
040° are by far the longest (Fig. 5). Color-coding
fractures by strike emphasizes that fractures within a
particular strike-range tend to cluster together into
distinct structural domains (Fig. 6). While there is
considerable overlap of areas, the following general
statements are valid: 1) fractures striking between 000°
and 020° (red) lie along an east–west zone primarily in
the southern half of the peninsula, intersecting volcanic
fissure swarms at an angle 40° counterclockwise to their
trend; they generally increase in length from west to
30 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
Fig. 3. Map of fractures with color symbology according to strike, overlain on an aerial photomosaic of the Reykjanes Peninsula. Th =Lake Thingvallavatn. Inset: Rose diagram of fracture strikes using
the same color symbology.
31A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
east, but are longest in the center of the peninsula; 2)
fractures striking between 021° and 040° (yellow) tend
to cluster within the volcanic fissure swarms and are
parallel to the strike of eruptive fissures; 3) fractures
striking between 041° and 060° (green) tend to cluster at
the outer edges of the volcanic fissure swarms and are
most densely concentrated north of the Reykjanes and
Grindavík swarms; 4) fractures striking between 061°
and 080° (blue) are rare within the volcanic fissure
swarms and are almost exclusively found in the northern
part of the peninsula. The number of fractures striking
N080° (purple) is negligible. In areas where fractures
striking 041–060° and 021–040° overlap, one set is
always longer than the other. Near eruptive fissures,
021–040° fractures are longest, whereas outside the
zone of eruptive activity 041–060° fractures are longer
and 021–040° are generally subordinate.
Field and seismic data confirm that fractures striking
between 000° and 020° are primarily right-lateral strike-
slip faults, and all should be considered active. This was
driven home on June 17, 2000 when three M≥5
earthquakes occurred within seconds of a M6.6 event
over 80 km to the east in the SISZ (see Clifton et al.,
2003; Pagli et al., 2003; Árnadóttir et al., 2004). These
right-lateral strike-slip events occurred on N–NNE-
striking faults, spaced about 10 km apart. Only one of
the faults had been previously mapped (Erlendsson and
Einarsson, 1996). All are along the so-called “active
plate boundary”zone which extends westward from the
SISZ transform zone (Einarsson, 1991). An example of
the surface manifestation of strike-slip faults on the RP
is shown in Fig. 7. The map in Fig. 7A shows (in black)
arrays of northeast striking en echelon fractures in the
area between the Bláfjöll and Hengill volcanic swarms.
Each member of the array is in turn made up of smaller
en echelon fractures. This is typical of the “double en
echelon”arrangement described for strike-slip faults in
the SISZ (e.g., Erlendsson and Einarsson, 1996;
Bergerat et al., 2003; Clifton and Einarsson, 2005).
These short 021–040° striking fractures, which were
mapped from aerial photographs, are the surface
expression of north-striking, right-lateral, strike-slip
Fig. 5. Plot of all fractures in Fig. 3 showing the relationship between
length and fracture strike.
Fig. 4. Map of post-glacial lava flows from fissure eruptions. Historic lava flows (erupted in the last 1100 years) shown in dark grey, older flows in
lighter grey. Fractures are in black.
32 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
Fig. 6. Maps showing each strike color-bin separately, overlain on volcanic fissure swarms from Fig. 1B. Individual eruptive fissures are shown in
black. A) strike=000° to 020°; B) strike = 021° to 040°; C) strike =041° to 060°; D) strike =061° to 080°.
33A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
faults. The aperture of the extensional fractures and the
size of compressional “push-up”structures observed in
the field are consistent with their formation during a
moderate (M5–6) strike-slip earthquake (Einarsson,
2005, personal communication). Superimposed on these
fracture arrays in Fig. 7A are grey lines which represent
the N-striking faults at depth that are assumed to connect
the arrays. Therefore, the map contains both the
underlying strike-slip fault (interpretive) and the tension
fractures that are its surface manifestation (observation).
Fractures in the westernmost part of the peninsula show
a similar pattern (Fig. 7B). In this case, the push-up
structures (light grey areas) are much more prominent
and closely spaced than extensional fractures along
Fig. 7. Maps illustrating how strike-slip faults were mapped in this study. Inset map shows locations. A) Arrays of short northeast-striking open
fractures shown in black. Light grey lines show underlying strike-slip faults (see text for further explanation). Longer black lines are normal and
oblique slip faults. B) Faults mapped from field data collected with Differential GPS. Light grey areas are push-up structures, black lines are open
fractures. Light grey lines show underlying strike-slip faults (see text for further explanation). C) Faults 2, 3, and 4 mapped from seismic data alone.
Fault 1 mapped by combining field, photo and seismic data. Fractures mapped from photos and field data are shown in black; underlying strike-slip
faults are shown in white (see text for further explanation).
34 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
these short strike-slip faults. It is only in hyaloclastite
ridges that N-striking faults may occur as through-going
structures. Three strike-slip faults were mapped from
earthquakes alone (see Fig. 7C). Two of these faults (2
and 4) slipped on June 17, 2000 and the third (fault 3) on
August 23, 2003. Aftershocks from these M∼5 events
clearly delineated N–S striking fault planes in the
subsurface which coincide with structural lineaments in
the local topography. The fault labeled 1 in Fig. 7C was
mapped by combining these methods. Extensional
fractures along the central part of the fault trace were
mapped in the field whereas those at the northern end
were mapped from aerial photographs. The fault trace
was extended further to the south based on the
occurrence of small earthquake swarms. Therefore, the
total length of this fault is an estimate.
5. Discussion
An oblique spreading center can be classified in
terms of the relative obliquity of the plate boundary and
the plate motion direction (e.g., 30° in the case of the RP
ridge segment). Given this angle, the principal stress and
strain directions can be calculated (Withjack and
Jamison, 1986), and used to make general theoretical
predictions about fault and tension fracture orientations.
Alternatively, clay analog models of oblique spreading
capture the actual patterns of fault growth and the effects
of interactions between growing faults (see Fig. 2).
Different angles of obliquity have been shown to be
associated with specific fault patterns (e.g., Withjack
and Jamison, 1986; Clifton et al., 2000), which can be
applied to oblique spreading centers worldwide.
In nature however, oblique spreading results in a
level of complexity that cannot be thoroughly captured
by laboratory or analytical methods. Our detailed
mapping of fracture orientations and spatial distribution
on the RP reveal heterogeneities that suggest the
influence of local controls on fracture patterns which
result in the development of distinct structural domains.
These domains are not only controlled by ridge
obliquity, but also by proximity to the ridge axis,
proximity to volcanic centers, magmatic periodicity and
associated temporal variability in the stress and strain
fields, changes in rift orientation, reactivation of old
structures, and the development of fractures in locally
perturbed stress fields.
At the 10 s of km scale along the RP ridge segment,
the orientation of the volcanic fissure swarms (Fig. 1B)
can be attributed to the long-term effect of oblique
motion at the plate boundary. Volcanic fissure swarms
on the RP have a general orientation about 35°
counterclockwise to the strike of the presumed ridge
axis, approximately perpendicular to the direction of
predicted maximum principal strain for the case of 30°
spreading obliquity. Nonetheless, the strike of fractures
within these fissure swarms is highly variable (Fig. 3)
and thus difficult to attribute to a single, cumulative
strain direction. In fact, when fractures are grouped by
strike (Fig. 6), there is a clear correlation between their
distribution and the locations of both the ridge axis and
the volcanic fissure swarms.
The relationship of fracture locations and orientations
to the volcanic fissure swarms implies that magmatism
is important for fracture pattern development. However,
magmatic episodes on the RP are periodic, with a post-
glacial repeat time on the order of 1000 years. In fact,
MOR spreading centers are in general characterized by
periodic magmatic episodes and variable magma
production rates. Such complexities are not incorporated
into analytical models of long-term principal strain
directions or clay laboratory models. If all fractures
developed only during magmatic periods, it might be
expected that the strike of fractures would be fairly
uniform, and similar to that of eruptive fissures. The
strikes of fractures closest to the volcanic systems
(yellow fractures in Fig. 6B) do tend to be parallel to
eruptive fissures. These fractures are comprised of both
normal faults and non-eruptive tension fractures, and are
likely to represent extensional structures that formed in
the perturbed stress fields above and alongside upwardly
propagating dikes (Rubin, 1992; Kattenhorn, 2003).
On the outer edges of the volcanic fissure swarms,
fracture strikes differ from those within the swarms (green
fractures Fig. 6C)andthuscannotbelinkeddirectlytothe
effects of dike intrusion. These fractures cannot simply be
assumed to represent old features that were formed at the
ridge axis and rafted outwards by plate spreading. They
are oriented clockwise to fractures within the fissure
swarms, whereas long-term counterclockwise rotation of
structures would be expected to occur in a left-lateral
oblique spreading environment. The fractures striking
041° to 060° (green in Fig. 6C) must therefore represent
the effects ofa stress field that is specific to the outer edges
of the region of volcanic activity. The stress field in this
region is likely to be less affected by magmatism along the
ridge axis and controlled predominantly by tectonic
stresses. In fact, fracture strikes approximately 20°
counterclockwise to the ridge axis are closer to the
predicted orientation of normal faults in α=30° clay
models (Clifton et al., 2000), which do not incorporate a
magmatic element (see Fig. 2).
Many of the faults in this same region, where there is
considerable overlap of green and blue fracture
35A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
orientations, are comprised of en echelon fracture
segments at the 10s of meters scale, where they have
broken through the most recent lava flows at the surface
(e.g., blue in Fig. 8A and B). The consistent left-stepping
arrangement of these fractures indicates that the
underlying faults have undergone right-lateral oblique
slip, resulting in surface fractures that are oriented about
20° clockwise to the overall fault trend. Hence, part of
the reason why the blue fracture field in Fig. 6D overlaps
with the green field is due to oblique motion on the green
faults and resultant en echelon fracturing.
Nonetheless, some of the blue fractures represent a
distinct population of normal faults (Fig. 8A–D) and
must have formed through some mechanism other than
oblique slip on underlying faults. We noticed that the
more easterly striking fractures tend to cluster together
forming distinct structural domains in various locations.
For example, in the region of Stóra–Sandvík (Fig. 8D),
the green and blue fault sets are distinctly different in
character. The green faults form narrow (10s of meters
wide) grabens and are likely to be associated with
underlying dike tips, whereas the blue fault forms a
prominent SE-facing scarp that represents the NW
boundary of the ridge axis where it comes onshore. It
has been suggested (Clifton and Schlische, 2003) that
part of the stress field in the region outside of the zone of
volcanic activity may be related to warping of the crust
in the transition between thicker crust away from the
ridge axis and extended crust closer to the ridge axis,
perhaps explaining some of features that strike 061–
080° (blue in Fig. 6D). Supporting evidence for an
abrupt change in crustal properties comes from
magnetic intensity data (Jonsson et al., 1991). The
central part of the RP is characterized by several
elongate positive magnetic anomalies centered on the
volcanic fissure swarms, whereas the northern part of
the peninsula is characterized by a broad magnetic low
indicating a thick sequence of reversely magnetized
material covered by a thin layer of positively magne-
tized younger lavas (see their Figs. 4 and 6). Three-
dimensional surface motion maps (Guðmundsson et al.,
2002) which combine data from GPS (Hreinsdóttir et
al., 2001) and InSAR (Vadon and Sigmundsson, 1997)
suggest the presence of a broad ENE-trending axis of
subsidence connecting the centers of the volcanic fissure
swarms. The 061–080° fractures lie almost exclusively
to the north of, and are parallel to, this axis of
subsidence, supporting the idea that they formed in
response to crustal flexure related to subsidence along
the trend of the volcanic fissure swarms.
The area northeast of the Reykjanes fissure swarm
(Fig. 8A), variously referred to in the literature as the
Vogar graben or the Vogar fissure swarm, has been studied
in great detail (Gudmundsson, 1980, 1987; Clifton and
Schlische, 2003; Grant and Kattenhorn, 2004). Fractures
at Vogar have been broadly described as “sinuous”or
“anastomosing”(Gudmundsson, 1987; Clifton and
Schlische, 2003). This is certainly the case in the western
part of the graben where Plio-Pleistocene rocks are cut by
large normal faults. However, our fracture map indicates
that the sinuosity is actually the result of interaction
between distinct fault sets (blue and green; Fig. 8A). In
many places, the northeast normal faults in this area cross-
cut what appear to be older, easterly striking open
fractures. These fractures have considerably more lichen
and vegetation growing on their inner walls and have
degraded edges, whereas the northeasterly striking faults
have sharp edges, several meters of throw and much less
vegetation growing on their inner walls.
In Fig. 8A, blue faults commonly appear to emanate
away from the tips of green fault segments, and often act
as linkage structures between the segments. Where this
pattern occurs, the blue faults consistently occur on the
eastern sides of the north tips of the green fault segments
and on the western side of the south tips. Such a
geometry is reminiscent of the pattern of tailcracks that
occur at the tips of right-lateral faults, and is consistent
with the right-lateral oblique motion inferred from the
en echelon fracture pattern. Given that the blue fracture
set is apparently older, they cannot be primary linkage
structures. Instead, reactivation of the blue fault set
likely resulted from a period of oblique motion on the
green fault set and resultant perturbation of stresses at
the fault segment tips. A similar but larger scale
perturbation effect occurs at the northeast end of the
Krísuvík fissure swarm (Fig. 8E). A large cluster of
normal faults in this region defines a distinct structural
domain in the green color bin of Fig. 3. These normal
faults are concentrated on the eastern side of the north
tips of a number of right-lateral oblique-slip fault
segments (yellow in Fig. 8E), consistent with the
development of tailcracks associated with right-lateral
motion. Again it is likely that the normal fault set is
older (the scarps are up to 10s of meters high) and has
been reactivated by motion on the oblique-slip faults.
The north-striking (red) fractures in Fig. 6A are
strike-slip faults that form an east–west trending zone
which continues directly into the SISZ at the southern
end of the Hengill fissure swarm. The individual faults
cross-cut the volcanic fissure swarms and are not
dependent on their locations. Fault length generally
increases towards the east, probably in response to a
thickening of the crust in that direction (Weir et al.,
2001). Strike-slip faults are commonly identifiable in
36 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
Fig. 8. Maps showing complex fractureinteractions in three locationson the RP (inset map shows locations). Color symbology same as used in Figs. 3 and
6(see text for further explanation). Fractures in (B) shown in lightershades were mapped from aerial photos, fractures in darkershades mapped in the field.
37A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
the field as N–S arrays of en echelon fractures and push-
ups (Fig. 7), with individual fractures oriented clock-
wise to the fault orientation in response to right-lateral
motion. These short, en echelon fractures explain some
of the linear domains of yellow fractures in Fig. 6B
which fall outside of the volcanic fissure swarms.
Additional structural complexity at the northern and
southern margins of the RP (e.g., Fig. 8E) appears to
occur along the lateral projection of strike-slip faults,
raising the possibility that the strike-slip faults are
longer than their more obvious surface manifestation.
Many of the strike-slip faults have a very subtle
surface manifestation (Fig. 7) and have thus been
mapped using earthquake data. Such mapping has been
greatly facilitated by the fact that most seismic energy
release on the RP currently occurs along N–S strike-slip
faults. Such activity is consistent with the left-lateral
transcurrent motion presently occurring along the plate
boundary according to GPS velocity measurements
(Hreinsdóttir et al., 2001). The faults thus accommodate
left-lateral transcurrent motion via right-lateral “book-
shelf”faulting, analogous to the SISZ further to the east
(Sigmundsson et al., 1995). The current predominance
of strike-slip faulting raises an important question about
when the large number of normal and oblique-slip faults
on the RP are active. Fault throws may exceed 10 m in
b12,000-year-old lava flows, implying long-term slip
averages of ∼1 mm/yr. Repeated precision leveling data
collected between 1966 and 2004 across the Vogar
graben (Tryggvason, 1970; Anell, 2004) shows that at
least 3 mm/yr of throw has occurred on several normal
faults during this time period. Although some of the
background seismicity across the RP can be attributed to
normal and oblique-slip faults, they are clearly not the
dominant active features in the current amagmatic
period. Given this pattern of fault activity, and the
clear spatial association of normal and oblique-slip
faults with the locations of volcanic fissure swarms, as
described above, they are likely to be more active during
magmatic periods. Such behavior was observed during
the last eruptive episode at the Krafla volcano in the
NVZ when several meters of throw occurred along
normal faults during periods of dike injection (Einarsson
and Brandsdóttir, 1980).
The high fracture density in the Reykjanes volcanic
fissure swarm (RFS) relative to the other swarms
suggests that fault development here is somewhat
unique. The RFS, at the western end of the peninsula,
extends offshore and is part of the gradual transition zone
from the RR to the RP proper. This zone is oriented less
oblique to the plate spreading direction than the rest of
the RP. Hreinsdóttir et al. (2001) and Árnadóttir et al.
(2005, personal communication), have modeled the plate
boundary on the RP as two screw dislocations with a
change in strike at 22.42°W longitude (at the Grindavík
swarm) so that west of this point strike is 060° and east of
this point it is 079°. Clay models of oblique spreading
(Clifton et al., 2000) indicate that fault density increases
with decreasing obliquity. This would appear to be the
case for the RFS indicating that there is a spatial
heterogeneity in the distribution of strain across the RP.
Hreinsdóttir et al. (2001) posited that the 8.5 ±
0.9 mm/yr extension component of plate spreading that
cannot be accounted for in GPS velocity measurements
is principally accommodated during magmatic periods.
We concur with this suggestion and advocate that the
fracture patterns and fault behavior described in this
work imply temporal variability in stress field char-
acteristics associated with the tectono-magmatic cycle.
For example, individual faults may show variable slip
histories, such as normal faults reactivated as oblique-
slip faults during breakthrough of surface lava flows
(Fernandes, 2005). These temporal changes in fault
behavior result from the principal strain direction during
the magmatic periods being different to that associated
with transform motion characteristic of the current
amagmatic period. Although different fault sets may be
dominant in different time periods (such as strike-slip
faults in the current amagmatic period), we have
documented that all fault types may be somewhat active
throughout the magmatic cycle, depending on the
relative orientations of the principal strain and the
fracture strike at any point in time.
Theoretical models for infinitesimal displacements
(Withjack and Jamison, 1986) predict a principal
extensional strain direction oriented approximately
half-way between the displacement vector and the
normal to the plate boundary. Therefore, left-lateral
transform displacements along an E–W axis (i.e., the
zone of strike-slip faults across the RP) create an
extensional strain direction oriented approximately 45°
to this zone in a clockwise sense. This strain field is
extremely conducive to right-lateral slip along N–S
strike-slip faults, and would promote normal motion
along NE-striking faults. Such fault behavior is
confirmedbybothα=0° clay model results (Clifton
et al., 2000) and the recent earthquake history on the
RP. In contrast, when the extension component of the
absolute plate motion vector is accommodated during
magmatic periods, the principal extensional strain
orientation is likely to be N60° to the zone of plate
spreading. This strain direction would thus promote
normal to right-lateral oblique-slip along NE to ENE-
striking faults, in agreement with both our field
38 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
observations (Fig. 8A and B and clay model results for
αin the 30–60° range. These fault orientations are
therefore most active during the magmatic periods,
with strike-slip faulting becoming subsidiary.
6. Conclusions
In summary, the pattern offaults and fractures on the RP
ridge segment (Figs. 3 and 6) represents the combined
influences of obliquity angle (which changes slightly near
the western end of the RP), proximity to volcanic centers,
reactivation of older structures in the perturbed stress fields
of differently oriented younger structures, and the point in
time during the waxing and waning of the tectono-
magmatic cycle. These factors have combined to produce
a range of fault orientations in distinct structural domains
that indicate a heterogeneous accommodation of strain
across the RP ridge segment. The most active fault set at
any point in time is variable. Strike-slip faulting dominates
during amagmatic periods (such as currently) whereas
normal faulting is more prominent during magmatic
periods. Nonetheless, normal faults are somewhat active
in amagmatic periods, but tend to slip in an oblique sense
during these times due to the temporally variable stress
field. We conclude that magmatic periodicity on the RP
ultimately results in both spatial and temporal strain
partitioning, with different components of the long-term
spreading direction given by the NUVEL-1A model being
accommodated at different times in response to an
oscillatory tectono-magmatic cycle.
Acknowledgements
The authors would like to acknowledge Freysteinn
Sigmundsson and Thora Árnadóttir for useful discus-
sions. We thank Leslie Fernandes, Kristen McCurdy,
and Francien Peterse for assistance in the field. We
thank Mike Sandiford, Haakon Fossen and an anony-
mous reviewer for their helpful suggestions. Funding for
this project was provided by a grant in support of
research from the Icelandic Center for Research
(RANNÍS), European Community contract EVG1-CT-
2001-0044 (RETINA) and European Community con-
tract EVG1-CT-2002-00073 (PREPARED) (all to Clif-
ton) and NSF grant EAR-0309016 (to Kattenhorn).
References
Anell, I., 2004. Subsidence in rifts zones: analyzing results from
repeated precision leveling of the Vogar profile on the Reykjanes
Peninsula, Southwest Iceland. M.Sc. Thesis Nr. 178, University of
Lund, Lund, Sweden, 54 pp.
Árnadóttir, Th., Geirsson, H., Einarsson, P., 2004. Coseismic stress
changes and crustal deformation on the Reykjanes Peninsula due to
triggered earthquakes on 17 June 2000. Journal of Geophysical
Research 109, B09307. doi:10.1029/2004JB003130.
Bergerat, F., Angelier, J., Gudmundsson, A., Torfason, H., 2003. Push-
ups, fracture patterns, and palaeoseismology of the Leirubakki
Fault, South Iceland. Journal of Structural Geology 25, 591–609.
Clifton, A., Einarsson, P., 2005. Styles of surface rupture accompa-
nying the June 17 and 21, 2000 earthquakes in the South Iceland
Seismic Zone. Tectonophysics 396, 141–159.
Clifton, A.E., Schlische, R.W., 2001. Nucleation, growth and linkage of
faults in oblique rift zones: results from experimental clay models
and implications for maximum fault size. Geology 29, 455–458.
Clifton, A.E., Schlische, R.W., 2003. Fracture populations on the
Reykjanes Peninsula, Iceland: comparison with experimental clay
models of oblique rifting. Journal of Geophysical Research108 (B2).
Clifton, A.E., Schlische, R.W., Withjack, M.O., Ackermann, R.V.,
2000. Influence of rift obliquity on fault-population systematics:
results of experimental clay models. Journal of Structural Geology
22, 1491–1509.
Clifton, A.E., Pagli, C., Jónsdóttir, J.F., Eythorsdóttir, K., Vogfjörð, K.,
2003. Surface effects of triggered fault slip on Reykjanes Peninsula,
SW Iceland. Tectonophysics 369, 145–154.
DeMets, C., Gordon, R., Argus, D., Stein, S., 1994. Effect of recent
revisions to the geomagnetic reversal time scale on estimates of
current plate motions. Geophysical Research Letters 21, 2191–2194.
Einarsson, P., 1991. Earthquakes and present-day tectonism in Iceland.
Tectonophysics 189, 261–279.
Einarsson, P., Brandsdóttir, B., 1980. Seismological evidence for
lateral magma intrusion during the July 1978 deflation of the
Krafla volcano in NE-Iceland. Journal of Geophysics 47, 160–165.
Einarsson, P., Sæmundsson, K., 1987. Earthquake epicenters 1982–
1985 and volcanic systems in Iceland (map). In: Th. Sigfússon (ed.),
Í Hlutarsins Eðli: Festschrift for Thorbjorn Sigurgeirsson, Mennin-
garsjóður, Reykjavík.
Erlendsson, P., Einarsson, P., 1996.The Hvalhnúkur Fault, a strike-slip
fault mapped within the Reykjanes peninsula oblique rift, Iceland.
In: Thorkelsson, B. (Ed.), Seismology in Europe, XXV ESC
General Assembly, Reykjavík Iceland, September 9–14, 1996,
pp. 498–505.
Eyolfsson, V., 1998. Kortlagning sprungna og nútíma eldvarpa í
Fagradalsfjalli á vestanverðum Reykjanesskaga (Mapping of
fractures and Holocene volcanic vents in Fagradalsfjall, Western
Reykjanes Peninsula. In Icelandic). B.Sc. thesis, University of
Iceland, Department of Geology and Geography, 70 pp.
Fernandes, L.C., 2005. The magmatic influence on temporal strain parti-
tioning at an oblique spreading center, southwest Iceland. Unpublished
MS thesis, University of Idaho, Moscow, Idaho, USA, 119p.
Grant, J.V., Kattenhorn, S.A., 2004. Evolution of vertical faults at an
extensional plate boundary, southwest Iceland. Journal of Structural
Geology 26, 537–557.
Gudmundsson, A., 1980. The Vogar fissure swarm, Reykjanes
Peninsula, SW-Iceland. Jökull 30, 43–63.
Gudmundsson, A., 1987. Geometry, formation and development of
tectonic fractures on the Reykjanes Peninsula, Southwest Iceland.
Tectonophysics 139, 295–308.
Guðmundsson, S., Sigmundsson, F., Carstensen, J.M., 2002. Three-
dimensional surface motion maps estimated from combined
InSAR and GPS data. Journal of Geophysical Research 107.
doi:10.1029/2001JB000283.
Hreinsdóttir, S., Einarsson, P., Sigmundsson, F., 2001. Crustal
deformation at the oblique spreading Reykjanes Peninsula, SW
39A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
Iceland: GPS measurements from 1993 to 1998. Journal of
Geophysical Research 106, 13,803–13,816.
Jakobsson, S.P., Jónsson, J., Shido, F., 1978. Petrology of the Western
Reykjanes Peninsula, Iceland. Journal of Petrology 19, 669–705.
Jónsson, J., 1978. Jardfraedikort af Reykjanesskaga (Report on the
Geology of Reykjanes Peninsula. In Icelandic). Orkustofnun
Report OS-JHD-7831, Reykjavik.
Jonsson, G., Kristjansson, L., Sverrisson, M., 1991. Magnetic surveys
of Iceland. Tectonophysics 189, 229–247.
Kattenhorn, S.A., 2003. Upward growth of normal faults in southwest
Iceland: driven by dike magmatism? Proceedings of the RIDGE
2000/Nordvulk Tectonic–Magmatic Interaction Workshop, Gey-
sir, Iceland, Aug. 31–Sept. 9, 2003.
Macdonald, K., 1998. Linkages between faulting, volcanism,
hydrothermal activity and segmentation on fast spreading centers.
In: Buck, W.R., Delaney, P.T., Karson, J.A., Lagabrielle, Y. (Eds.),
Faulting and Magmatism at Mid-Ocean Ridges. Geophysical
Monograph, vol. 106. American Geophysical Union, pp. 27–58.
Pagli, C., Pedersen, R., Sigmundsson, F., Feigl, K.L., 2003. Triggered
fault slip on June 17, 2000 on the Reykjanes Peninsula, SW-
Iceland captured by radar interferometry. Geophysical Research
Letters 30 (6), 1273. doi:10.1029/2002GL015310.
Rubin, A.M., 1992. Dike-induced faulting and graben subsidence in
volcanic rift zones. Journal of Geophysical Research 97,
1839–1858.
Sæmundsson, K., 1979. Outline of the geology of Iceland. Jökull 29,
7–28.
Saemundsson, K. Einarsson, S., 1980. Geological map of Iceland,
sheet 3, SW-Iceland, second edition, scale 1:250,000. Museum of
Natural History and the Iceland Geodetic Survey, Reykjavik.
Sigmundsson, F., Einarsson, P., Bilham, R., Sturkell, E., 1995. Rift-
transform kinematics in south Iceland: deformation from Global
Positioning System measurements, 1986 to 1992. Journal of
Geophysical Research 100, 6235–6248.
Sigurdsson, F., 1985. Jardvatn og vatnajardfraedi á utanverdum
Reykjanesskaga (in Icelandic). Orkustofnun Report OS-85075/
VOD-06, Reykjavik.
Tryggvason, E., 1970. Surface deformation and fault displacement
associated with an earthquake swarm in Iceland. Journal of
Geophysical Research 75, 4407–4421.
Vadon, H., Sigmundsson, F., 1997. Crustal deformation from 1992 to
1995 at the Mid-Atlantic Ridge, southwest Iceland, mapped by
satellite radar interferometry. Science 275, 193–197.
Weir, N.R.W., White, R.S., Brandsdóttir, B., Einarsson, P., Shima-
mura, H., Shiobara, H., RISE fieldwork team, 2001. Crustal
structure of the northern Reykjanes Ridge and Reykjanes
Peninsula, southwest Iceland. Journal of Geophysical Research
106, 6347–6368.
Withjack, M.O., Jamison, W.R., 1986. Deformation produced by
oblique rifting. Tectonophysics 126, 99–124.
Woodcock, N.H., 1986. The role of strike-slip fault systems at plate
boundaries. Philosophical Transactions of the Royal Society of
London 317, 13–29.
40 A.E. Clifton, S.A. Kattenhorn / Tectonophysics 419 (2006) 27–40
