Fault Motion Along the Eastern Margin of the Forlandsundet graben, Sarstangen, Svalbard

Preliminary report on field work conducted by Steffen Bergh*, Alvar Braathen**, Frode Karlson* and Harmon Maher in summer of 98. *-University of Tromsø, **-Norwegian Geological Survey.

Report compiler. Harmon Maher, Dept. of Geography and Geology, University of Nebraska at Omaha, Omaha, NE, 68182-0199, U.S.A.

Orginal posting: 8/17/98. Revised 4/21/99.


The Forlandsundet 'graben' is a fault bounded basin with Tertiary fill within the crystalline hinterland basement of Spitsbergen. Outcrops in the Sarstangen area give insight into the eastern margin of this basin. Two differing Tertiary sequences occur, both dominated by proximal conglomerates, are likely related to two different faults. Slip-linear data and other structural observations are consistent with a history of sinistral, then dextral motions along the east most border fault, followed by orogen-perpendicular extension. Such a history likely records the transpressive to transtensive change in plate motions between northeast Greenland and Svalbard. Locally elevated temperatures permitted thin mylonitic fault rocks to develop. Overall the results continue to emphasize a picture of more complex and varied Tertiary tectonism in the hinterland area, related to decoupling of the oblique plate motions.


A Tertiary basin, the Forlandsundet 'graben', is preserved in the hinterland portion of the Tertiary fold-thrust belt on Spitsbergen, and is known to have a complex history (Kleinsphen and Teyssier, 1992; Gabrielsen and others,1992). Given a probable Eocene to Oligocene age for the fill (Gabrielsen and others, 1992), the basin formed partly coeval with contractional deformation to the east (Maher et al. 1995) indicating a complex regional strain partitioning (Braathen et al., in review). This regional deformation is related to the opening of the Norwegian-Greenland sea and the dextral motion between Greenland and Svalbard (Harland, 1969).

The western border fault is exposed on Prins Karl Foreland, and consists of graben-parallel normal faults stepping down into the interior of the basin (to the east, Kleinsphen and Teyssier, 1992 and Lepvrier, 1992). Adjacent Tertiary fanglomerates locally rest unconformably on basement rocks. The syn-tectonic Tertiary fill is deformed by both extensional and oblique contractional features (Gabrielsen and others, 1992).

The eastern fault margin of the Forlandsundet graben is intermittently exposed at the base of Sarstangen and Kaffiøyra (Fig. 1). This boundary lies along the northern strike continuation of the SEDL, a major fault zone lineament of Carboniferous rocks intensely deformed by a complex history of Tertiary motions. These include an early orogen sub-perpendicular contractional phase, a major sinistral, then dextral orogen-parallel phase (Maher and others, 1995), and a late and minor extensional phase (Guddingsmo, 1996). The more complex history and orogen parallel motions are consistent with decoupling of the dextral transpressive plate motions into a contractional foreland and a transpressional hinterland (Maher and others, 1995).

The purpose of this report is to discuss slip linear fault data and stratigraphic/sedimentologic observations from two well exposed transects on Sarstangen from the basment rocks and into the Tertiary fill obtained during field work. These data further inform about the polyphase history and complex regional strain pattern of this hinterland basin.

Sarstangen Tertiary stratigraphy

An important realization is that there are two Tertiary sequences of very different character and likely different age exposed on Sarstangen. An eastern sequence exposed mainly along the interior moraine margin is relatively well lithified, consists of brown-orange polymict conglomerates and sandstones with minor thin shales (Fig. 2). Coaly fragments are relatively common, lying in concentrations along bedding planes. There is some decrease in overall grain size from E to W, suggesting an eastern source. These rocks also have a host of brittle and ductile faults indicative of a polyphase deformational history (discussed later). Expanded clasts are intruded by the sand matrix, and a brittle, but healed fracture set with an E-W orientation cuts through pebbles commonly with minor offsets (Fig. 3). A cross section of Gabrielsen and others (1992) would suggest that this is the Sarsbukta Formation, and Manum and Throndsen (1986) indicate that dinoflaggelates indicate these "sediments are no younger than the Upper Eocene", but indicate they may be older.

A western sequence is dominated by polymict diamictites to conglomerates that are more poorly sorted and of overall coarser grain size than the eastern sequence and have a dark greenish cast. It is mainly exposed along the southern shore of Sarstangen. Wojcik (1981) indicates these are glacial deposits. The consistent dip (up to 30°), the degree of lithification, the stratigraphic thickness (several hundred meters), and the sedimentology suggest otherwise to us. Again, a simple reading of Fig. 2 of Gabrielsen and others (1992) suggests this more basinward unit is the Sarstangen Fm.. Phyllite and marble clasts are common (consistent with a local basement source), and larger clasts are meters across. Both the clast type (up to 50% phyllite), clast shape ('fragile' elongate pieces with kinks) and size are indicative of a proximal source and a debris fan setting . The western Sarstangen sequence, though more towards the basin center, appears more proximal than the eastern sequence. As outcrops were traced basinward and with stratigraphic ascent, they become finer grained and better sorted, and include good lithic arenites.

The contact between the two units can be narrowed down to a distance of several hundred meters. There appears to be no transitioning of the units; i.e. they retain their distinctive characters even though in such proximity. Also, Krasilscikov and others (1995) describe a strong magnetic anomaly that is an appropriate position to demarcate a fault contact between the two (Fig. 1 b).

The difference in coloration, lithification, grain size and proximity to source, and deformational history is consistent with two units of different age. Importantly large clasts of the eastern sequence were seen in the western, clearly indicating the western sequence is younger. A clear relative age relationship is thus demonstrated. While tilted and gently folded at a variety of angles these strata appear unfaulted, even though shore exposures provided the opportunity for such to be found. Some sutured pressure solution contacts between clasts was evident, but the semi-penetrative fracture set seen in the older sequence was also absent. The greater degree of lithification and structural complexity of the eastern complex is consistent with it being the older sequence. Note that this is the opposite age relationship to that of Gabrielsen and others (1992). Considering the geologic setting and plate history and limited paleontologic data the older unit may very well be Eocene and the younger Oligocene. Based on Sr isotope signatures Eidvin & others (1998) assign an Oligocene (27.5-29.5 Ma) age to sampled Forlandsundet strata.

Structural features

Major bounding faults are not well exposed but their strikes can be inferred from the sparse outcrop data and from geophysical data (Ohta and others, 1995). Exposures constraining the basement-Tertiary contact at Dahltoppen and at Kapp Graarud can be connected by a distinct and strong linear anomalies adjacent to an area of low values and anomaly relief. The former characterize the basement and the later the Tertiary sediments. This is taken to be the outer major basin bounding fault (Fig. 1).

A second strong linear magnetic anomaly trend (Krasilscikov and others, 1995) is an appropriate position to separate the western and eastern stratigraphic sequences described above (Fig. 1). Considering the differences in structural character of the two sequences described above it likely represents a later fault, and may be the fault scarp source for the western sequence.

An overall persistent westward dip exists that general decreases basinward (Fig. 2), but is as high as 70 degrees next to the margin (Fig. 4). In the Kapp Graarud section a map scale fold pair exists (Fig. 1), with a fault near the crest. The axis is subhorizontal and trends 330 degrees, the same general trend as exists in the fold-thrust belt to the east. The general basinward tilt and very gentle folding affects both Tertiary units and thus can be considered a later event. The steeper marginal dips have been attributed to normal fault drag (Gabrielsen and others, 1992), but the folding and continued basinal dip distal from the fault are likely best explained in other ways. Basinward tilts can also result from differential compaction over a complex stepped basin. Seismic sections suggest that there is an overall synclinal sag for the Tertiary basin fill in the Sarsbukta area and that over a kilometer of Tertiary strata underlie surface exposures on Sarstangen (Gabrielsen and others, 1992). This would be similar to geometries seen on sections from the Loppa High. The folding is consistent with observations of contractional structures in Tertiary strata from the west side of the graben on Prins Karls Foreland (Gabrielsen and others, 1992; Kleinsphen and Teyssier, 1992). In elucidating the earlier kinematics this later rotation needs to be considered, and as Lepvrier (1992) indicates some fault populations appear to have been rotated, suggesting an earlier age (discussed below).

A strong and intense fracture set cuts through the Tertiary conglomerates proximal to the eastern border fault. In a number of instances sandy matrix material intruded the fractures indicating they were tensile, and that they were likely penecontemporaneous with deposition (Fig. 3). A large clast framework must have permitted the transmittal of stresses and resulting brittle failure, while a water-rich (with high pore-pressures) sandy matrix was as yet unlithified and intruded the fractures. These fractures were consistently due east-west in orientation and subvertical, suggesting at least local N-S extension. Given a northwest trending border fault and regional tectonic fabric such a strain could be produced by a sinistral shear couple.

Outcrop-scale fault populations

Due to difficulties in meeting basic assumptions (Unruh and Twiss, 1998) we are not using the fault inversion technique for deducing paleostresses, but the slip-linear technique that helps to constrain strain patterns associated with distributed cataclastic flow. This continues an approach used before (e.g. Braathen, 1995).

Three different plots from three different areas show consist fault patterns at each locality, but quite a difference between localities. A shore section from Kapp Graarud to the SW starts in basement rocks and then includes gently folded Tertiary strata. The fault slip linear plot documents a conjugate strike-slip system with a N60E acute bisector and inferred shortening direction (Fig. 5). This is oblique to the more or less N-S oriented border fault in a fashion consistent with dextral motion. It is also sub-parallel to the major contractional axis in the fold-thrust belt (Braathen et al., in review). A substantial number of predominantly sinistral strike-slip faults do not fit this pattern, in that they have distinctly more northwesterly strikes. Similar faults are seen in the plot of faults in basement rocks proximal to the major bounding fault at Dahltoppen. These faults likely represent a separate population from the conjugate set.

At Dahltoppen (Fig. 5) slip linear data from Tertiary strata show a very well developed and dominant conjugate normal fault set. This set was also found by Ohta and others (1995). The extension direction is roughly the same as the obtuse bisector of the conjugate strike-slip set described above. Again, the inferred strain is consistent with dextral motion along the major bounding fault. Minor populations of normal faults consistent with E-W extension exist, and these were measured in proximity to the fault bounding the western sequence.

Slip linear data from basement rocks at Dahltoppen show a more complicated pattern but have a predominance of sinistral faults (Fig.5 b). This is in distinct contrast to the results of Ohta and others (1995). These faults showed a increase in intensity of occurrence with approach to the border fault and an orange staining that is suggestive of a later, and likely Tertiary age. Sinistral motion was documented along the SEDL along strike to the SE (Maher and others, 1997). A lack of these sinistral faults in the adjacent Tertiary strata is most easily explained by motion preceding deposition of these strata. In that several slip surfaces showed multiple slip directions a polyphase history is supported (Ohta and others, 1995).

The distribution of the Tertiary strata is in keeping with the two sequences being related to two different faults - the older sequence with a more eastern fault. Much less fault slip-linear data exists for the more western fault, but suggests it is a normal fault. Individual fault planes with both dip and strike-slip striae were observed, suggesting a 2-stage faulting history.

Thermal history

Thin mylonites and several cm wide fault zones with good composite surfaces, evidence of pressure solution, and mineralization were observed associated with the conjugate normal faults (Fig. 6). Samples were taken and thin sections are being made. Their character suggests conditions of deformation may have involved considerable overburden and/or elevated temperatures. Evidence for local conditions producing microscopic mica growth were noted for Tertiary rocks on the western side of Forlandsundet as were elevated vitrinite reflectance values (Kleinsphen, 1991). However, samples from the Dahltoppen area and from the eastern margin outcrops exhibited much lower values. This suggests these rocks were not in thermal equilibrium and that hydrothermal fluid flow along the faults permitted the more 'ductile' fault rocks to develop. Differing coloration and lithification was observed to be spatially associated with the faults and is in keeping with such an hypothesis.

Kinematic history and conclusions

The following is a model for the kinematic history based on extant literature and our observations. The penecontemporaneous extensional fracture set and the sinistral fault population in the basement rocks would be consistent with pre-Eocene to Eocene sinistral motions along major bounding faults. Sinistral motions have been documented from the SEDL along strike to the southeast (Maher and others, 1997).

A tilted conjugate strike-slip fault set is consistent with contraction with a 60-240° degree axis. Subsequent tilting and folding is consistent with a similar strain field. A post-tilting normal fault set suggests 120-300 degree extension. The rocks were lithified and at elevated temperatures during the formation of the normal fault set. The approximate symmetry of strain axes between the conjugate strike-slip, the conjugate normal, and the folding events, is suggestive of a genetic linkage via a protracted and evolving deformation event. In general it may mark the change from transpression to transtensive plate motions between Svalbard and Greenland that occurred at 38 Ma. Both sets of conjugate faults indicate extension subparallel to the general trend of the Forlandsundet basin border faults or oblique to the anomalous 350° trending border fault at Sarstangen. This pattern is most easily explained by a component of dextral motion along the eastern margin of the basin. The preponderance of conjugate normal faults may be a consequence of the anomalous more northerly orientation of the eastern border fault, which would produce a releasing bend during dextral motions.

Normal faults associated with 70-250 degree directed extension occur (Lepvrier, 1992). Given that they occur in the basin interior and cut the younger sequence, we would suggest they are Oligocene or younger in age, consistent with the 27.5-29.5 Ma assignment based on Sr isotope compositions for some Forlandsundet Tertiary deposits (Eidvin & others, 1998). Considering the concurrent plate motions, these would represent a decoupled extensional component of overall later transtension. Tilting, perhaps due to differential compaction likely occurred at different stages.

This history differs from that of Ohta and others (1995) in having an early sinistral component, and a late extension phase, but is otherwise in broad agreement, although different approaches were used in determining the relative ages. The differing character of the eastern and western border faults of this Tertiary basin may be due to either diachronous development, or to decoupling in an overall transtensive setting similar to that for the Dead Sea (Ben-Avraham and Zoback, 1995). The clear parallelism between the Tertiary border fault and basement fabrics and faults (Ohta and others, 1995) clearly indicates that weak surfaces in the basement played a critical role in the development of the eastern border fault. The general regional pattern is one of differential responses for different portions, suggesting transpressive/transtensive decoupling. The Forlandsundet 'graben' is indeed not a simple graben with multiple kinematic and fill events (Gabrielsen and others, 1992; Kleinsphen and Teyssier, 1992).


Ben-Avraham, Z. and Zoback, M., Transform-normal extension and asymmetric basins: An alternative to pull-apart models; Geology, 20, p. 423-426, 1992.

Braathen, A., and Bergh, S., 1995, Kinematics of tertiary deformation in the basement-involved fold-thrust complex, western Nordenskiøld Land, Svalbard: tectonic implications based on fault-slip data analysis: Tectonophysics, 249, p. 1-29.

Eidvin, T., Goll, R. M., Grogan, P., Smelror, M, & Ulleberg, K., 1998, The Pleistocene to Middle Eocen stratigraphy and geological evolution of the western Barents Sea Continental Margin well site 7316/5-1 (Bjørnøya West Area); Norsk Geologisk Tiddskrift, vol 78, 99-123.

Gabrielsen, R. H., Oddbjørn, S. K., Haugsbø, H., Midbøe, P. S., Nøttvedt, A., Rasmussen, E., Skott, P. H., 1992, A Structural outline of Forlandsundet Graben, Prins Karls Forland, Svalbard; Norsk Geologisk Tidsskrift, vol. 72, p. 105-120.

Guddingsmo, J., 1996, Strukturgeologisk analyse av tertiaert deformerte karbon/perm-bergarter ved Svartfjells, nordvestlige Oscar II Land, Spitsbergen; Candidatus scientarium, Universitet i Tromsø, 148p.

Kleinspehn, K. L. & Teyssier, C., 1992, Tectonics of the Palaeogene Forlandsundet Basin, Spitsbergen: a preliminary report; Norsk Geologisk Tidsskrift, vol. 72, p. 93- 104.

Leprevier, C., 1992, Early Tertiary palaeostress distribution on Spitsbergen: implications for the tectonic development of the western fold-and-thrust belt; Norsk Geologisk Tidsskrift, vol. 72, p. 129-135.

Krasilscikov and others, 1995, Surface magnetic anomaly study on the eastern part of the Forlandsundet Graben; Polar Research, 14, 55-68.

Maher, H. D., Jr., Bergh, S., Braathen, A., & Ohta, Y., 1997, Svartfjella, Eidembukta, and Daudmannsodden Lineament - Tertiary orogen-parallel motion in the crystalline hinterland of Spitsbergen's fold-thrust belt; Tectonics, v. 16, p. 88-106.

Manum, S. B. & Throndsen, T., 1986: Age of Tertiary formations on Spitsbergen. Polar Research., 4, 103-131.

Ohta, Y. Krasilscikov, A., Lepfrier, C., and Tebenkov, A. M., 1995, Northern continuation of Caledonian high-pressure metamorphic rocks in central-western Spitsbergen; Polar Research, 14, p. 303-315.

Wokcik, C., 1981, Geological observations in the eastern part of the Forlandsundet graben between Dahlbreen and Engelsbukta, Spitsbergen. Studia Geologica Polonica, 73, 25-35.

Twiss, R. J. and Unruh, J. R., 1998, Analysis of fault slip inversion: Do they constrain stress or strain rate?: JGR, 103, 12,205-12,222.

Return to H. Maher home page.