Fennoscandian Earthquakes: Whole Crustal Rupturing Related to Postglacial Rebound

Abstract

Local and regional earthquake locations provide seismic evidence that large shield earthquakes have occurred in northern Fennoscandia. These paleoearthquakes, with fault lengths of up to 160 kilometers and average displacements of up to 15 meters, were triggered by nonisostatic compressive stresses caused by the removal of the ice at the end of the last deglaciation. The Fennoscandian faults were probably formed by single events that ruptured through most of the crust. The largest event, moment magnitude M, -8.2, was larger than other known stable continent earthquakes outside failed rifts or extended crust.Large-scale surface lineaments (up to 160 km) in northern Fennoscandia (1) have been interpreted as postglacial fault scarps (PGFs) formed by thrust earthquakes shortly after the local deglaciation 9000 years aeo. The formation of these faults mav have " been due to stresses caused hy early postglacia1 rebound. and if so. the faults can be used to estilnate the rate of early postglacial rebound and to infer lithospheric dynamics (2). Offsets of Quaternary sediments and the clustering of paleolandslides and liquefaction phenomena (3) along the scarps indicate that these faults mark Quaternary earthquakes. Such evidence would not be observable if the PGFs formed before deglaciation (3) and restricts the oldest age for the displacements on the PGFs to the time of local deelaciation. 8500 to 9000 vears " ago. The younger age limit is set by undis---turbed postglacial sediments, including littoral sediments, that overlie parts of the Lansjarv PGF, glacier meltwater channels that cut through the Parvie scarp, and 14C dates of peat in a landslide close to Lansjarv of 8200 ? 200 years ago (3). The shallow littoral sediments have depths of (10 m and were formed within 200 years after deglaciation, after which, on the basis of the rate of landrise ( 4 ) , they were raised above sea level. In view of the length of the faults in the stable shield of northern Fennoscandia, the causative earthquakes would have had magnit~~des of -8. Intraplate earthquakes of such magnitude have elsewhere only been located in failed rifts or extended crustal regions (5). I analyzed seismic data to determine whether the observed lineaments are due to earthquakes, and to estilnate the likely magnitudes, fault dimensions, and numher of events that produced the PGFs. The existence of an active fault zone is

A mass movement origin for cirques

“…The possibility that the retreat of ice from Scandinavia and elsewhere after the last glaciation gave rise to severe earthquakes, up to M w = 8·0, due to differential crustal stresses resulting from substantial ice unloading and sea-level rise (e.g. Stewart et al, 2000;Arvidsson, 1996;Zoback and Grollimund, 2001;Olesen et al, 2004) is being increasingly debated, while Davenport et al (1989) and Firth and Stewart (2000) summarize evidence for post-glacial seismic activity in Scotland that was apparently much more intense than the present-day low level of seismicity. The repeated Pleistocene glaciations and deglaciations seem likely to have caused similar or greater crustal stresses and seismic activity many times during the last two million or so years.…”

A generic approach for simulation of large earthquakes

“…Making the same approximation for all earthquakes modeled here and comparing with corresponding stress drop approximation for the glacially induced earthquake in the Pärvie fault in northern Scandinavia suggests that four of the models displayed are well on the conservative side. The slip/width ratio for the Pärvie earthquake is based on estimates by Arvidsson [5] and Wood [12]. For the Uemachi fault and for the 1999 Chi-Chi earthquake it is based on models by Sekiguchi et al [15] and Ma et al [16], respectively.…”

“…The continuum was assumed to be linearly elastic with generic parameter values; Young's modulus 75 GPa, Poisson's ratio 0.25 and density 2700 kg/m 3 . Postglacial, or end-glacial, stress states are characterized by high horizontal stresses which are due to down-warping and/or tectonic strain accumulated under the stabilizing ice cover, whereas the vertical stress corresponds to the rock overburden [5,12,13]. Here, the major initial horizontal stress was oriented normally to the fault strike and calibrated to give the average fault displacement needed to obtain the intended seismic moment.…”

“…In SKB's safety assessment, canisters sheared 100 mm or more count as failed [1]. Although Sweden presently is an area of low seismicity [2] it is generally agreed that large earthquakes (about M8) have occurred in Sweden in connection with the retreat of the latest ice cover some 10,000 years ago [3][4][5]. While all unambiguous traces of such large postglacial earthquakes appear to be located in northern Scandinavia where the ice cap was thickest and had the longest unbroken duration, it cannot be entirely excluded that some may have occurred also in central and southern Sweden [6].…”

Postglacial Faults in Fennoscandia: Targets for scientific drilling

“…If these PG faults were formed in single seismic events, their formation would have required the occurrence of earthquakes with magnitudes up to M w < 7 -8 (e.g. Muir-Wood 1989;Arvidsson 1996;Bungum & Lindholm 1997;Olesen et al 2004). Hence, their presence indicates a much stronger seismicity than the present-day intraplate seismicity of stable continental regions that are generally characterized by smaller earthquakes (M w # 4) (Ahjos & Uski 1992;Bungum & Lindholm 1997).…”