Subaquatic paleoseismology: records of large Holocene earthquakes in marine and lacustrine sediments

Abstract

A review of the early Holocene paleoseismicity in areas covered by huge ice sheets in the interior of Fennoscandia and North America shows that observational data and numerical modelling are consistent with a Glacial Isostatic Adjustment response to ice removal from thick old lithosphere during the last deglaciation. We primarily relate this climate-earthquake relationship to the combined effects of the state of the lithosphere and upper asthenosphere, mainly temperature, rheology and stress, and the state and trend of the climate. The changing ice load modulates the viscoelastic Earth response and the effective stress in such a way that it may cause crustal earthquake ruptures and earthquakes by positive augmentation of pre-existing sources of stress, potentially reactivating old structural features, i.e., inheritance and rejuvenation. The review also introduces issues related to modelling methods and assumptions, such as using realistic ice loads in space and time and appropriate lateral and vertical models of structure, composition and physical properties of the crust and upper mantle. With such caveats in mind, the difference in postglacial tectonic response in North America and Fennoscandia is yet not satisfactorily understood. However, differences in stress augmentation, a greater and longer lasting ice load over a lower relief bedrock, and potential release of stress at greater depth may be specific to North America. Also, while Archean crust characterizes most of the Canadian Shield, almost all of the Fennoscandian paleoseismicity is found in Proterozoic terrains, between the Caledonides to the west and the Archean shield to the northeast. Moreover, the plate boundary configuration yields a variable intraplate ridge-push trend not yet properly accounted for in modelling. In Fennoscandia, there is a clear spatial relationship between rebound, earthquakes and old structures in the Lapland Fault Province and beyond. While the main seismic activity earlier was assumed to have occurred during ~11.5–8.2 ka, there is now increasing evidence also for mid- and late-Holocene ruptures and for fault segmentation instead of previously assumed whole-scarp ruptures, which yielded magnitude estimates up to M ~ 8.2. However, the rupture dates and magnitudes of most postglacial events in Fennoscandia are still uncertain, and even more so in North America.