At two occations, 1755 and 1872 earthquakes are known to have caused large destruction in the village of Húsavík in North Iceland. They are related to the Húsavík - Flatey Fault (HFF) which passes through the village. Among objectives of our projects, PRENLAB and PRENLAB-2 is to provide answers to the following questions: 1) Is a destructive earthquake to be expected soon at or near Husavík, i.e. during the next years or the next tens of years? 2) How big will it be in terms of surface movements, magnitude, size of fault area, fault slip? 3) What faulting mechanism should be expected in large Húsavík earthquakes? Are they different depending on different strain conditions of the Tjornes Fracture Zone (TFZ) as a whole? Are there more than one active fault planes within the zone? What is the relation between Husavík earthquakes and the strain conditions and the strain or stress changes within the TFZ? 4) Is a short term prediction possible, i.e. hours or days before a large earthquake, making it possible to mitigate risk for people? 5)Is it possible that risk might be mitigated by a minor relocation of the village or of some houses of the village?
Besides hinting at answers to these questions we will point out some research tasks that have to be undertaken to be able to provide better answers. These suggestions are made mainly from the point of view of seismologists.
Data will be presented about historical earthquakes since 1700, their estimated magnitudes, lengths of fault planes and slips, not only in the HFF but also in the TFZ as a whole.
Difference in mechanism as well as in location of the center of energy release in the two large earthquake epizodes will be demonstrated as well as difference in general conditions prevailing in the North Volcanic Zone (NVZ) and the TFZ at these epizodes. It is estimated that the cumulative slip on these zones is at least 3 - 4 meters in these epizodes. Based on geological evidence it has been concluded by some authors that during the past few tens of thousands to a few hundreds of thousands years the HFF has only taken up a small fraction of the total plate divergency in this area, which is 18 mm, i.e. only a mm or a few mm per year. This would mean that thousand years should pass before the next magnitude 7 earthquake would take place in the area.
The interplay between the hot spot activity, general rifting, and variable weight of the icecap causes large pulsations of rifting activity. We propose that it is possible that at present the right lateral motion across the HFF is relatively fast, i.e. much faster than the mean value for the last hundreds of thousands of years. This means that presently the HFF takes up most of the plate divergency, which would mean that repitition time of magnitude 7 earthquakes on the fault would only be 2-3 hundred years. The historical seismicity and the rifting epizodes since 1700 support this. The historical seismicity may indicate that most of the rifting would be taken up by the 130 km long proposed transform fault, from north of Skagafjordur to the western part of NVZ.
If the above assumption about the present day activity of the HFF fault system is justified it should be analysed in a similar way as some of the very large transform fault systems of world, where the motion is intermittend in time and on segments of large earthquakes and segments of aseismic motion, and the stresses and smaller earthquakes around the fault system are controlled by the major slips on the main fault system. The seismicity and rifting adjacent to the 130 km transform fault from Skagafjordur to Husavik is thus partly controlled by the stresses which the intermittend strike slip movements along discrete parts of the fault build up. The high activity of the HFF during our times also means that only a fraction of the plate divergency would be taken up on the Grimsey lineament. The earthquake activity on the Grimsey lineament following the Krafla rifting epizode needs not necessarily have involved much net plate motion. The rifting, which was up to 10 meters in the rift area close to Krafla, was less than 2 meters in the Öxarfjörður area. Although it was only two meters it caused a temporary stress disturbance in the area to the north of Öxarfjörður, in a system that is ballanced by pore pressures exerted by basaltic fluids from below the crust and the shear stresses related to the general plate motion in the area. It is obvious that the model described here deviates significantly from some models that have been previously hinted at. However any modelling of the present day dynamics of this area is poorly constrained by geological as well as geophysical observations. Among significant tasks related to seismology that should be carried out to try to constrain better a model for Husavík earthquakes are the following:
Comprehensive seismicity study, including SIL data, OBS data, catalogue data 1927 - 1990, special data, international and Icelandic related to epizodes in the TFZ and NVZ.
Various maps created by former studies in the TFZ, like gravity maps, magnetic maps, various depth maps should be included in the study and prepared to be compared directly to seismicity maps, and maps of faults.
Special tasks should be carried out with the objective to constrain better the present time (1000 years time scale) movements of the fault, especially near Húsavík. This could possibly include geological studies of sediment movements, by borehole coring or by digging trenches in Husavík or by other methods? How well are rifting epizodes in the NVZ known before 1725? Is it possible to enhanche this knowledge?
The possibility of major fluctuations in activity in this area may limit the possibilities of being able to achieve long term prediction or hazard assessment. This would increase the significance of medium term or short term prediction research. Generally speaking large earthquakes are considered to be preceded by aseismic motions on the earthquake fault, leading to modification of the stresses adjacent to the fault. How long this premonitory movement lasts, is not well understood yet. However it should be continuously monitored. This is done now with the SIL seismic system and measurements have been carried out by GPS survey at the eastern end of the fault, near Húsavík. It is a question if more monitoring should be proposed like additional seismometers near the fault, continuous GPS, volumetric strainmeters etc.?