Fault activation and fault dynamics during the 1993-1998 uplift episode in Hengill, SW-Iceland

The Hengill geothermal area in SW-Iceland is located on a triple junction where two extensive and one conservative plate boundary meet. In this complex tectonic setting magmatic and geothermal processes are at work as well and all these processes are accompanied by seismic activity which we use to learn about fault activation and to analyze the dynamics and fault development over time. The mapping of fractures provides valuable information on permeability at depth which can help to position future well bores which is of great economic interest since the Hengill geothermal area is the biggest exploited geothermal area in Iceland. In the 1993 to 1998 the Hengill area underwent a period of unusually high seismic rates when an uplift episode caused a land rise of 8 cm and triggered more than 90.000 earthquakes. This uplift was caused by a small inflation source in about 7 km depth, possibly of magmatic origin. As modeling revealed, the stress induced by the uplift was not sufficient to explain the release of seismic energy, indicating that the crust must have been at high stress and close to failure prior to the uplift episode which acted only as a trigger. We use cross-correlation and apply a relative relocation algorithm to the catalogue earthquakes which significantly improves the accuracy of both the relative and absolute locations. After relocation alignments of earthquake become visible and show mostly orientations in NNE and ENE direction. We recalculate the focal mechanisms for the new locations and jointly interpret the results together with the focal mechanisms to calculate the planes that best describe the faults. From the focal mechanisms the average slip vector on each fault can be estimated and we learn about the stress field. The relocated earthquake distribution shows that the stresses induced by the uplift event must have been small in comparison to the regional stress since the activated faults do not respect the geometry of the uplift source but are rather in agreement to the regional stress field. The uplift did not cause any new breaks in the crust but rather reactivated existing faults which sub-optimally oriented in relation to the uplift. Some faults are active constantly with varying earthquake rates. Other faults and fault groups are activated repeatedly for short time intervals over the observation period.