S51D-01 INVITED 08:00h
Geophysical Characterization of Seismogenic Structures in Northern California
In preparation for the centennial of the great 1906 San Francisco earthquake, we are developing a regional three-dimensional seismic wavespeed model for Northern California. The model will be used to compute strong ground motions in a simulation of the 1906 event and to help characterize both well-known and hidden seismogenic structures. The dataset combines P-wave arrival times from a well-distributed set of more than 6,000 earthquakes with a complete archive of available active-source P-wave travel times from the region. The wavespeed modeling is being carried out with a combination of conventional and double-difference seismic tomography. As a first stage, a preliminary model has been developed covering the region west of the Great Valley from Hollister to Clear Lake. The geometries of the major faults are clearly defined by the relocated seismicity, and most of the seismogenic faults are marked by significant wavespeed features. We take advantage of our high-quality wavespeed model and the abundant seismicity located with this model to examine the power of other geophysical observables to detect and characterize seismogenic faults in the region. Seismically active strike-slip faults with reasonable surface exposures predominate in the region, but we concentrate our analysis on faults or fault segments that are more challenging to characterize. We explore the degree to which gravity, magnetic, deformation, and topographic data, alone and in various combinations, can identify these seismogenic zones. For example, basement structures identifiable in the gravity and magnetic data (and in the wavespeed model) appear to control the seismicity associated with the Ortigalita fault (which runs through San Luis Reservoir), an unmapped strike-slip fault connecting the Calaveras and Greenville faults, and an apparent thrust fault on the edge of the Great Valley near Vacaville. We also investigate the degree to which active-source seismic data alone can resolve structural features associated with these faults. Our findings can provide some guidance for geophysical studies of faults in areas with more limited seismicity and seismic networks.
S51D-02 INVITED 08:20h
Characterization of Seismogenic Faults of Central Japan by Geophysical Survey and Drilling
Integrated investigations on seismogenic faults by geophysical survey and drilling are indispensable to better understand deep structure and physical properties of a fault fracture zone. In central Japan, three large active faults, Neodani, Atotsugawa and Atera faults, exist and are remarkable for research because of the potentiality of a scale of magnitude 7 to 8 class earthquake and the different characteristics of the seismogenic activities in these faults. Each individual fault shows its own characteristic features, which may reflect different stages in an earthquake cycle. High seismicity is concentrated with a clear lineation on and around the Atotsugawa fault, which is recognized as aftershocks from the latest event of the 1858 Hida earthquake (M=7.0). On the other hand, extremely low seismicity is found around the Atera fault, of which some parts seemed to be dislocated by the 1586 Tensyo earthquake (M=7.9). As an example of the results of study at the Atera fault, we obtained a wide variety of fault structures, composed materials, states of crustal stress and strengths of the fault from the geophysical survey (resistivity and gravity) and in-situ borehole experiments. Our findings are as follows: (1) The fracture zone around the Atera fault shows a very wide and complex fracture structure, from approximately 1 km to 4 km wide. (2) The average slip rate was estimated to be 5.3 m /1000 yr by the distribution of basalt in the age of 1.5 Ma as determined by radioactive dating. We inferred that the Atera fault has been repeatedly active in recent geologic time; however, it is in a very weak state at present. (3) Stress magnitude decreases in the area closer to the center of the fracture zone. These are important results to evaluate fault activity. Recent in-situ downhole measurements and coring through active faults have provided us with new insights into the physical properties of fault zones. In the vicinity of the epicenter of the 1995 Hyogo-ken Nanbu (Kobe) earthquake, we have conducted an integrated study by using 1,000 m to 1,800 m deep drilling wells. In particular, the Nojima-Hirabayashi borehole was drilled to a depth of 1,838 m and directly intersected the Nojima fault. Three possible fault strands were detected at depths of 1,140 m, 1,313 m and 1,800 m. Major results obtained from this study include the following: (1) Shear stress around the fault zone is very small, and the orientation of the maximum horizontal compression is perpendicular to the surface trace of faults. (2) From the results of a heat flow study, the lower cut-off depth of the aftershocks was estimated to be roughly 300 _E#8249;C. (3) Cores were classified into several types of fault rocks, and an asymmetric distribution pattern of these fault rocks in the fracture zones was identified. (4) Country rock is characterized by very low permeability and high strength. (5) Resistivity structure can be explained by a model of a fault extending to greater depths but with low resistivity. The integrated study by geophysical survey, drilling and core analyses, downhole measurements and long-term monitoring directly within these fault zones, provide us with characteristic features and dynamics of active faults.
S51D-03 INVITED 08:40h
Definition of Seismogenic Faults by Combining Seismological, Geological and Geophysical Data: Case Studies From the Aegean Sea Area
Geological and geophysical data are often combined with seismological observations to define faults that are related to past, as well as recent catastrophic earthquakes. Such a combination of information from different disciplines is rather required in areas such as the Aegean, where the sparseness of permanent seismological networks on one hand, and the high seismotectonic complexity on the other hand, pose constraints on the efficient definition of seismogenic sources by using seismological methods alone. The 1999 Athens earthquake forms a characteristic example, where combination of seismological, geological and interferometric data was required to accurately locate the seismogenic fault. Proposed hypocenter locations of this earthquake differ as much as 16 km in the horizontal direction and 22 km in depth. The different sets of hypocenter parameters were tested through forward calculations of the surface displacement field, taking into account the distribution of slip during this earthquake, and the resulting fields were compared to interferometric and geological data to determine the best source model. Another example of the effectiveness of the combined geological, geophysical and seismological data sets is the recent determination of the western continuation of the Ganos Fault (Turkey) that ruptured in 1912, within the Gulf of Saros, in northeastern Aegean Sea. Until recently, geological observations supported the theory that the continuation of the Ganos Fault in the Gulf of Saros coincides with the fault along the northern coasts of the Gelibolu peninsula. Nevertheless, the combination of geological information with recent geophysical and seismological data tends to cancel this theory and rather supports that the western continuation of the Ganos fault is located close to the northern border of the Saros depression. These combined studies introduce new data for future seismic hazard studies in the broader area.
S51D-04 09:00h
Identifying seismogenic faults: Case studies from SW Iceland
Many large earthquakes originate from either poorly constrained or unknown faults, making hazard assessment difficult in earthquake-prone regions. Here we demonstrate the utility of integrating results from field surveys, seismic observations, and remotely sensed data to identify earthquake faults on the Reykjanes Peninsula, Iceland. This Peninsula is an oblique segment of the Mid-Atlantic ridge, and because of its extreme obliquity with respect to the direction of absolute plate motion, it experiences both extension and shear, resulting in slip along both normal and strike-slip faults. Large normal faults and eruptive fissures are prominent on the Peninsula, but strike-slip faults have a much less obvious surface expression. Since 1998, six earthquakes with local magnitudes around five have occurred on the Peninsula, with seismic and geodetic data indicating their occurence along generally north-striking, right-lateral, strike-slip faults. Only one of these faults had been mapped previously, and primary rupture associated with the earthquakes was found along only two faults. Survey work along faults subject to aftershock activity constrained surface disruption paralleling local topographic features and structural lineaments. We review the above examples and explore the feasibility of using similar techniques in order to identify seismogenic faults in SW Iceland before damaging earthquakes occur. We will illustrate how GIS is allowing time-dependent insights into the seismic history of the Reykjanes Peninsula.
S51D-05 09:15h
Fault Mapping in the Hengill Region, SW Iceland by Joint Interpretation of Microearthquake Distribution and Collective Focal Mechanisms.
Joint interpretation of microearthquake distribution and individual focal mechanisms is used to map faults and slip directions during a period of uplift and large-scale faulting in the Hengill region, Southwest Iceland. Earthquake locations are obtained through cross-correlation of P and S waveforms and subsequent inversion of relative arrival times; focal mechanisms are obtained by inversion of spectral amplitudes of P and S waves. The tectonic activity, during the years 1993 to 1998, at the intersection of the Western Volcanic Zone and the South Iceland Seismic Zone (SISZ) induced over 80 thousand microearthquakes. The cause has been attributed to emplacement of magma into the crust, under the Hengill region. The often episodic seismic activity was mostly confined to the region north of the center of uplift, which is at the northern margin of the SISZ, but occasional swarms also occurred at the southern margin of the SISZ. Faulting, however, did not ruptured through the Seismic Zone until June 1998, when the seismic activity culminated in in a three-day swarm producing over two thousand earthquakes. The activity started north of the uplift center and finally ruptured south, through the SISZ on the second day. This breaking through the SISZ barrier produced a M~5 event and two M~4 events. Later the same year the volcanio-seismic episode finally ended in another three-day swarm confined to the southern margin of the SISZ, which produced a second M~5 earthquake. In the region north of the uplift center, the faults revealed by the microearthquake analysis show an interweaving pattern of strike directions, mostly varying between NE and E, but with some short N trending sections. The fault sections are mostly near-vertical and commonly a few hundred meters to a kilometer long. Slip directions are generally right lateral on NE striking faults and left lateral on E striking faults. Normal faulting is also observed. In the SISZ, West and South of the center of uplift, the predominant fault direction is N and fault patches can be as long as 3 km. Slip direction on these is predominantly right-lateral strike-slip, characteristic for faulting in the SISZ. Event distribution around the epicenter of the two main shocks indicate that they occurred on near-vertical N striking faults in agreement with the focal mechanisms. An approximately 1 km long arching fault scarp, which was mapped on the surface, 1 km northeast of the M~5 June epicenter, therefore is probably not due to the main event.
S51D-06 09:30h
Characterization of the Seismic Stress/Strain Fields Around the Bend of the Western/Central Alps
The western Alps tectonic regime is characterized by ongoing widespread extension in the highest zones of the belt and transpressive/compressive tectonics at the external limits of the belt [Sue et al., 1999, Delacou et al, 2004]. This contrasted tectonics is examined using a global synthesis of 389 reliable focal mechanisms covering the whole belt. In term of deformation state, an original method of regionalization allowed us to precisely map the different tectonic modes in the belt. Extension appears as the main feature of the current activity in the western Alps, and affects their inner areas as a whole, following the arcuate geometry of the chain. Shortening is limited to local areas at the outer limits of the chain. Strike-slip is observed in the whole alpine realm. In term of stress analysis, we inverted the whole database of focal mechanisms in homogeneous areas determined using our regionalization analysis. The stress state is confirmed to be radial both concerning s3 in the inner extensional zones, and s1 in the outer transcurrent/tranpressional zones. Extensional areas are correlated with the part of the belt which presents the thickest crust, as shown by the comparison with the Bouguer anomaly map and the smoothed topography of the belt. The overall geodetic strain corresponds also to a radial extension across the western Alps [Calais et al., 2002]. Indeed, there is a quite good qualitative coherency between seismotectonic and geodetic approaches. To compare the seismic strain with the overall geodetic strain, we attempted to quantify the seismic part of the deformation. We used the regionalization of the seismic deformation to determine sub-areas of tectonically homogeneous seismic strain. In each sub-area, we computed the total seismic moment tensor, and the associated yearly seismic strain rate. This rate allows to compare different areas of the belt, in which the seismic catalogue may cover various range of time. Thus we can estimate the seismic strain per sub-area in the whole belt. Only some percent to some tens of percent of the geodesy-related deformation could be explained by the alpine seismic activity. This result rises up the issue of aseimic deformation in the Alps. This approach brings new quantitative elements to understand the ongoing geodynamic processes in the alpine belt. The low seismic strain rates we obtain in a belt characterized by a high tectonic contrast in a quite limited area, could suggest that the Alps are currently in a meta-stable tectonic state, ruled by isostasy/buoyancy forces rather than European/Apulia plate tectonic collision. REF: Calais et al., 2002, Geology, 30-7, 651-654; Delacou et al.,2004 GJI 158, 753-774; Sue et al., 1999, JGR, 104, 25611-25622.