Seismicity and Seismo-tectonics of Alaska: Alaska as a Natural Laboratory for Earthcsope

Transcription

1 Seismicity and Seismo-tectonics of Alaska: Alaska as a Natural Laboratory for Earthcsope Roger Hansen, Natasha Ratchovski, Trilby Cox, and Douglas Christensen Geophysical Institute, University of Alaska Fairbanks, Box , Fairbanks, AK roger@giseis.alaska.edu Alaska is characterized by high seismicity due to the active subduction of the Pacific Plate beneath the North American Plate (~5 to 7 cm/year), and is the most seismically active region in the United States by a large margin (See Figure 1). Nearly the entire state of Alaska is seismically active and thus offers an excellent natural laboratory for the study of earthquake occurrence and a variety of tectonic and volcanic processes. Sources of Earthquakes in Alaska Number of Major Earthquakes International Seismological Center Major (M 5.5) Earthquakes During 18 Year period Year Compiled by Steve Estes, September Alaska California We believe it is very important that Alaska be included as a strong component of Earthscope. In particular the PBO and USArray components of Earthscope can benefit scientifically from the complex setting of Alaska. First, a strike slip type plate boundary is located in the Southeast. Many of the faults continue into western Canada then through Alaska creating some of the largest faults in the U.S. (Fairweather, Queen Charlotte, Denali, Tintina, etc.) The plate boundary then transitions to a continental collision of the Yakutat block with Alaska, raising the Chugach and St. Elias ranges. This region is better known to seismologists as the region of the Yakataga seismic gap. Tectonically, understanding the effects of the collision of the Yakutat block is the linchpin to understanding the present tectonics of most of southern and western Alaska. Yet we do not know which faults absorb most of this relative motion, or how this collision may drive motion and seismic activity on the Denali fault and faults in western Alaska. The Aleutian arc is an ideal place to study how changes in obliquity of subduction affect subduction processes. Seismicity rates for Alaska as a whole are large (e.g. Figure 2). Alaska is a vast, untapped resource for seismic studies that if properly instrumented will provide a wealth of data. Ten great earthquakes have occurred along the Aleutian trench since Alaska averages one M8 event every 13 years and one M7 every year. M7 Page 1

2 events are a possibility virtually anywhere in Alaska, and M6-7 events occur at a rate of at least 5 per year. With modern instruments installed at densities such as big foot, much could be learned about larger earthquakes and their effect. In some areas, transportation and communication costs are slightly higher in Alaska and this increases the cost of operating seismic networks. However, the probability of capturing a significant earthquake is also much greater than in other regions resulting in a greater scientific and engineering payoff which more than offsets any higher installation and maintenance costs. Installing geophysical instrumentation in Alaska is a safe, cost-effective, scientific investment. Distribution and Segmentation of Subduction Zone Seismicity The deep seismic zone extending from Cook Inlet north to beneath Denali (the highest mountain in North America) is located beneath land, and therefore accessible to intensive study. The strike of the deep seismic zone runs nearly NS, and dips at an angle of about 45 to the W, approximately, reaching about 200 km depth. As a function of depth, the seismicity rate peaks at about 70 km, remains strongly active to about 130 km, but declines rapidly below that depth. This deep seismic zone produces earthquakes at a very high rate. Between 1988 and 2000 more than 16,000 events with M1.9 and depths larger than 50 km have been located and they were produced at a steady rate. Thus, at least 1,300 events per year are available for study. 60 N 58 N 62 N 156 W 156 W Kodiak block Kenai block McKinley block C D B A The distribution of seismicity is illustrated in Figure 3, taken from a recent article on the relocation of subduction zone events (Ratchkovski and Hansen, 2001). A total of 14,102 subduction zone earthquakes from the AEIC catalog were relocated using the Joint Hypocenter Determination method. The relocated subduction zone earthquakes have more accurate relative locations and provide more detail on the structure of Alaska Wadati-Benioff zone. In particular, we were able to identify more precisely boundaries between southwestern (Kodiak), central (Kenai) and northeastern (McKinley) blocks of the Alaska WBZ. The boundary between the Kodiak and Kenai blocks seems to be represented by a smooth bend, with no evidence for a tear in the subducted plate. The boundary between the Kenai and McKinley blocks, however, is represented by a tear in the plate that can be traced as far as Prince William Sound. In addition, there is evidence for plate segmentation within the McKinley block. The boundaries between the segments are marked by seismicity gaps and offsets and abrupt dip changes. We suggest that the density of the existing seismograph network is not sufficient for future studies of the Alaskan subduction zone. The Bigfoot array will allow studies such as analysis of the effect of heterogeneous slab structure on generation of the anomalous wave phases, relationship between the Alaska and Wrangell subduction zones, investigations of the subduction zone volcanism and/or absence of such, and GPS studies of the crustal movements above the subducting slab. 62 N 60 N 58 N Figure 3. Interpretation of the plate segmentation in southern and central Alaska based on the new locations of the subduction zone earthquakes (each color corresponds to a 25 km depth interval, starting from 25 km depth). Thick dashed lines are the boundaries between the three plate segments: southwestern (Kodiak), central (Kenai), and northeastern (McKinley). Thin dashed lines and letters A through D show segmentation of the McKinley block. Red diamonds are the volcano Page 2

3 Study of a Sequence of Strong Intraplate Earthquakes in the Kodiak Island Region of Alaska A sequence of strong earthquakes was registered in in the Kodiak Island region of the Alaska-Aleutian subduction zone (Figure 4). Two M W 7 earthquakes occurred in December, 1999 and January, 2001 and an M W 6.5 event occurred in July, These events and their aftershocks were recorded by the regional seismograph network. Regional broadband data were used to obtain seismic moment tensors for the main shocks and their largest aftershocks. Relocation and moment tensor inversion results indicate that these events originated inside the subducting Pacific plate, similar to the recent Nisqually earthquake. The focal mechanisms indicate down-dip tension with the fault planes being nearly vertical and parallel to the strike direction of the subducting plate. The 1999 and 2000 events were located downdip of the locked portion of the megathrust, while the 2001 event was located directly beneath it. Concentration of moderate and strong earthquakes near the edges of the 1938 rupture zone may be an indication of the higher stress concentration due to the plate interface being locked in the area. Could the 1999 and 2001 Kodiak Island M W 7events be a precursor activity of the next rupture of the Figure 4a. JHD 1938 asperity or a part of the 1964 rupture zone? We can not locations of the regional earthquakes, 1999 MW 7, answer this question with a high degree of certainty. Besides 2000 MW 6.5, and 2001 MW 7 earthquakes and their the 1999 and 2001 intraplate events, the only other known aftershocks. Focal mechanisms of the main shocks and strong events of the same nature in the Kodiak Island vicinity their largest aftershocks are shown. are the two 1912 m b 7.3 and 6.9 events that occurred at 90 km depth and within a 20 km radius of the 1999 M W 7event. They preceded the 1938 and 1964 earthquake by 26 and 52 years, respectively. The estimated recurrence times of large earthquakes in the Kodiak Island area is approximately 60 years [Nishenko and Jacob, 1990]. According to these estimates, the rupture zone of 1938 event may be due for another event of similar magnitude. Unfortunately, the time span of the instrumental seismic observations in the Kodiak Island region is not enough to enable us to observe a causal relationship between the megathrust ruptures and the occurrence of the strong intraplate events. This new information on the sequence of strong intraplate events in the Kodiak Island vicinity lays the groundwork for an opportunity of a focussed study with USArray. Figure 4b. Schematic representation of the subduction beneath the Kodiak Island region. Locations of the 1999 and 2001 MW 7 earthquakes and their mechanism are from the regional study. Position of the upper surface of the subducting slab is constrained using the relocated seismicity. Width of the seismogenic interface is a down-dip extension of the 1964 rupture zone. Page 3

4 Focussed Study of the Tectonics and Stress Regime in Interior Alaska 63 N 154 W 67 N 66 N 65 N 154 W SHmax 0-10 km km 150 W 150 W km km 146 W km km 146 W 67 N km km The tectonic framework of Alaska is dominated by subduction of the Pacific plate underneath the North American plate. Stresses due to the plate convergence are transmitted across great distances (> 500 km) into interior Alaska where the deformation causes substantial crustal seismicity. While some of the earthquakes are clearly associated with the large-scale strike-slip fault systems of Denali in the south and Kaltag and Tintina in the north, the majority of the shocks are located in a zone of distributed shear deformation between the two fault systems (Figure 5). These earthquakes are aligned in three major northnortheast trending zones which are called Minto Flats (MFSZ), Fairbanks (FSZ) and Salcha (SSZ) seismic zones. Among the largest earthquakes in the interior are the 1904 M S 7.3, 1929 M S 6.2 and 6.5, 1937 M S 7.3, 1947 M S 7.2, and 1968 M S 6.5 Rampart earthquakes. Locations for the three earlier shocks are rather uncertain, but can be generally identified with the area between the Denali and Tintina fault systems (Page et al., 1991). Little is known about geological structures that produce this broadly distributed seismicity. The area between the Tintina and Denali fault systems is occupied by the Yukon-Tanana terrane, an assemblage of Paleozoic and older metasedimentary, metavolcanic, and metaplutonic rocks that are multiply deformed and regionally metamorphosed (Foster et al., 1994). These rocks are cut by a suite of Cenozoic northeast-striking lineaments and sinistral-slip faults. Several of these faults show evidence suggestive of late Cenozoic displacement. Suggestive, but not definitive, evidence from trenching studies and geomorphic features suggests late Pleistocene displacements (Page et al., 1995). The northeast-trending earthquake lineaments are diffuse which may indicate a presence of overlapping parallel faults, splays and cross faults. Page et al. (1995) suggested a block-rotation model to characterize the deformation in interior Alaska. In this model, the crustal blocks are rotating clockwise in a dextral shear zone between the Denali and Tintina fault systems. The deformation accommodates shortening across the region in response to northnorthwest compression resulting from plate convergence. The Kantishna cluster is the most active crustal seismic source in the interior. It is located just north of the Denali fault and is composed of two distinct strands of seismicity aligned in a roughly SW-NE and WNW-ESE directions, respectively, and directly above the deepest seismicity in the Benioff zone north of Mt. Mckinley. The distinct separation between these two strands is the major improvement after the joint relocation over the AEIC locations. The majority of the earthquakes are located above 20 km depth. There is, however, a lineation of the earthquakes in the western part of the cluster that extends to a depth of at least 50 km. This deeper seismicity is rather enigmatic, since the thickness of the crust in this area is believed to be about 40 km (Meyers et al., 2000). Figure 5 shows a total of 3,611 earthquakes that occurred from 1988 to 1999 which were relocated using the Joint Hypocenter Determination method. The Minto Flats seismic zone has the most coherent structure among the three major seismic zones in the interior and extends from the Kaltag fault in the north to the Kantishna cluster in the 66 N 65 N 63 N SHmin Figure 5. Relocated earthquakes and orientations of the best fitting maximum (green) and minimum (blue) horizontal stresses for 7 target areas. Dashed polygons outline separate inversion regions with larger symbols representing earthquakes included into the inversion. Page 4

5 south. The Fairbanks and Salcha seismic zones do not extend as far north and south as the MFSZ. Given the blockrotation model of the deformation in central Alaska is valid, the FSZ and SSZ may continue propagating farther north and/or south in a number of moderate- to large-sized earthquakes. The new earthquake locations provide more detail on the structure of the Kantishna cluster and better locations for the aftershock sequence of the 11/29/2000 M L 5.6 earthquake in the Minto Flats seismic zone. The Kantishna cluster is composed of two distinct earthquake lineations of SW-NE and WNW-SES orientations. The JHD locations for the aftershocks of the 1995 M W 6.0 Minto Flats earthquake and 10/22/96 M W 5.7 earthquake near the Denali fault are also available. Locations and mechanisms of the 1995 and 2000 Minto Flats seismic zone events are consistent with the block-rotation model of deformation in central Alaska. A catalog of 196 fault plane solutions consisting of the moment tensor solutions for the earthquakes with magnitude 4.0 or above and P-wave first motion solutions for the earthquakes with magnitude 3.4 and above that occurred from 1988 to 2000 is composed for interior Alaska. This catalog was used to calculate principal stress orientations in the crust. The stress orientations change systematically across central Alaska. In particular, the maximum principal stress orientation rotates clockwise from SE-NW to SSW-NNE direction as one moves from west to east across central Alaska. These stress orientations agree with the stress trajectories inferred from the geological stress indicators (Estabrook and Jacob, 1991) and are consistent with a stress regime we would expect if the Pacific plate were a rigid intrusion into a plastically deforming North American plate (Nakamura and Uyeda, 1980). Stress orientations within the Kantishna cluster are complex showing rapid changes across short distances. Orientations of the maximum principal stresses are identical to those west of the MFSZ and range between ESE- WNW and SE-NW. The intermediate and minimum stresses, however, vary considerably. This complexity may be caused by the location of the Kantishna cluster at the junction of the major tectonic features in the region: the Denali fault in the south and the MFSZ in the northeast. Detailed GPS measurements of the crustal movements and focussed seismic arrays in the area could help resolve this question. Focal mechanisms and stress orientations in the northern part of the cluster indicate thrust type of faulting, while in the southern part they are consistent with the right-lateral movement on a hypothetical fault coinciding with the predominant WNW-ESE alignment of the earthquakes. Analysis of the fault plane solutions for the earthquakes within the Minto Flats seismic zone shows that the predominant orientations of the fault planes changes across 64 o N latitude from a SW-NE orientation north of it to a SSW-NNE orientation south of it. This change coincides with a change in the earthquake alignment. The earthquakes form a coherent southwest-northeast oriented lineament north of 64 o N latitude, while the seismicity is more diffuse to the south. North of 64 o N, the orientation of the maximum compressive stress is rotated approximately 35 o from the prevailing seismicity trend. This angle is indicative of strong, less-well-developed faults (Provost and Houston, 2000) within the MFSZ, which is consistent with the assumption that the MFSZ is composed of a series of subparallel faults and fault splays. Tests with different velocity structures were necessary in the moment tensor inversion procedure for calculating Green s functions for the earthquakes that occurred in the interior. Furthermore, the area south of the Denali fault requires yet a different structure. The moment tensor inversion study shows that it is possible to obtain a reliable moment tensor solution for moderate-sized earthquakes ( M L 4 ) using three-component records from a single broadband station when the epicentral distances are between 50 and 300 km, but the implementation of big foot with some stragically placed special study arrays could greatly improve the resolution of these seismo-tectonic questions, and improve routine catalog production through well calibrated velocity models. Earthscope would supply a wealth of information about Alaska and active tectonics. This contribution (along with colleagues at the University of Alaska) to this workshop outline some of the major studies that we would hope to undertake within PBO, the deployment of the Big-Foot array, and the special study array in Alaska. Page 5