S13A-1044
Fault model of the 2007 Solomon earthquake estimated from the crustal deformation survey data
On April 1, 2007, a large earthquake occurred off the Solomon Islands along the Solomon subduction zone. The earthquake generated a large tsunami that killed more than 40 people in Gizo and Simbo Islands near the epicenter. The one day aftershock distribution showed that the source region was located in the subduction zone where the Woodlark ridge system subducted beneath the Pacific plate. Because of the subduction of the ridge, no trench exists near the plate boundary. Instead, two Islands, Simbo and Ranongga Islands, exist unusually close to the plate boundary. About two week after the earthquake, the Japanese survey team was in the Solomon Islands to conduct the coseismic deformation survey in Gizo, Simbo, Ranongga, Vella Lavella, and Kilimbangara Islands near the source area of the 2007 Solomon earthquake. A whole island of Ranongga was uplifted by the earthquake because a large area of coral flats around the island was now appeared above high tide levels. In Simbo Island, located about 20km south of Ranongga Island, we found small subsidence. In Vella Lavella Island, the most part of the island was subsided except the most southeast tip of the island. In Gizo Island, the small subsidence was found along the most part of the coast. Using those crustal deformation data, we estimated that the strike of the fault is 315 degree, the width of the fault is 35 km, the dip of the fault is about 35 degree which is much larger than a typical dip of the plate interface near a trench, the shallowest edge of the fault is located between Simbo and Renongga Islands, and the depth of the shallowest edge is less than 5km. This unusual splay fault type earthquake occurred in this area where the Woodlark ridge subducted beneath the Pacific plate.
S13A-1045
Source Rupture Process of the Solomon Islands Earthquake of April 1, 2007 Inferred from Teleseismic Body Waves
A large earthquake (Mw 8.1) followed by a tsunami took place in the southeast Pacific along the New Britain subduction zone on April 1, 2007. This region displays a complex tectonic nature where 4 plates intersect. Along the Solomon Island convergent margin the relatively small Woodlark and Solomon plates enter into the subduction zone side-by-side with the much larger Australia plate and the boundaries of these three plates are defined by transform faults. On a regional scale this subduction zone plate boundary is characterized by the occurrence of large earthquake doublets in 1971, 1974, 1975, 1977 and 2000. We investigated the source process of the April 1, 2007 earthquake using three different faulting patterns; (1) fixed thrust mechanism over the entire extend of the fault, (2) varying slip directions over the area of rupture and (3) varying focal mechanisms along the faulted region. The teleseismic body wave inversion technique that we used in our analysis of the source parameters for the three different fault models consistently yield a relatively large fault area (approximately 300 km by 50 km) with an overall seismic moment on the order of 1.0 x 1028 Nm. For the fixed mechanism and varying slip direction models, a major part of the seismic moment is released in the form of two pulses separated by 15-20 seconds. The second pulse is the largest one and it is located northwest of the hypocenter, implying a northwestward directed unilateral rupture. Our results indicate that the location of this maximum seismic moment release is close to the centroid location determined by Global CMT and also spatially coincides with the projected subduction of the transform boundary between the Australian and the Woodlark plates. The distribution of the dislocations computed for the varying slip direction model along the megathrust is characterized by three isolated patches of varying slip amounts and directions. One of these patches is located at the hypocenter of the event and the other two spatially coincides with boundaries of the three subducting plates. These transform boundaries display bathymetric highs (ridges) all along their trend which might have acted as asperities in the coupling zone of the plate interface and serve as the locations of major seismic moment release and slip during the rupture. The inversions for the varying mechanism model indicate some component of seismic energy was released with a strike-slip mechanism near the intersection of the transform boundary.
S13A-1046
Successful Forecast of the Place of Great Solomon Islands Earthquake of April 1, 2007, Mw8.1
Predicting the large and great earhquakes proved to remain at present a complicated and most difficult problem, regarding the place and magnitude, and especially the time, on medium- or long-term interval, which is the principal component of any prediction. As a result a few (scientific) predictions are really successful, with a desired precision. False alarms and failures are by far dominant. However, the incresed number of higher accuracy data, advancements in the rupture- imaging at a smaller scale, seimotectonics details, increase of GPS data, different deterministic and probabilistic methods and numerical experimnts, retrospective predictions, etc. All these have to contribute, to increase a knowledge necessary to find better prediction-laws and methods in order to make particular earthquakes in specified regions more predictable. Along this view, we present a successful prediction (deterministic prediction) of a large earthquake in the Solomon Islands subduction zone: 154°E to 165°le E, where large earthquakes are occurring. We analized in detail historical earthquakes and aftershock areas (Kelleher et al, 1973, 1974; McCann et al 1979; Lay and Kanamori, 1980) of large events, their distribution and space and time, and the fluctuations of earthquake activity. We identified, in 1984, a seismic gap in the Central region of the subduction zone, with the boundary to NW: 156°E and to SE: 157°E. Its dimension (L = 180Km) corresponds to a magnitude of 7.8-8.0. No time prediction was made. The gap bundaries were fixed to NW at the SE-end of aftershock areas of the Jan - Feb. 1974 earthquake doublet (M7.0, M7.1) and to SE at the location of the Jan. 30, 1939 (Ms7.9-8.0) and its aftersock area was not known. The 2007 Solomon epicenter is well inside the gap. No evidence or rationale at that times would at least suggest that a future large earthquake in the gap will possibly rupture further to NW, and if the contrary than it is only a coincidence. Therefore the forecasted earthquake has closed very well the gap. The 6-days (also the 8-days) earthquake aftershock zone (PDE - data), 155°E - 157.5°E, is grater than that of the gap. This interval is very well consistent with the lengths of rupture models based on different inversion methods (C. Ji, Sanchu, Y. Yagi). Also we found the gap remained very stable until the occurrence of the 2007 Solomon earthquake. Fnally, a new seismic gap, bounded 158°E - 160°E, is proposed following the 2007 event. If the 1939 earthquake occurred in this gap is, however, presently unknown to us. If strong evidence can support the 1939 rupture place in the gap, then the future earthquake will close the first known seismic cycle for this plate boundary segment. The space-time evolution of the gap must therefore be continously under observation and further analyzed.
S13A-1047
Geologic Survey of the 2 April 2007 Solomon Islands Earthquake and Tsunami
The 2 April 2007 magnitude 8.1 Solomon Islands earthquake and tsunami caused extensive damage to coral reefs, coastal erosion, and in some locations, 3 meters of uplift, subsidence, and numerous landslides in the Western and Choiseul Provinces. Extensive damage to the coral reefs ranged from shattered branching corals to 4 meter head corals snapped off their bases and toppled over. The fringing reef on the east coast of Ranongga sustained the greatest degree of damage as it was uplifted 3 m above sea level and remains completely exposed. Sediment samples were collected along transects extended from offshore to onshore environments for larger islands, such as Ghizo, where the tsunami did not pass over the entire island. Smaller islands, such as Nusa Aghana, a transect was conducted from the outer barrier reefs, through the lagoon, across the island, and offshore on the opposing side of the island. Offshore data was collected using a side-scan sonar system that records bathymetry and images coral reef morphology. This data was coupled with snorkeling and SCUBA diving to ground truth the offshore lagoon and reef environments. Sediment samples were collected offshore every 5 m and were documented by underwater photos and GPS coordinates. Offshore to onshore sediment transects reveal that sediment was eroded from seaward facing shorelines, deposited a thin veneer of sediment on islands, and transported the majority of the sediment on coral reefs on the lagoon side of islands, essentially burying coral and lagoonal sediment. Coral reef damaged by the earthquake and tsunami represents a major concern for an already threatened ecosystem. Recovery of the fishing and dive tourism economies rely on the healthy reestablishment of the reef.
S13A-1048
Spatial Distribution and Sedimentary Facies of the 2007 Solomon Islands Tsunami Deposits
We conducted a field survey of the extent of damage, crustal deformation, and onshore deposits caused by 2007 Solomon Islands tsunami in Ghizo and adjacent islands in the western Solomon Islands, from 13th to 18th April, 2007. Our survey team was comprised of six Japanese and one American researcher. Three of us, the authors, mainly investigated tsunami deposits in three villages (Titiana, Suva, and Pailongge) in southern Ghizo Island. One member of our team re-investigated the deposits in June 2007. The tsunami generated sheet-like deposits of coral beach sand on the flat plain in Titiana. Beside the sea coast, the tsunami wave eroded ground surfaces and formed small scarps at 30 m from the sea. Just interior of the scarps, tsunami deposits accumulated up to 9 cm in thickness. The thickness decreased with distance from the sea and was also affected by microtopography. No sandy tsunami deposits were observed on the inland area between 170 m and 210 m from the sea. The upper boundary of inundation was recognized at about 210 m from the sea because of accumulation of driftwood and floating debris. In Suva and Pailongge, the outline of sand-sheet distribution is the same as it in Titiana. The tsunami had a maximum thickness of 10 cm and two or three sand layers are separated by thin humic sand layers. These humic layers were likely supplied from hillslopes eroded by the tsunami and transported by return-flows. These successions of deposits suggest that tsunami waves inundated at least two times. This is consistent with the number of large waves told by eyewitnesses. In the Solomon Islands, the plentiful rainfall causes erosion and resedimentation of tsunami deposits. Furthermore, the sedimentary structures will be destroyed by chemical weathering in warm and moist environment, and bioturbation by plants, animals, and human activities. The sedimentary structures had been preserved till the end of June 2007, but had already been penetrated by plant roots and sandpipes of crabs. We believe that the knowledge of weathering process of tsunami deposits is important for interpretation of sedimentary structures of paleo-tsunami deposits.
S13A-1049
Tsunami Field Survey for the Solomon Islands Earthquake of April 1, 2007
Two weeks after the 2007 off-Solomon earthquake, an international tsunami survey team (ITST) of Japanese and US researchers performed a post tsunami survey in Ghizo and adjacent islands. Main purpose of the team was to provide information on the earthquake and tsunami to the national disaster council of the Solomon Islands, who was responsible for the disaster management at that time. The ITST had interview with the affected people and conducted reconnaissance mapping of the tsunami heights and flow directions. Tsunami flow heights at beach and inland were evaluated from watermarks on buildings and the position of broken branches and stuck materials on trees. These tsunami heights along the southern to western coasts of Ghizo Island were ca. 5m (a.s.l.). Tsunami run-up was traced by distribution of floating debris that carried up by the tsunami and deposited at their inundation limit. The maximum run-up was measured at Tapurai of Simbo Island to be ca. 9 m. Most of the inundation area was covered by 0-10 cm thick tsunami deposit that consists of beach sand, coral peaces and eroded soil. Coseismic uplift and subsidence were clearly identified by changes of the sea level before and after the earthquake, that were inferred by eyewitness accounts and evidences such as dried up coral reeves. These deformation patterns, as well as the tsunami height distribution, could constrain the earthquake fault geometry and motion. It is worthy of mention that the tsunami damage in villages in Ranongga Island has significantly reduced by 2-3 m uplift before the tsunami attack.
S13A-1050
Tsunami Waveform Modeling Of The 2007 Solomon Earthquake
On 20:40(UTC) April 1st, 2007, a large earthquake (M 8.1) occurred off the south coast of Solomon Islands, and accompanied with tsunami and crustal deformation. The tsunami was recorded at many tide gauge stations in the Pacific Ocean with amplitudes of several to ten and several cm. Tanioka et al. (AGU 2007) proposed a suitable fault model that explains the crustal deformation data obtained by the field survey (Nishimura et al., AGU 2007). The assumed fault is located between Simbo Island and Gizo Island, and its length, width and slip are 100 km, 35 km and 7.5 m, respectively. However the fault size cannot be constrained from the crustal deformation, because most of the fault plane is located in the ocean. In order to examine whether the assumed fault model can explain the observed tsunami, we carried out the tsunami numerical simulation and compared the observed and calculated waveforms in Solomon, Vanuatu, and Australia. The governing equations in the simulation are linear long wave equations in the spherical coordinate system, and the Coriolis effect and the bottom friction are ignored. The calculated tsunami waveforms from the proposed model are similar to the observed ones, in terms of the arrival time, amplitude, and waveform of the first cycle or two. [Acknowledgement] We appreciated the Bureau of Meteorology, Australian government for providing us the tide gauge records.
S13A-1051
Near-Field Population Response During the 2 April 2007 Solomon Islands Tsunami
When the magnitude 8.1 earthquake and subsequent tsunami hit the Solomon Islands on 2 April 2007 it killed 52 people. On Ghizo Island, home of the capital of the Western Province, Gizo, waves approaching 4 m in height inundated the south coast villages. Eyewitness accounts supported by geologic data from the offshore coral reef and sediment deposited on land suggest a wave that came in as the shaking stopped as a rapidly-rising tide rather than a turbulent bore- vehicles and houses were floated inland with very little damage. Those that survived in villages affected by the tsunami had indigenous knowledge of prior events, whereas immigrant populations died in higher proportions. While buoy-based early warning systems are necessary to mitigate the effects of teletsunamis, they would have done little good in this near-field environment. In Pailongge, a village of 76 indigenous Solomon Islanders on Ghizo's south coast, there were no deaths. Village elders directed the people inland following the shaking and the almost immediate withdrawal of water from the lagoon, and heads of household made sure that children were accounted for and evacuated. Of the 366 Gilbertese living in Titiana, however, 13 people died, 8 of which were children who were exploring the emptied lagoon. A large proportion of the dead were children (24) as they were likely too weak to swim against the non-bore flow. The Gilbertese migrated from Kiribati in the 1950"s, and had not experienced a major earthquake and tsunami, hence had no cultural memory. In the case of the Solomon Islands tsunami, as was the case in the 2004 Indian Ocean tsunami, indigenous knowledge served the people in the near-field well. In the case of the Indian Ocean where there was 10-20 minutes separation between the time the shaking began and the waves arrived, the combination of an in-place plan and a suitable physical geography allowed the population of Simeulue Island and the Moken people of Thailand to escape before the waves hit. In the Solomons, there was less than 3 minutes separation time, and the populations with indigenous knowledge were able to save themselves. Mitigation strategies for those that live adjacent to tsunamigenic subduction zones must include a community-based disaster management plan to educate a variety of populations with different cultural knowledges. This education can be in concert with development of an basin-wide early warning system.
S13A-1052
Regional seismic observation of the October 9, 2006 underground nuclear explosion in North Korea and its interpretation
The recent underground nuclear explosion (UNE) in North Korea on October 9, 2006 leaves many issues to be resolved. One outstanding question is how big the event was. Another question is why the Lg wave, which is typically the most prominent regional phase even for UNEs, is not observed much in regional observations at South Korea. This Lg phase is analyzed importantly both for constraints to the size of detonation and for the discrimination of nuclear explosion from natural earthquake. We find that the Lg from the nuclear explosion was significantly dissipated by energy leakage into the mantle due to the variation in crustal structures along the ray paths. A part of the leaked energy develops in a form of mantle-lid waves, causing large energy addition to the Sn. These features are confirmed with numerical waveform modelings and comparisons with waveforms from natural earthquakes. The discriminative Lg attenuation dependent on ray path causes underestimation of the magnitude. We redetermine the magnitude of the UNE from regional records with pure continental ray paths. The body-wave magnitude of the UNE is estimated to be 4.0±0.2.
S13A-1053
Preliminary Seismological Report on the 6 August 2007 Crandall Canyon Mine Collapse in Utah
A large and tragic collapse occurred in the Crandall Canyon coal mine in east-central Utah on 6 Aug 2007, causing the loss of six miners and generating national attention. This collapse was accompanied by a local magnitude (ML) 3.9 seismic event having an origin time of 2:48 am MDT (8:48 UTC) and a location near the collapse. Two lines of evidence indicate that most of the seismic wave energy of this event was generated by the mine collapse rather than an earthquake: (1) the observation that all of the P-wave first motion directions are down and (2) the results of a moment tensor inversion by Ford et al. (2007; http://seismo.berkeley.edu/seismo/Homepage.html). The Crandall Canyon mine is in an area of Utah where there is abundant mining-induced seismicity, including events with both collapse and shear-slip sources. Prior to the 6 Aug collapse, and within a 3 km radius of it, there were 28 seismic events during 2007 that were large enough to be detected and located as part of the routine processing of University of Utah regional seismic network data: 8 in the 2.5-week period prior to the collapse (ML ≤ 1.9) and 15 during an earlier period of activity in late February and early March (ML ≤ 1.8). The 6 Aug collapse was followed by 37 locatable seismic events of ML ≤ 2.2 before the end of August. One of these "aftershocks" (ML 1.6) occurred in conjunction with the violent burst of coal from the mine walls on 17 Aug (UTC) that killed three rescuers. The aftershocks have an exponential frequency-magnitude distribution with a lower ratio between the frequencies of smaller- and larger-magnitude events (lower b-value) than for the prior events in the area. Aftershock rates generally decreased with time through August but there was a noteworthy 5.8-day hiatus in activity that began 37 hours after the collapse. The University of Utah deployed a 5-station temporary network near the mine beginning on 8 Aug. Data from these stations are being used to help develop travel-time corrections for these and other stations in order to improve the computed locations of seismic events that occurred in the area both before and after the 6 Aug collapse.
S13A-1054
Overview of the August 15th, 2007 Pisco Earthquake (Mw=8.0), Central Peru.
With 7 subduction related earthquakes of ~Mw=8.0 during the 20th century, Central Peru is one of the most seismogenic areas in the world. On August 15th, 2007 a large subduction related earthquake (Mw=8.0, GCMT), occurred along the coast of Central Peru, near the town of Pisco. The earthquake was felt as far as 600km away from the epicenter and produced extensive damage in Pisco, where the intensity reached a maximum of Imax=VII (MM), and. Accelerometric records in Lima, 250 km, away from the epicenter, indicate up to 110 cm/s2 peak acceleration. Records from Ica, 80 km from the epicenter, indicate a peak acceleration of 488 cm/s2 on the EW component. Closer to the epicentral area ground shaking was severe enough to produce extensive liquefaction and ground rupture along the coast. First motion polarities indicate a thrust event with a dip angle of about 14°, which is somewhat steeper than the estimated dip angle of the subduction interface in the area. Waveform modeling shows that in fact, the source consists of 2 sub-events. The early afershock distribution well delineates these events which ruptured a portion of the trench that was left unbroken by the more recent 1942 earthquake (Mw 8) to the south, and the 1974 earthquake (Mw 8) to the north. Historic records indicate that the portion of the subduction zone which ruptured August 15th, 2007 had not experienced any comparable or larger rupture since at least 1746 and possibly 1687, which is remarkable given that the local convergence rate between the Nazca and South America plate ( 6-8 cm/yr).
S13A-1055
Source Characteristics of the 2007 Pisco Earthquake (Mw=8.0), Central Peru, Derived From Seismic Waveforms Inversion and Tsunami Modeling
We have undertaken a detailed analysis of the source of the 2007 Pisco earthquake based on the modeling of teleseismic and accelerometric records, and taking into account available information on uplift or subsidence along the coastline, as well as run-up heights. Preliminary analysis of the teleseismic P waves, show that the 2007 earthquake was characterized by two distinct sub-events, occurring 60s apart. The inversion of the teleseismic body waves (P and SH) reveals that the first sub-event was located at the hypocenter, and that most of the energy was released by the second patch, in a similar way to the 2001 South Peru earthquake (Mw=8.4). Assuming that the aftershock distribution is a good approximation of the rupture extent (130 km), the duration of the P wave signal bounds the rupture velocity to values below 2 km/s. The small move-out observed between the P wave phases supports this hypothesis, but also limits our ability to clearly resolve between different low rupture velocities. This resolution limitation is also enhanced by the complex geometry of the slab which prevents the correct inversion of the slip history based on a single 2D fault plane. We use earthquake catalogues to better constrain the subduction interface, and consider a multi-plane geometry for our inversion. The directivity of this event confirms the previously identified tendency of large South Peru, and Chile earthquakes to rupture unilateraly to the south.
S13A-1056
Seismological and Geodetic Observations of the 15th August 2007 Pisco, Peru Earthquake.
On 15th August 2007, a M8.0 earthquake occurred on the interface between the Nazca and South American Plates causing widespread damage in the towns of Chincha Alta, Ica and Pisco, with 514 casualties and 35,500 buildings reported destroyed. The 2007 Pisco earthquake occurred in a region between two large recent earthquakes: the 1996 M7.7 Nazca earthquake to the south and, in 1974, a M8.1 earthquake to the north. The existing local GPS network was resurveyed within 2 weeks of the earthquake and significant displacements measured at 8-10 sites. Co-seismic and postseismic InSAR data has been collected from several satellites, including Envisat, ERS-2, Radarsat and ALOS. In the two weeks following the mainshock, 42 aftershocks were recorded teleseismicly with magnitudes in the range 4-6.3. We will present results from the analysis of this geodetic data along with seismological analysis of teleseismic recordings of the mainshock and aftershocks.
S13A-1057
The August 15, 2007 M=8.0 Ica, Peru Earthquake
On August 15 at 6:40 PM a magnitude 8.0 earthquake occurred off the coast of central Peru. The hypocentral depth was 40 km. The earthquake was located 180 km southeast of the capital city of Lima, near the cities of Ica and Pisco. At least 500 people were killed, and over 1500 were injured. Tens of thousands of buildings were ruined and up to one-hundred thousand people were left homeless with most of the damage occurring in rural areas. In the two week period following the earthquake, at least 16 aftershocks hit Peru, with magnitudes ranging from 4.5 to 6.3. The earthquake occurred at the boundary between the Nazca and South American tectonic plates. The two plates are converging at a rate of 77 millimeters per year, and the earthquake was caused by the downward and eastward movement of the Nazca plate under the South American plate. Since Peru is located at the boundary of these plates, earthquakes are quite common in this area, the last being an M=8.1 near the city of Arequipa in southern Peru in 2001. Prior to this event, the deadliest earthquake to hit Peru was in 1970 when an earthquake in the Peruvian Andes triggered landslides that buried the town of Yungay and killed more than 66,000 people. Since the epicenter of the recent earthquake was in the Pacific Ocean, it was feared that a major tsunami would be generated, however only minor waves were recorded along the Peruvian coast.
S13A-1058
The Mozambique Earthquake Sequence of 2006: High-Angle Normal Faulting in Southern Africa
We report source mechanisms for the six largest shocks of the Mozambique earthquake sequence of February 22, 2006. The main shock of this sequence is one of the largest (Mw = 7) to occur in Africa over the past 40 years and its P-waveforms alone are sufficient to show that north-south trending normal faulting near the surface continues to depths of more than 15 km along an exceptionally steep dip of 76°+/-4°; so globally seismogenic normal faulting spans a wide range of dip from about 30° to 75°. S-waveforms, when combined with a minor component of left-lateral slip observed in the field, indicate the rake to be between -80° and -89° which, in turn, places a new constraint on relative plate motions: Complications from a nonspecific Rovuma micro-plate notwithstanding, our results favor the Euler pole between Somalia- and Nubia-plates to lie southward of the epicenters.
S13A-1059
Rupture Process of the 2007 Noto Hanto Earthquake Inferred From Local P-wave Records: Second hypocenter, Initial Rupture, Main Rupture, Asperity and Killer Pulse
The 2007 Noto Hanto earthquake (Mjma6.9; Mw6.7) occurred on 25 March 2007 (JST) in the Noto Peninsula area of central Japan, which shook strongly the near-source region. Many strong-motion stations recorded the seismic motion near the source region. In this study we analyzed the P-wave part of such records. In the waveform a typical characteristic can be seen. That is an emergent onset. The strong-motion records at local stations except the three nearest stations show nearly two seconds of small but increasing amplitude arrival (hereafter, called "initial rupture phase") followed by the onset of the main energy release (hereafter, called "main rupture phase"). Such an emergent onset is seen on strong-motion records of other earthquakes (e.g., the 2000 Western Tottori earthquake; the 2005 West Off Fukuoka Prefecture earthquake; the 2007 Niigataken Chuetsu-oki earthquake). For the three nearest stations, S wave may have masked the main rupture phase with the lager amplitudes. The P- wave part of the records has another remarkable feature. In the first P onset part of the initial rupture phase two events can be seen. The difference between the two arrivals is about 0.5 s. Using a master-event technique (Takenaka et al., 2006, EPS) we determined the location of the second event (hereafter called "the second hypocenter") relative to the location of the first event (i.e. hypocenter), and then estimated the relative location and time of the onset of the main rupture with respect to the second hypocenter. We furthermore mapped the asperity area on the main fault plane using a source imaging technique based on the back-projection (Yamamoto and Takenaka, 2006, AGU Fall Meeting). From these analyses, we derived the following rupture history of this earthquake: The initial rupture plane and the main rupture plane are different as suggested by the focal mechanism solutions of the P-wave polarities (JMA, 2007) and the CMT (F-net, NIED, 2007). The rupture first propagated toward the direction of strike of the initial rupture plane for 0.5 seconds and changed to the main rupture plane, where this changing point is the second hypocenter. It located at about 1 km apart from the hypocenter. The rupture then restarted toward the direction of Wajima City along the main rupture plane. The main rupture (breaking of the asperity) began at a time of 2.4 s after the second hypocenter onset at a position of 1.0 km and 1.4 km in the strike and updip directions, respectively, from the second hypocenter. The asperity distributes along an updip and strike direction from the second hypocenter and the imaging peak locates around the coast near Monzen-cho, Wajima city in the Noto Peninsula. This estimated rupture history is consistent with damage distribution and spatial pattern of the observed killer pulse (forward rupture directivity pulse) in S-wave portion of the strong-motion records. (Acknowledgements) We used the strong-motion records supplied by the National Institute for Earth Science and Disaster Prevention (NIED; K-NET, KiK-net, F-net), and the Japan Meteorological Agency (JMA).
S13A-1060
Non-linear site response characteristics of the Anamizu (K-NET ISK005) station and relation to the earthquake disaster during the 2007 Noto-Hanto earthquake, central Japan
On March 25, 2007, the 2007 Noto-Hanto earthquake of MJMA6.9 occurred. We analyzed the strong ground motion records at the Anamizu (K-NET, ISK005) station, where recorded JMA seismic intensity of 6+, to get the ground motion characteristics at the station. Comparing the H/V spectral ratios of the mainshock records and those of records for before events and for aftershocks, non-linear site response behaviour during the mainshock ground motions was observed. As the H/V characteristics were already recovered at the coda wave part of the mainshock records, non-linear site response was restricted only the strong shaking part. From the borehole site information, soft subsurface layer of 10m thickness with low shear wave velocity (approximately 100m/s), would control this non-linear site response. Second, we conducted aftershock and microtremor observations by the temporally-installed stations to understand extension of the area whose subsurface layer is consistent with that at the ISK005 station. Large amplifications of S-wave portion of aftershock records at the ISK005 and the HOS sites were observed by comparing those at the ANC, where is regards as a rock station. As the site specific ratio, amplification of approximately 10 at the frequency of 1-1.5Hz among aftershocks. Single station microtremor H/V ratio also show the similar predominant frequency at Anamizu downtown sites and the low-velocity subsurface layer seems to extend the disaster area during the mainshock. During the 2007 Chuetsu-Oki earthquake of MJMA6.8, occuured at 16th July, similar non-linear subsurface response were observed. Those ground motion characteristics are also discussed. We use K-NET and JMA data. http://sms.dpri.kyoto-u.ac.jp/e- index.html