T31G-01 INVITED
16 July 2007 Niigata-ken Chuetsu-Oki earthquake; its characteristics, tectonic background and significance for active fault evaluation
On July 16, 2007, the Mw 6.6 (MJMA 6.8) Niigata-ken Chuetsu-Oki earthquake occurred along the Japan Sea coast south of Sado Island. Kashiwazaki-Kariwa nuclear power plant, situated at the eastern margin of the source area of this earthquake, recorded significantly severe ground motion, which was more than twice, at the maximum, as strong as the designed acceleration in the E-W direction at the foundation base mat level of 3 nuclear power units. Exceedance of the designed ground motion by the Chuetsu-Oki earthquake became the center of public attention. The epicenter of the 2007 Chuetsu-Oki earthquake is located about 30 km northwest of the epicentral area of the 2004 Chuetsu earthquake, and both of the two earthquakes were caused by reverse faulting. Although the source of the 2004 Chuetsu earthquake has been identified as an NW-dipping reverse fault, the dip direction of the source fault of the 2007 Chuetsu-Oki earthquake, SE or NW, has not been settled yet as of September 3, 2007. The 2004 Chuetsu and 2007 Chuetsu-Oki earthquakes occurred in the Strain Concentration Zone characterized by shortening tectonics with E-W- to NW-SE-trending compressive axis. The Strain Concentration Zone traces its origin back to the formation of Japan Sea under E-W extensional tectonics ca 20 Ma. Along the eastern margin of the Japan Sea, grabens or depositional basins were formed in association with the opening of the Japan Sea, being filled with ~5-km-thick Neogene-Quaternary volcano-sedimentary sequences. Around 3 Ma, the eastern marginal area of the Japan Sea changed into a compressional tectonic field, and reverse faulting and folding have been prevailing in the area ever since. The 2004 Chuetsu earthquake was produced by reverse faulting on a northern segment of the known Muikamachi fault system, which probably had been formed as the eastern boundary fault of the Shin'etsu depositional basin. For the 2007 Chuetsu-Oki earthquake, two possibilities exist regarding its relation to known faults. If the source fault dips southeast, a deep part of a 20-km-long submarine active fault, which had possibly been formed as a western border fault of the Shin'etsu basin, should have caused the earthquake. If the source fault dips northwest, on the other hand, west-dipping reverse faults in the coastal hilly area might have relation to the recent earthquake, although there is no known active fault that exactly corresponds to the location and extent of the 2007 earthquake source area. This presentation will brief the latest situation of the controversy on dip direction and its implication to active fault evaluation for seismic safety of nuclear power plants, based on the newest results of post-earthquake researches by related institutions.
T31G-02
Coastal Deformation Sssociated with the 2007 Noto Hanto Earthquake, Central Japan, Determined by Uplifted and Subsided Intertidal Organisms
The March 25, 2007 Noto Hanto earthquake (Mj 6.9, Mw 6.7) caused a vertical crustal movement along the northwestern coast of the Noto Peninsula, central Japan. We promptly determined pattern and amount of coseismic coastal movement by using uplifted and subsided intertidal sessile organisms. Our observations show a broad 20-km-wide asymmetric surficial deformation above and across the south-dipping source fault, with steep north-facing frontal limb and gentle south-facing back limb. The maximum uplift was ~40 cm at the crest. Our forward modeling suggests that the south-dipping source fault is buried at a depth of ~2 km and that 1.2 m slip on that fault yields best fit to our surface observation. Our results show that traditional field mapping should be combined with modern instrumental observations such as GPS and InSAR for the most effective and reliable estimate of spatio-temporal crustal movement associated with large earthquakes.
T31G-03
Tsunami source of the unusual AD 869 earthquake off Miyagi, Japan, inferred from tsunami deposits and numerical simulation of inundation
The 869 Jogan earthquake, off Miyagi, produced unusually large tsunamis, according to a historical document and tsunami deposits. One of the oldest official documents in Japan reported that about 1,000 people were drowned from the tsunami in Sendai plain, indicating much larger tsunami than the 1896 Sanriku tsunami (the worst tsunami disaster in Japan caused by a tsunami earthquakes) or the 1933 Sanriku tsunami (caused by the outer-rise normal fault event). Our systematic field surveys revealed the distribution of tsunami deposits in Sendai and Ishinomaki plains. In both plains, the 869 tsunami deposits are identified as sand layers just below the regional tephra (To-A from Towada volcano in AD 915). In Sendai plain, the tsunami deposits extend about 1 to 3 km from the coast line at that time, which is estimated as about 1 km inland of the present coast. In Ishinomaki plain, the tsunami deposits extend > 3 km from the estimated coast line, which is about 1-1.5 km inland of the present coast. Multiple sand layers indicate recurrence of such unusual tsunamis with approximately 1,000 yr interval. We computed tsunami inundation in both plains from several types of tsunami source models such as outer-rise normal fault, tsunami earthquakes (narrow fault near trench axis), interplate earthquakes with fault widths of 50 and 100 km. Comparison of the computed inundation area with the distribution of tsunami deposits indicates that only an interplate earthquake source with 100 km width (depth range of 20 to 50 km) can reproduce the observed distribution of tsunami deposits in both Sendai and Ishinomaki plains. This source (Mw=8.1 to 8.3) is much larger than the anticipated Miyagi-oki earthquake (M~~7.5) with 99% probability in the next 30 years.
T31G-04
Numerical simulation of the 1771 Yaeyama tsunami in the southwestern Ryukyu arc
Numerical simulation of 1771 Yaeyama tsunami was computed and showed that M7-class earthquake would cause very high run-up locally without landslide. The 1771 Yaeyama tsunami occurred on April 24, 1771 in southwestern Ryukyu area, Japan. Maximum run-up heights reached over 30 m. About 12000 people were dead by the tsunami. Previous tsunami source models for this event have included both seismological faults along the Ryukyu Trench and submarine landslides. However, no evidence of landslides in the source has been obtained, despite marine surveying of the area. The fault along the Ryukyu trench also cannot have reproduced the distribution of run-up heights. Recently, CMT solutions of earthquakes and GPS measurement showed that the Ryukyu arc was extending parallel to the Ryukyu Trench. This suggests that the normal faulting cutting the island arc would be possible source for tsunami. Although the arc-cutting normal faults had been observed previously, they had not been examined as a source of the tsunami. Then the numerical tsunami simulation was employed and the possibility of the source of the 1771 Yaeyama tsunami was investigated. The tsunami propagation was computed using nonlinear shallow-water equations. The area for computation extended from 23.7N to 25.5N, and from 123.7E to 125.75E. The bathymetry and topography data were gridded at 75 m intervals over the entire area, and a finer grid interval of 25 m was employed for six areas for detailed analysis. The time interval for computation was 0.1 s. The model was computed for a wave travel time of two hours. Assumed fault length is 50 km (Mw7.6). As a result, the arc-cutting normal fault model proposed in this study successfully reproduced the distribution of tsunami run-up. Since the fault is located at the shallow area (about 300m in water depth), the tsunami waves refract at the shelf and concentrate locally to coast of islands. The arc-cutting normal fault and focusing effect would have caused very high run-up height in this area.
T31G-05
Implications of river morphology response to Dien Bien Phu fault in NW Vietnam
In northern Vietnam, most rivers are flowing southeastward sub- or parallel to the valley of Red River and characterized by long but narrow catchments. The Dien Bien Phu fault is associated with the most seismically active zone in Vietnam and situated in the potential eastern boundary of the rotating southeastern Tibetan block. It cuts the Da River, the largest tributary of Red River in northwest Vietnam and has distorted the drainage basin resulting in complex river patterns. To assess the river morphology response to active Dien Bien Phu fault, we use 1/50,000 topographic data and ASTER images to map the precise river courses and digital elevation model data of SRTM to retrieve and analyze the river profiles. From the mapping results, the N-S striking fault results in three conspicuous north-trending river valleys coincided with the different fault segments to facilitate the measurement and reconstruction of the offsets along the fault. Further combining the longitudinal profile analysis we obtain ca. 10 km offsets by deflected river as the largest left-lateral displacement recorded along the active fault. The restored results show the downstream paleochannel of the Da River had been abandoned and becomes two small tributaries in opposite flow directions at present due to differential crustal uplift. Also the present crisscross valley at the junction of the Da River and the fault is resulted from the capture by another river which has been also deflected by the neotectonics. Based on our observations on river response, the Dien Bien Phu fault is a sinistral dominant fault with an uplift occurring in its eastern block. Furthermore the active Dien Bien Phu fault does not cut through the Red River northward indicating the western block of the fault can not be regarded as a single rigid block. There should be possible to find NW-SE trending faults paralleling to Red River to accommodate the deformation of the western block of the fault.
T31G-06
Docked Or Accreted Indian Ocean Fracture Ridges Along The Sumatra Subduction Zone Northern Tip
Detailed multibeam mapping coupled with echo sounder data collected offshore the northern tip of Sumatra over the rupture area of the Dec. 26 2004 Mw 9.3 earthquake during the Sumatra aftershock cruise reveal dominant sub-meridian dextral wrenching at the western termination of the Sunda subduction zone. N10°W trending dextral wrench faults with both west and east vergency affect the wedge that overlies in this area the northern prolongation of the NS trending oceanic fracture zone ridges system that absorbs the Indian/Australian plates relative motion between the 90° E and 92°E ridges. Three main N10°W dextral wrenched fault zones with a discrete westward vergency were continuously mapped on the basis of bathymetry and low frequency sounder profiles over 150km across the wedge. The Upper Splay Fault located immediately west of the Inner Ridge (backstop or innermost wedge?) extends from south of Simelue/Nias Island to the vicinity of the Nicobar Islands to the north. The N/S Median Splay Fault within the core of the wedge is the site of the most intense strike-slip deformation evidenced by flower structures and micro-seismicity (OBS). Piggy-back basins are dextrally wrenched as shown by echo sounder data. The westernmost half of the wedge is dominated by NNW elongated pop-up structures with both west and east verging thrusts. Many of these are interpreted by us as oceanic ridges deformation in progress. We propose that the structural fabric of the wedge is due to the interaction of the 90°-92°E oceanic fracture ridges system with the Sumatra backstop and inner wedge. The ridges deformed at depth by NS left-lateral strike-slip faulting due to the distributed Indian/Australian motion dominate the fabric of the hyper- oblique convergent zone. As a consequence, if the ridges are not yet incorporated within the wedge, as we believe, the inter-plate boundary that is well defined by after-slip seismicity near 40-50km depth below the Aceh basin and is known to extend westward with a gentle dip under the upper wedge, does not extend below the outer oceanic fractures system that forms the backbone of the lower wedge. It would have to steepen upward to merge into the MSF in the median part of the wedge. Thus, in our preferred interpretation, the inter-plate boundary does not extend below the lower part of the wedge that is essentially constituted by deformed but still not accreted oceanic fracture volcanic ridges and the steep dip of the decollement east of the MSF below the shallow upper wedge may have played a major role in the generation of the Sumatra Tsunami. Alternatively one would have to assume that the ridges are already incorporated within the wedge which we have at this stage no evidence for. We show that this type of interaction of the submeridian Indian ocean fracture ridges with the subduction zone extends far north into Myanmar and Bangladesh along the NS trending Andaman trench.
T31G-07
Structure and Kinematics of the Indo-Burmese Wedge
The Burma subduction trench and the associated Indo Burmese wedge mark the present eastern boundary of the Indian plate in the northern Bengal area. The initiation, duration and history of the Bengal crust subduction beneath Burma is still debated. The aim of this paper is to provide a structural and kinematic analysis of the Indo- Burmese wedge in order to better constraints the Bengal crust subduction history beneath Burma. On the basis of field observations, seismic reflection data interpretation and well logs data we present a structural analysis of the Outer Indo-Burmese Wedge. We also constrain the onset of this Outer Wedge to be younger than 2Ma, implying a recent and fast westward growth (~10cm/yr) since Late Pliocene in close relationship with the onset of the Shillong plateau. Restoration process of a synthetic cross section through the Outer Wedge allowed us to estimate the amount of EW shortening accommodated in the Outer Wedge to be 5.1mm/yr since 2Ma. These results combined with previous available GPS data from central Myanmar suggest strain partitioning at wedge scale. The core of the wedge is affected by shear deformation and acts as a buttress for a frontal wedge that accommodates a more compressive strain component. Finally we propose that the main characteristic of the Indo-Burmese wedge growth mechanism is the progressive incorporation of the most internal part of the wedge, formerly affected by transpressive thin-skinned tectonics, to the buttress where they are subsequently affected by shear deformation. The crustal structure boarding the newly formed buttress seems to be guided by the subducting crust fabrics. We are in favour of a very recent (Late Miocene) onset of the present Indian crust subduction beneath Burma coeval with the global plate kinematics reorganisation related to the Indian/Australian plate spliting. This subduction postdates the Indo Burmese range onset that must have started in early Miocene. This range first began to uplift because of shear deformation at the India/Burma plate boundary in a highly oblique strain context. The subduction may have started when a lower obliquity was reached after the Late Miocene plate kinematics reorganisation.