T14A-01 INVITED
Plate boundary structure controlling megathrust earthquakes in the Nankai subduction zone: results and future perspectives of an active-source seismology
Active-source seismic studies in the Nankai subduction zone have revealed that structures of the plate boundary play a key role for controlling rupture propagations during the 1944 Tonankai and the 1946 Nankai earthquakes. Onshore - offshore active-source seismic surveys crossing the Tokai segment and the Nankai segment revealed a large scale subducted ridge and seamount, which are interpreted as northwestward members of the rear-arc seamount chains on the Izu arc and the Kinan seamount chain on the Shikoku basin, respectively. Those ridge and seamount are imaged at non-rupture areas of the last megathrust earthquakes. I proposed, from those observations, that the large scale convex structures may prevent lateral rupture propagation of the magathrust earthquakes. In the Nankai - Tonankai segment boundary which is clearly shown by an inversion of tsunami data, active-source seismic imaging detected a high seismic velocity body forming a strongly coupled patch at the segment boundary. The numerical simulation incorporating those structural variations explained the historic rupture patterns, and shows the occurrence of a giant earthquake along the entire Nankai trough. Although, the numerical simulation well demonstrated the historic rupture patterns of each segment, the simulated slip distributions within each segment does not coincide with those obtained by observed earthquake and tsunami data. Moreover, the inversions of the earthquake and the tsunami data showed different rupture patters within the Tonankai segment; i.e., in the east of the Shima peninsula both data showed large slips, but beneath the Kumano basin (i.e., in the west of the Shima peninsula) large slips were only obtained by the tsunami data. This indicates that slip having a long time constant may only occurred beneath the Kumano basin where the seismic imaging showed a smooth geometry along the plate boundary. In order to further improve the physical model of the plate boundary which explains the detailed rupture pattern and cycles, new active/passive onshore-offshore seismic profiles densely covered in the entire Nankai seismogenic zone are planned in the next five years. In this presentation, I summarize the result of the active-source seismic surveys since the last a decade and briefly introduce the new seismic project.
T14A-02
Integrating plate coupling and transient slip events along the subduction zone interface in Nankai Trough, SW Japan using GEONET GPS time series
Large earthquakes (M~=8) have occurred repeatedly along the Nankai trough, Southwest Japan, over the past 1000 years. Recently, time-varying slow earthquakes such as short-/long-term slow slip transients, non- volcanic tremor, low/very-low frequency earthquakes have been observed in the region. To understand the interseismic plate loading process and its relation to large earthquakes and slow transient events, we use recently reanalyzed GEONET GPS data to re-examine the plate loading rate and time variable deformation in space and time.In order to estimate interseismic velocities from the GPS time ser,es, we developed tools to identify earthquake and instrument related offsets. We use QOCA to fit an empirical model to the raw time series, estimate and remove seasonal variations and common mode errors across the network. As a result, we were able to determine the accurate horizontal and vertical interseismic velocities for the entire network using 10 years data (1996-2006), as opposed to the short-time velocity estimates in most previous studies. We inverted for the back slip (or slip deficit rate) along the Nankai trough using 3-D velocity field using a realistic interplate fault geometry. To allow for the nonuniform deformation source, we model interplate coupling beneath Shikoku and Kii Peninsula and Nankai-Suruga trough separately. The results show that strong coupling were found at 10-30 km beneath Shikoku and Kii Peninsula while Tokai region has maximum backslip rate of ~2-3cm/yr at the depth of 5-25 km. There is a good spatial correspondence between the inferred strong coupling region with the past large earthquake rupture zones. We applied the extended Network Inversion Filter [MacGuire and Segall, 2003] to model the spatial and temporal distribution of two major slow slip events: 2000-2004.5 Tokai slow slip event and 2002-2004 Bungo Channel event. The cleaned GPS time series provides much better resolution for transient slip analysis. The inverted slip histories show complex patterns with each event consisting of several subevents. Comparison of the plate coupling pattern and transient slip shows that part of strong coupling area beneath western Bungo Channel overlaps with the aseismic slip locus, indicating at least part of the accumulated slip deficit is released by episodic slow slip events in the region. The transient slip zones is located at the downdip portion of strong coupling region but updip of low frequency earthquakes. Excellent temporal coincidence between the resolved slow slip rate and LFE occurrence were found in both events, supporting the hypothesis that they are closely related and possibly reflect the modulation of slow slip on seismicity.
T14A-03
Amplification of Tsunami Heights by Delayed Rupture of Great Earthquake along the Nankai Trough
Great earthquakes (MW>8) have occurred along the Nankai trough with an approximately 100 years interval. The source areas of Tokai, Tonankai and Nankai earthquakes consist of four segments, but the rupture delay on these segments is variable. We investigate the effect of delayed rupture along the Nankai trough on the tsunami heights on the Japanese coast by using tsunami numerical method. Our tsunami simulation is for nonlinear long wave theory, with the grid size of 270 m and time interval of 0.6 s. We used a source model of Annaka et al. (2003), which consists of four segments; N1 is Tokai segment, N2 is Tonankai segment, N3 and N4 are Nankai segment. This model reproduces the heights of the 1707 tsunami traces very well. We varied the delay times of each segment and examined 13 cases. In Cases 1 to 6, N1 and N2 segments rupture simultaneously, followed by N3 and N4 ruptures delayed by 10 to 40 minutes. In Cases 7 to 12, the rupture starts in N2 segment, followed by N1 segment by 10 minutes delay; the delay times of N3 and N4 rupture are the same as those of Cases 1 to 6. To evaluate average amplification of the coastal tsunami heights with respect to the simultaneous rupture case, we use geometric mean value K. The coastal tsunami heights are grouped into four types. The tsunami heights are significantly amplified (K > 1.2) on Type A coasts (the east coast of Kyushu, the west coast of Shikoku) for delayed rupture of N3 and N4 segments, because the tsunami from the delayed ruptures of Nankai (N3 and N4) segments overlaps with the westward-propagating tsunami from Tokai (N1) and Tonankai (N2) segments. The amplification becomes over 1.4 in Case 4 or 10 in which N3 delayed 20 minutes and N4 delayed 30 minutes. The tsunami heights along Type B coasts (the Suruga Bay) are slightly amplified (1.0 < K < 1.2) by the delayed rupture of N1 segment. On Type C coasts (the south and east coasts of Shikoku, the west coast of Kii peninsula), the tsunami heights are almost the same (K ~ 1.0) for the simultaneous and delayed ruptures. Finally, on Type D coast (the south coast of Chubu and the east coast of Kii peninsula), tsunami heights decrease (K < 1.0) for the delayed rupture. However, tsunami heights are locally amplified at places of a bay, for example, Owase and Toba.
T14A-04
Structural and Stratigraphic Setting of the NanTroSEIZE Kumano Transect From 3D Seismic Data
In preparation for IODP drilling, we contracted acquisition of a 3D seismic data set across the NanTroSEIZE transect. The 12 x 56 km 3D box extends from the Kumano basin seaward to the deformation front in the dip direction and extends along strike ~ 4 km northeast and ~ 8 km southwest of the NanTroSEIZE drilling sites. A full 3D pre-stack migration was carried out at IFREE/JAMSTEC in 2007. The 3D volume images the frontal accretionary prism, where the frontal detachment/d'collement is a strong, continuous positive-polarity reflection through most of the frontal region. Thrust packages in the frontal prism are highly variable along strike, due to several oblique ramps. The thrust packages are overlain by slope basin sediments that are progressively more deformed with depth. A megasplay fault rises from the oceanic crust at ~ 10 km depth, cutting across the older part of the accretionary prism all the way to the seafloor in the frontal prism region. This splay fault reflection exhibits areas of reverse polarity, possibly reflecting fluid flow from the seismogenic zone to the surface. The hanging wall is composed of a few hundred meters of Recent to upper Pliocene hemipelagic slope sediments overlying strongly deformed lower Pliocene strata that are believed to have been accreted at the toe of the accretionary prism. The block is truncated at its seaward edge, presumably by a steeply-dipping back thrust. As the hanging wall advances, it is blanketed by hemipelagic slope sediments that slump off steep slopes and into the adjacent trench slope basin. The more consolidated (and thus stronger) hanging wall block then overrides these redeposited slope sediments. The slope sediments in the footwall are 100-400m thick and are faulted and folded as they are overridden. More than 2200m of sediment in the Kumano forearc basin are imaged. The deepest part of the section is strongly folded, while the seaward portion is progressively tilted landward due to repeated motion on the megasplay fault. The forearc basin strata onlap older slope sediments, which in turn overlie an older part of the accretionary prism.
T14A-05
NanTroSEIZE Stage 1: Overview of Results From the Nankai Trough Seismogenic Zone IODP Drilling Project
IODP Expeditions 314-316 took place over five months of continuous drilling operations in the off-Kumano
region of the Nankai Trough subduction zone (September, 2007 – February, 2008). This effort was both the
maiden scientific voyage of the new drilling vessel Chikyu and the first stage in the Nankai Trough
Seismogenic Zone Experiment (NanTroSEIZE), a multi-year effort designed to investigate fault mechanics
and seismogenesis along a subduction megathrust through direct sampling, in situ measurements, and long-
term monitoring. The principal elements of the larger project include (a) the 2006 3D seismic survey of the
intended transect, (b) four planned stages of IODP drilling, sampling, and downhole measurements, and (c)
long-term real-time monitoring of the plate interface in cabled borehole observatories. The three Stage 1
expeditions accomplished drilling, logging, and sampling of 8 sites as deep as 1400 meters below the sea
floor. During Expedition 314, we obtained a comprehensive suite of geophysical logs at several sites along a
transect focused on the up-dip transition from seismic to aseismic fault behavior, using state-of-the-art
logging-while-drilling (LWD) techniques and drilling to depths of 400 to 1400 m. Expeditions 315 and 316
were devoted to coring at these same sites, plus two others. Integrated results of all three expeditions shed
light on the lithologies, physical and hydrogeological properties, structural features, tectonic history,
geochemical and biological systems, and present-day state of stress to 1.5 km below the seafloor.
One highlight of log and cores results is evidence for the orientation of present-day principal stresses from
borehole breakouts, and paleostress from fractures and faults in cores. The orientation of the maximum
horizontal stress axis (SHmax) from breakouts across the outer wedge is consistently perpendicular to the
local strike of major structure, but is subparallel to strike in the forearc basin/inner wedge domain, consistent
with a compressional to transpressional stress state in the outer wedge and an extensional stress state in the
inner wedge. Further results include: (a) contrasting internal structure of recently active faults depending on
their apparent structural, burial, and or diagenetic history, (b) timing and evolution of the major forearc basin
and "mega-splay" thrust system activity, (c) progressive disruption of sedimentary strata with distance into
the wedge, and (d) interaction of syn-deformational slope deposits with recent thrust motion and slope
failure. These first expeditions in NanTroSEIZE set the stage for coming deep drilling efforts to reach the
seismogenic zone, scheduled to begin in March, 2009.
http://www.iodp.org/nantroseize
T14A-06
Lithostratigraphy of the NanTroSEIZE Transect Area in Nankai Trough: A Summary of Results from IODP Expeditions 314, 315, and 316
Stage 1A of the Nankai Trough Seismogenic Zone Experiment achieved success by coring six sites during IODP Expeditions 315 and 316 offshore SW Japan. When combined with LWD data acquired during Expedition 314, clarity emerges in our collective view of the regional lithostratigraphy. At Site C0001, we crossed a pronounced unconformity between a 200-m-thick slope-apron succession and the underlying accretionary prism (hanging wall of mega-splay fault). The lower 10 m of slope cover consists of enigmatic turbidite sand, and the Pliocene accretionary prism below (3.7-4.5 Ma) is unusually muddy and difficult to drill. Coring within Kumano Basin (Site C0002) penetrated more than 1050 m, passing completely through the forearc basin into accreted strata of Miocene age (5.0-5.9 Ma). The contact between the deformed prism and forearc basin is a major unconformity with a hiatus of nearly 1.5 m.y. A condensed section of mud, with local concentrations of glauconite and widespread clay-filled vein structures, rests above the unconformity. Deposition of sand and silt turbidites did not commence until about 1.6 Ma, and the basin filled rapidly thereafter. The frontal thrust system of the accretionary prism was cored at Sites C0006 and C0007. Strata within the footwall are very sandy (with poor recovery), and the hanging wall comprises an upward coarsening and thickening succession of upper Shikoku Basin facies (mud and volcanic ash) and sandy turbidite deposits that may have spilled out of the confines of the bathymetric trench. Additional complications include multiple packets of sand and gravel from migrating channel-levee complexes and displacement of facies units along multiple subsidiary faults above the main frontal thrust. The principal coring targets at Sites C0004 and C0008 were a shallow splay fault and the slope apron. A prominent erosional unconformity and a remarkable set of mass transport deposits (with intraformational mud clasts) highlight the base of slope apron above and seaward of the fault's tip, respectively. Slope failure was probably triggered by periods of intense fault activity on the mega-splay.
T14A-07
Initial results from the Nankai Trough shallow splay and frontal thrust (IODP Expedition 316): Structure and Evolution
Integrated Ocean Drilling Program (IODP) Expedition 316 is the third drilling expedition of the Nankai Trough Seismogenic Zone Experiment (NanTroSEIZE), which is designed to investigate the processes that govern the strength, physical character, and distribution of slip along subduction zone plate boundary fault systems. The first target of Expedition 316 was the shallow portion of the megasplay fault system, just seaward of the break in slope marking the boundary between the inner and outer accretionary wedge. The scientific objectives of drilling the shallow portion of the megasplay fault system are to characterize the slip, deformation, and fluid flow behavior and evolution of the updip region of the megasplay, in particular to assess whether it is an active blind fault or an inactive fault. Two sites were drilled to investigate the shallow region of the megasplay fault system. Site C0004 was selected to cross the splay fault at a depth of ~300 mbsf; the total drilling depth was 400 mbsf to sample underthrust slope basin material. Site C0008 sampled the material of the slope basin 1 km seaward of Site C0004. This site was selected to provide age control on splay fault movement and reference material for sediments being overridden by splay fault movement. Samples of the fault zone sample were successfully recovered and reveal a long term history of activity spanning more than a few hundreds thousand of years. The second target of Expedition 316 was the frontal thrust system. The scientific objectives of drilling this region are to understand the function of the frontal thrust, including its slip and fluid flow behavior, with respect to large earthquakes. An important component of this understanding is to reveal why this particular frontal thrust behaves differently (as evidenced by its longevity and large taper angle) from frontal thrusts in other accretionary margins. The results suggest that the lowest part of the accretionary prism is composed of Plio-Miocene Upper Shikoku Basin sediments and abruptly changes to very coarse underthrust trench-filling deposits accross the narrow plate boundary frontal thrust. Thrust faults within the frontal part of the prism appear to be inactive and are starting to gravitationally collapse due to over-critical state.
T14A-08
Accretionary wedge growth and the limit of the seismogenic zone.
One question asked is whether the updip limit of the seismogenic zone in subduction zones is primarily controlled by present day conditions on the interplate, such as temperature (Hyndmann et al., 197), or by a discontinuity of hanging wall mechanical properties (Byrne et al., 1988), which could results from geological history. Phase 1 Nantroseize drilling already provide evidence that the splay faults at the updp limit of the seismogenic zone are recent structures and do not correspond to a major geological boundary. Structural units recognized all along the Nankai margin are a Quaternary accretionary wedge of 30 to 50 km width and a forearc basin which northern part is underlain by pre-Miocene accretionary complex. This older accretionary complex (part of Shimanto belt) was built during the subduction of the Pacific plate beneath the Japanese Island arc and was heated and affected by diagnesis, metamorphism and igneous intrusions during the Miocene. The amount of material accreted during the early phase of Philippine Sea plate subduction is poorly known. Drilling at Sites C0001 and C0002 constrains the age of accretion between 2 and 3.8 Ma in the splay fault area, and between 3.8 and 5 Ma near the outer edge of the forearc basin. The splay fault system is thus found within the recently accreted part of the margin. Furthermore a simple solid grain balance calculation suggests that a steady state grain flux of 16 to 20 km3/Ma could account for the volumes accreted seaward of C0001 and C0002. This flux is significantly lower than the present day flux into the subduction, which ranges from 27 to 45 km3/Ma assuming a frontal convergence rate of 40 mm/Ma. However,The input flux may have varied in time as a consequence of cyclic arc collision in the Izu area and basement ridge subduction. These are known to affect the growth and tectonic of the Tokai segment, located immediately east of the Kumano (or To-nankai) segment. We wonder whether these processes could influence sedimentary influx, wedge growth and splay fault activation on the Nantroseize cross section.