T24A-01
Temporal Variation of Strain Field Around the Dedana Fault in NE Japan, Before and After the 2008 Iwate-Miyagi Nairiku Earthquake (M7.2)
The Dedana fault (DF) is a southern part of the Kitakami-teichi-seien fault zone (KTSFZ), which consists of
some typical active thrust faults along the eastern margin of Ou backbone range (OBR) in northeastern
Japan. Based on the recent GPS studies, Miura et al. [EPS04] have found a strain concentration zone with
EW contraction along OBR. In order to clarify more detailed strain field around the KTSFZ, we have installed
7 new continuous GPS sites along the EW line crossing the DF to complement the pre-existing sites operated
by the Geographical Survey Institute (GSI) and the National Astronomical Observatory (NAO) in October
2007. Ohzono et al. [JPGU08] have pointed out a possibility of abrupt strain rate increase in OBR bounded
by the DF from the time series of the site coordinates by March 2008.
The Iwate-Miyagi Nairiku (inland) earthquake (M7.2) occurred beneath OBR on 13 June 2008 (UT). Its
hypocenter is located at the southwest about 20 km away from the DF. This earthquake caused the large
coseismic displacement around the source area. Ohta et al. [in review to EPS] proposed a fault model to
explain the coseismic deformation observed by GPS. Soon after the earthquake, additional 14 GPS sites
were deployed by the Japanese University Consortium for GPS (JUNCO) to reveal the postseismic
deformation, which is distributed in the shallower part of the coseismic fault and the DF [Iinuma et al.,
submitted to ASC08].
Examining the time series of the site coordinates before and after the earthquake, we can summarize the
temporal variation of the strain field around the DF as follows:
1) Before the earthquake, 8 months GPS data provide the clear strain increase in OBR around the west of
the DF. However, this may be caused by the long-term strain accumulation as already suggested. We should
investigate the data more carefully.
2) During the earthquake, the DF does not seem to be displaced because the EW profile of the coseismic
displacement cannot be explained by the movement of the DF.
3) On the contrary, the postseismic deformation can be attributed to minor but certain after slip that occurred
on the DF. This idea is supported by looking at two profiles of coseismic and postseismic displacements
showing the different pattern. The 2008 Iwate-Miyagi Nairiku earthquake induced the aseismic displacement
on the DF.
Acknowledgements. This study is based on GPS data provided by the research project conducted by JNES to
establish evaluation techniques of seismogenic faults, and those from continuous sites operated by GSI and
NAO. We thank all people who cooperated the GPS deployment for the temporary network.
T24A-02
Style and Rate of Late Pleistocene – Holocene Deformation of the Poukawa Fault Zone, Central Hawke's Bay, New Zealand
The Poukawa Fault Zone is one component of a complex system of contractional faulting in eastern North Island, New Zealand. It is located within the actively uplifting Hikurangi Margin where the Australian plate meets the Pacific plate at a convergence rate of over 40 mm/yr. The most destructive earthquake in New Zealand history, the 1931 Hawke's Bay earthquake of M 7.8, occurred just off the northern termination of the Poukawa Fault Zone. To the south and probably within the Poukawa Fault Zone, another strong earthquake struck near Waipukurau in 1863. We have characterized the contemporary style of faulting along the zone on the basis of an integrated analysis of a broad spectrum of data, including exploratory trenching; geomorphic data aided by 1m resolution digital orthophotos, a LIDAR-derived Terrain Model, and GPS-RTK surveys; stratigraphic and paleoseismic analysis; radiocarbon and OSL dating and tephra correlation. We have also made a detailed reconstruction of the terrace sequences formed where the Kaikora Stream crosses at a high angle to the Poukawa Fault Zone. These data show that the Poukawa Fault Zone is a contractional fault system formed by a series of NE-SW strands with style varying, from west to east, from high-angle east-dipping reverse to low-angle west-dipping thrusting. The geometry of the system suggests that these faults may merge at shallow depth into a single large structure capable of generating strong earthquakes similar to those that occurred in the past on nearby sections. All these faults variously displace the top of the Ohakean aggradation surface (12-15 ka) thereby generating scarps of several meters. The Kaikora Stream terrace sequences also testify to a series of uplift events associated with the late-Holocene growth of two of the eastern thrust faults. Two reaches of Kaikora Stream show evidence of uplifted and abandoned inset Holocene stream terraces found in association with a surface-rupture trace and an active fold. The four terraces in each case correspond in number with paeloearthquake events recognized in trenches nearby (Kelsey et al. 1998). Based on these relations the recurrence interval of surface faulting and folding is c. 3000-3700 yr. The abandonment of a low inset terrace capped by peat and Waimihia Tephra (c. 3400 yr BP) is consistent with this average recurrence. Based on the deformation of the dated Ohakean surface across the entire Poukawa Fault Zone, its reverse slip rate is c. 1-2 mm/yr.
T24A-03
The 2008/01/09-22 Ms=6.4-5.4 earthquakes in southern central Tibet from InSAR observations and numerical modeling
A sequence of three earthquakes of Ms=6.4, 5.9, and 5.4 occurred within two weeks in January 2008 near the center of south Tibet (85.3°E and 32.3°N). We use interferometric Synthetic Aperture Radar (InSAR) observations to study the fault structure and coseismic displacement of these events. Two pairs of ascending (Track 341 and Track 427) and one pair of descending (Track 348) Envisat data covering the three events are processed using the ROI-PAC software. The observed coseismic deformation shows that these events occurred on two approximately parallel normal faults trending ~25-30° east of north and dipping to the west. We modeled these events using two planar faults embedded in an elastic half-space and divided into 10x10 and 5x5 ~4.5 km- and 3 km-wide patches, respectively. The variable slip solution is obtained by least-square inversion of the observed displacement in the three interferograms while exploring a range of possible geometry and fault dip angles. The minimum misfit between the modeled displacement and InSAR data is obtained with the main rupture fault dipping ~65° and the secondary fault ~70°, both to the west. The coseismic displacement concentrates in the center of the main fault in the 2-16 km depth range with a maximum displacement of ~1.1 m. This slip distribution corresponds to a moment release of ~5.51 1018 Nm, consistent with the seismic moment released by the largest event on January 9. The second fault is located in the hanging wall of the main fault, ~8.5 km away to the west. Coseismic displacement reaches ~0.55 m on this fault, within a depth range of 1-6 km. This slip corresponds to a geodetic moment of ~8.44 1017 Nm, close to the seismic moment released by the January 16 event. The two ruptures could be responsible for the three main earthquakes of the January 2008 cluster, although the location of the January 22 event could not be identified in the radar data depicting the collective contributions of the three events. These observations suggest that active normal faulting in southern Tibet bears relatively high dipping angles and distributes mainly within the upper crust. To investigate the mechanical relationship between the two ruptures, we develop 3-D finite element models to compute the Coulomb stress change on the secondary fault after the main fault event. We explore both purely elastic and poroelastic solutions. Model results show that the Coulomb stress increased at shallow depth in the hanging wall of the main fault after the first event, favoring slip on the secondary fault plane. The poroelastic solution results in larger Coulomb stress changes in the early post-seismic period compared to the purely elastic solution, suggesting that pore fluids in the crust may play an important role in triggering earthquakes at shallow depth.
T24A-04
Modelling the Deformation Front of a Fold-Thrust Belt: the Effect of an Upper Detachment Horizon
Structures found at the deformation fronts of fold-thrust belts are variable in type, geometry and spatial organisation, as can be demonstrated from comparisons between structures in the Zagros Fold-Thrust Belt, Iran and the Sawtooth Range, Montana. A range of influencing factors has been suggested to account for this variation, including the mechanical properties and distribution of any detachment horizons within the cover rock succession. A series of analogue models was designed to test this hypothesis, under conditions scaled to represent the Sawtooth Range, Montana. A brittle sand pack, containing an upper ductile layer with variable geometry, was shortened above a ductile base and the evolution of the deformation front was monitored throughout the deformation using a high-accuracy laser scanner. In none of the experiments did the upper detachment horizon cover the entire model. In experiments where it pinched out perpendicular to the shortening direction, a triangle zone was formed when the deformation front reached the pinch out. This situation is analogous to the Teton Canyon region structures in the Sawtooth Range, Montana, where the Cretaceous Colorado Shale unit pinches out at the deformation front, favouring the development of a triangle zone in this region. When the pinch out was oblique to the shortening direction, a more complex series of structures was formed. However, when shortening stopped before the detachment pinch out was reached, the deformation front structures were foreland-propagating and no triangle zone was observed. This situation is analogous to foreland-propagating thrust structures developed at the deformation front in the Swift Dam region of the Sawtooth Range, Montana and to the development of fault-bend folds at the deformation front of the Zagros Fold-Thrust Belt, Iran. We suggest that the presence of a suitable intermediate detachment horizon within a sediment pile can be invoked as a valid explanation for the development of varied deformation front structures in fold-thrust belts. Specifically, the spatial extent of the upper detachment horizon with respect to the spatial extent of the deformed region is a key influence on the development of deformation front structures. However, we acknowledge that factors such as basement structure and variable sedimentation within the foreland basin may also be key influences on deformation front structures in other fold-thrust belts.
T24A-05
Coseismic Pit Crater, Normal Fault, and Extensional Fissure Formation in Unconsolidated Sediment and Basalt in Northern Iceland
Two rifting-related seismic events in 1975 and 1978 along the Mid-Atlantic Ridge near the northern coast of Iceland produced an array of surface deformation features in Holocene basalt flows and overlying unconsolidated sediments. New field mapping and aerial photograph interpretation is coupled with analysis of maps of seismic activity and level-line survey results to constrain the timing, style, and magnitude of this deformation. Fault scarps and fissures in basalts can be traced laterally down a gentle northward dip projecting into unconsolidated braided stream deposits, providing an impressive view of the deformation style in the two contrasting mechanical layers. We report on detailed field mapping of two of these laterally traceable structures conducted in the summer of 2008 and analysis of a suite of aerial photographs from 1958 to 1998. Map-scale structures in the basalts with little or no sedimentary cover include (i) fault scarps, (ii) fissures, and (iii) locally-developed gentle dip away from the related normal fault. Dilation of faults and extension fractures in the basalt has led to rock toppling and rock fall causing widening of fissures. Wedging of toppled rock blocks at the tops of fissures has locally produced keystone arches and bridges across the tops of open fissures. Different stages in the progression of fissure formation and collapse, including (i) fissure, (ii) widened fissure with cavern, (iii) localized collapse pit, and (iv) elongate collapsed fissure, can be observed over along-strike distances of 10's of meters. Where unconsolidated sand and gravel deposits >3 m thick cover the basalts (200 m to the north along strike) structural geomorphologic features are dominated by (i) grabens, (ii) pit craters, and (iii) elongate troughs. Graben-bounding normal faults cutting the sedimentary cover in many cases have displacements >1 m. Pit craters have cone to bowl shapes, commonly occur within grabens, and have depths up to 2.8 m. The mapped surface structures in the braided stream deposits formed during the 1975 and 1978 rifting events and were likely triggered by reactivation of faults and fissures in the underlying basalt. Fissures, caverns, pits, and troughs in the basalt and braided stream deposits are partially water filled, extending below the local water table. We interpret that pit craters formed in basalt by collapse into caverns formed by fissure-widening rock fall, and in sediments by draining of unconsolidated material downward into fault or fissure-related voids in the underlying basalt. In sediments, troughs appear to be produced by a combination of graben formation and pit crater amalgamation. A sequence of development from pit crater chain (or pit chain within graben) to trough to fissure occurs with increasing dilational fault displacement or fissure width, and as a function of sedimentary cover thickness. This field based study demonstrates that pit craters are readily explained by draining or falling of poorly consolidated material downward into subterranean cavities produced by coseismic fault and extension-fracture dilation in underlying cohesive material (basalt). Directly analogous geomorphologic patterns on Mars, clearly visible in high resolution surface imagery, suggest that similar mechanisms of deformation and surface collapse may be at work on Mars.
T24A-06
Geologic Insights From 3D Mechanical Models of a Field Outcrop Extensional System, Volcanic Tablelands, Bishop, CA
The Volcanic Tablelands, north of Bishop California, provide a world-class outcrop exposure of an extensional fault system and associated deformation. The welded upper unit of the Bishop Tuff provides a relatively unweathered surface which records throw on hundreds of fault scarps and associated deformation, accumulated over the past 750 ka. We use differential GPS, digital orthophotos, airborne LiDAR, and coarser resolution digital elevation models to constrain 3D mechanical models of faulting and associated deformation in the Volcanic Tablelands. Elastic dislocation models are used to constrain the subsurface geometry of faults observable only as surface scarps by minimizing the deviation of model surface displacement from surface displacement documented by the DEM. We address the issue of listric versus planar fault geometry as well as considering fault dip and the depth of faulting. Model predicted stress and strain distributions are compared with field observations of smaller scale faulting in order to assess our ability to predict off-fault deformation with mechanical models of large-scale deformation. Retrodeformational models are considered in addressing this question as well.
T24A-07
Recent Earthquake Breaks At The Sea of Marmara Pull-apart (North Anatolian Fault)
The North Anatolian Fault (NAF) makes a major transtensional step-over in the west which forms the lithospheric scale Sea of Marmara pull-apart, between the strike-slip Ganos and Izmit faults. Smaller strike- slip segments and pull-apart basins alternate within the main step-over, combining strike-slip and normal faulting. During the MARMARASCARPS cruise clear morphologic evidence of recent faulting activity was found along several segments of the NAF in the Sea of Marmara. Sets of well-preserved earthquake scarps extend offshore from the Ganos and Izmit faults on land. Our observations from visual exploration and ultra- high resolution bathymetry data (microbathymetry) suggest that those scarps correspond to the submarine ruptures of the purely strike-slip 1999 Izmit (Mw 7.4) and the 1912 Ganos (Ms 7.4) earthquakes. One break extends offshore eastward of the Ganos fault and cuts continuously the Tekirdag basin and Western High up to the Central basin over 60 km. Scarps, here, are very well preserved and show fine-scale morphology typical of strike-slip faulting. The age of the last earthquake break is difficult to assess directly with dating approaches. However, recent sedimentation rates can provide information on the age of the sediment covering the scarps. With that purpose, ROV (remote operated vehicle) sampled interface cores (up to 35 cm) into the disturbed sediment in the immediate vicinity of those scarps. Our first geochronological analysis with 210Pb seems to confirm the connection of fresh fault scarps to the 1912 Ganos earthquake rupture. Sedimentation rates determined from 210Pb profiles (excluding disturbed layers) on cores show a narrow range between 0.1-0.2 cm/yr. Another very fresh break is seen where the Izmit fault enters westward into the Cinarcik Basin. It crosses the bottom of a submarine canyon at 180 m depth, 10 km west of the Hersek peninsula. Microbathymetry suggests the 1999 fault scarp is there 0.5 m high. The flat floor of the canyon appears to result from leveling by significant sediment transport. So, on the average the sedimentation rate must be low. Under such conditions, only the last earthquake break can be preserved across the canyon floor. The break continues for some kilometers to the west and appears to end at the junction with Cinarcik basin normal faulting. It corresponds probably to the western end of the 1999 Izmit earthquake rupture and is consistent with an underwater rupture extension of 20-30 km westward as inferred from SAR interferometry. The direct observations of submarine scarps in the Sea of Marmara are critical to define barriers that have arrested past earthquakes. Incorporating the submarine scarp evidence modifies substantially our understanding of the current state of loading along the NAF next to Istanbul.
T24A-08
Paleoseismological analysis in Tehran region (Central Alborz, Iran)
The North Tehran, Taleghan and Mosha faults are three major active faults menacing the 15 millions peoples
leaving in Tehran metropolis and its suburbs areas. These three faults located at the southern piedmont of
Central Alborz and have been described as the sources of several large historical earthquakes in the past.
To assess the seismic hazard associated with these faults, we carried paleoseismological studies.
The North Tehran fault: Our study shows that the fault extends up to 110 km and corresponds to a reverse
fault associated with a left-lateral component within its north-western part. This fault zone is also
characterized by secondary active fold-and-thrust structures affecting the alluvial deposits within Tehran itself
(e.g. Milad Tower foreberg). Between Tehran and Karaj, where the fault trend changes from NE-SW
(eastwards) to NW-SE (westwards), we found a ~ 3 m fault scarp affecting the Pleistocene-Holocene
deposits. Trenching across the scarp showed a N 115° E trending 30° N dipping reverse fault.
We found evidences for 8 events (Mw > 6.5) during the past ~30000 years yielding a [3200- 4100
yrs] mean return period. The shortening rate across the fault is ~ 0.25 mm/yr during the Late
Pleistocene – Holocene.
The Taleghan fault: So far described as a south-dipping reverse fault, our study shows that the Taleghan
fault is not a reverse fault but a left-lateral strike-slip fault with a normal component. Its strike, dip and rake
within its eastern part are 105, 60° and -20/40°, respectively. Our paleoseismological analysis
shows that 2 (maybe 3) events with magnitudes Mw ≥ 7 occurred during the past ~ 3500
years. The recurrence interval for earthquakes is comprised between ~1200 and ~1800 years and
the horizontal slip rate is ~ 1.5 mm/yr.
The Mosha fault: As for the Taleghan fault, we found many evidences at different scales, of left-lateral strike
slip movements associated with a small normal component showing that the Mosha active fault is mainly a
left-lateral strike-slip fault, and not a revere fault as previously described Our paleoseismological
investigations allowed us to determine a minimum slip rate of 2.2 ± 0.5 mm/yr along the eastern part of
the Mosha fault. Along this segment, our analyses within several trenches brought evidences for several
seismic ground ruptures having occurred during the past ~10000 years - including probably one among
the two historical earthquakes having occurred in 1665 AD and 1830 AD. Combining stratigraphical and
kinematics evidences allowed us to conclude that these ruptures were caused by seismic events with
Magnitude Mw >7. Using radiocarbon and optically stimulated luminescence dates, we estimated the return
periods for these large events to be comprised between 1200 and 1600 years.