- Email: [email protected]
Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan
Quaternary International xxx (2017) 1e11
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan Hideaki Goto Department of Geography, Hiroshima University, 1-2-3, Kagamiyama, Higashi-Hiroshima, Hiroshima, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 October 2016 Received in revised form 20 November 2017 Accepted 6 December 2017 Available online xxx
The Median Tectonic Line active fault zone (MTLAFZ) extends for about 190 km through Shikoku, southwest Japan. Though the MTLAFZ is the most significant onshore active tectonic feature in southwest Japan, its late Quaternary slip rate has been estimated at only a few locations with reasonable references and ages. Better information on this feature's recent slip rates is critical to understanding the ongoing tectonic processes in the region and evaluating the seismic risk of this fault zone. In this paper, new estimates of the late Quaternary slip rate are reported from the Ikeda and Chichio faults in the central portion of the MTLAFZ. The author mapped late Pleistocene fluvial terrace surfaces and used tephrochronology and radiocarbon dating to constrain the age, and measured offset of terrace risers. The slip vectors of both faults are similar, as derived from piercing points on the bottoms of the terrace risers. The vertical component of displacement is 2e6% of the horizontal component. Long-term slip rates during the late Quaternary were calculated at 7.8e9.1 mm/yr, which is more precise than those in the previous studies and represents the highest slip rate in the MTLAFZ. This rate is also much faster than previously reported shorter-term slip rates of geodetic study. The long-term seismic risk of large earthquakes (>M8) along the Ikeda and Chichio faults in the next 30 years are evaluated to be more than 0.4e1.9% and 0.1 e3%, respectively, much higher than a previous estimate of 0e0.3%. © 2017 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Active fault Geomorphology Slip rate Digital elevation model Median tectonic line Southwest Japan
1. Introduction The Median Tectonic Line (MTL) is the most prominent geological boundary in southwest Japan (Fig. 1A). The Median Tectonic Line active fault zone (MTLAFZ) extends along the MTL for about 190 km through Shikoku (Research Group for Active Faults of Japan, 1991, Fig. 1). The MTLAFZ is an arc-parallel, right-lateral strike-slip fault related to the oblique subduction of the Philippine Sea plate beneath the Eurasian plate along the Nankai trough (Fitch, 1972), suggesting that the MTLAFZ is the most significant onshore active tectonic feature in southwest Japan. Though late Quaternary slip rate has been estimated at a few locations, these studies were conducted in 1960e90s based on the amount of offset measured on topographic map with a contour interval of 5e10 m. Improved constraints on the slip rate of this feature are critical to assessing the seismic hazard along this fault zone and to understanding the ongoing tectonic processes in the region. In this paper, new late Quaternary slip rate estimates are
E-mail address: [email protected].
reported for the Ikeda and Chichio faults along the central portion of the MTLAFZ in Shikoku, Japan. Towards this end, the author mapped fault traces and measured offsets within late Pleistocene fluvial terraces and dated terrace deposits using tephrochronology reported in the prior study and radiocarbon ages. Slip vectors and offset were measured from topographic profiles and maps derived from digital elevation models (DEM) developed from processing of aerial imagery taken in the 1975, before the construction of an expressway near the active fault traces.
2. Regional setting The MTL separates the low-pressure, high-temperature Ryoke metamorphic belt to the north and the high-pressure, low-temperature Sambagawa metamorphic belt to the south, along a line running from Shikoku to Kanto (Fig. 1A) (Hashimoto and Kanmera, 1991; Takahashi, 2006). The MTL is offset along the ItoigawaShizuoka Tectonic line (ISTL) in Chubu, suggesting that the MTL is older than the ISTL. The Upper Cretaceous Izumi group, composed of interbedded sandstone and mudstone, was deposited in the narrow half-graben to the north of the MTL, related to left-lateral
https://doi.org/10.1016/j.quaint.2017.12.013 1040-6182/© 2017 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
Fig. 1. A: Distribution of active faults and major tectonic lines in central Japan. The shaded relief map uses a 1-km grid DEM from SRTMplus Data provided by the USGS. White and red lines indicate active fault lines and tectonic lines, respectively, after Nakata and Imaizumi (2002) and Hashimoto and Kanmera (1991). The rectangle of dashed white line and purple line indicate location of Fig. 1B and C, respectively. Abbreviations are as follows: MTL, Median Tectonic Line; ISTL, Itoigawa-Shizuoka Tectonic Line. B: Active faults (red lines) along the MTL within the Shikoku region. Topographic map is after Okayama (1988). The stars with abbreviation indicates where the prior studies evaluated the long-term slip rates. Abbreviations show the site where the studies are conducted as follows: a: Goto (1996); b: Tsutsumi et al. (1991); c: Okada (1968); d: Okada (1970); e: Tsutsumi and Okada (1996). Abbreviations of the faults are as follows: I.F., Iyo fault; S.F., Shigenobu fault; Kw.F., Kawakami fault; O.F., Okamura fault; Iz.F., Ishizuchi fault; Ik.F, Ikeda fault; M.F., Mino fault; C.F., Chichio fault; Z.F., Zunden fault; It.F., Itano fault; N.F., Naruto fault. B: Enlarged geomorphological map of area C in Fig. 1A. The shaded relief map uses a 90-m grid DEM from SRTM Data provided by the USGS. Red lines indicate active fault lines, after Nakata and Imaizumi (2002). Abbreviations of the faults are same as Fig. 1B.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
faulting from the western part of the Kii Peninsula to Shikoku (Hashimoto and Kanmera, 1991). Evidence for late Quaternary faulting along the MTL was reported by Kaneko (1966) and Okada (1980), who observed fault scarps cutting fluvial terraces, and offset streams and terrace risers from the western part of Kii peninsula to Shikoku (Fig. 1A). The late Quaternary movement along the MTL has been found to be predominantly right-lateral strike-slip, with a vertical displacement less than one-tenth of horizontal displacement (Okada, 1980). The MTLAFZ is composed of several active faults, up to several tens of kilometers long and parallel to the MTL. Previous studies have estimated the long-term horizontal slip rate of the active faults along the MTLAFZ, based on the offset of channels that incised terraces and risers, as 1e3 mm/yr in Kii Peninsula (Okada and Sangawa, 1978), 5e10 mm/yr in the eastern to central part of Shikoku (Okada, 1968, 1970; Tsutsumi et al., 1991; Okada and Tsutsumi, 1997), and 1e2 mm/yr in the western part of Shikoku (Goto, 1996) (Fig. 1B). The slip rate in the eastern to central part of Shikoku is the highest, but is not well-constrained, because a few geomorphological studies was conducted in 1960e90s and the ages for the terrace surface of the offset references have not been clear. This study is focused on the Ikeda and Chichio faults which are the most prominent faults in eastern Shikoku (Fig. 1C). The recurrence intervals of faulting events in Shikoku during the late Holocene were revealed by paleoseismological studies to be about >1000 years (Okada and Tsutsumi, 1997; Morino and Okada, 2002). The most recent event was in Shikoku, concentrated at the end of the Middle ages in Japanese history (12-16th century) (Goto et al., 2001), probably corresponding to the 1596 Keicho earthquake recorded in historical documents from the Kinki district and Shikoku (Tsutsumi et al., 2000). Long-term seismic risk from this fault hosting a large earthquake (>M8) in the next 30 years is estimated to be 0e5% in Kii and 0e0.3% in Shikoku (Headquarters for Earthquake Research Promotion, 2011). Though the timing and slip amount of surface rupture associated with the most recent event are revealed in Shikoku (Goto et al., 2001; Tsutsumi and Goto, 2006), the long-term seismic risks are uncertain mainly because of unclear recurrence intervals (Morino and Okada, 2002). As the slip amount of the most recent surface faulting were detected in the previous studies, determining the long-term slip rate accurately is crucial for evaluating the earthquake risks along this fault zone. Based on the analysis of densely distributed campaign GPS measurements and block modeling results of Tabei et al. (2002), 5 mm/yr of contemporary dextral shear occurs across the MTL, when the fault dip angle is 35 and locked at 15 km from the earth surface. On the other hand, Aoki and Scholz (2003) analyzed the interplate coupling coefficient as the function of depth only, and the MTLAFZ slip rate was estimated to be quite variable, between 0.00 and 5.50 mm/yr assuming a vertical fault, and between 0.00 and 3.88 mm/yr assuming a 35 north-dipping fault. These studies have problems that the dip angle of the MTL and the MTLAFZ is uncertain
3
at depth, and the value of geodetic surveys may not be representative of the deformation within the intraplate zone because of the shortness of measuring periods. Thus, a long-term slip rate would be useful in providing insight to understanding the deformation through the Nankai trough to the intraplate region of southwest Japan. Two sites were selected along the Ikeda and Chichio faults where faults cut through several terraces, and laterally offset terrace risers. One is the Ashiro site on the Ikeda fault, and the other is the Kamigirai site on the Chichio fault (Fig. 1B and C). In the Kamigirai site, Okada and Tsutsumi (1997) discussed the slip rate based on the offset of Holocene terraces, and estimated that the slip rate was 6e7 mm/yr. However, the age of the terraces was determined only based on the height difference between the upper terrace developed during the last glacial age and alluvial plain near the fault scarps.
3. Material and methods To identify topographic features related to active faults, the author analyzed aerial photograph from the Geospatial Information Authority of Japan (GSI), taken in the 1960s and 1970s with the scale of 1: 8,000 to 1: 20,000. The author conducted the field survey to check the correlation of the terrace surfaces and observed the deposit of the fluvial terrace. The author also collected a wood sample buried in the terrace deposit for radiocarbon dating (Table 1). The ages of terrace formation were determined from radiocarbon age dating as well as from previous studies using tephrochronology (Mizuno et al., 1993; Morie et al., 2001). The original morphology in some of the study area was destroyed in the 1990s during construction of the Tokushima Expressway, and these vintage photos allow me to analyze the morphology before the disruption from construction. One pair of stereo-photos taken by GSI in 1975 was employed at each site (SI74-9 C15-31 and 32 in the Ashiro site, and SI-74-8 C11-4 and 5 in the Kamigirai site). To fix the DEM on the ground, we used the GCP points got from the topographical map with the scale of 1: 5,000 published from GSI in 1966 and 1967. After the photos were scanned with the resolution of 20 mm (1200 dpi), DEM were generated from these photos by aerial photogrammetry at the air survey company's laboratory. Topographic anaglyphs were then produced from the DEMs using Simple DEM Viewer software developed by Katayanagi (2016), and provide three-dimensional perspectives when viewed with red-cyan glasses Plate 1. The anaglyphs allow us to identify broad deformation as well as other small changes in topography (less than 1 m on the image from 1-m grid DEM), such as fault offsets and terrace risers (Goto, 2016; Goto et al., 2017). To determine the slip vector and the amount of faulting, the author examined pre-faulting geomorphology on these images by left-laterally realigning reference points along the fault using Adobe Photoshop software. The author divided and correlated
Table 1 Radiocarbon dates of the sample. Sample No.
Labo. No.
Method
Material
d13C PDB
14
C agea
Calibrated age rangeb
( ‰)
(years BP±1s)
(cal years BP ±2s)
References
e
IAAA-102836 TK-39
AMS beta
wood wood
27.16 ± 0.54 (-25)
13,940 ± 50 27,700 ± 800
17,122e16,639 31,781e28,653
This study Okada (1968)
e
TK-51
beta
humus
(-25)
25,200 ± 1400
31,096e25,141
Okada (1970)
e
Gak-6500
bata
humus
(-25)
17,830 ± 310
20,412e18,871
Okada and Tsutsumi (1990)
10101204
a 14
13
C ages were corrected by d C and calculated using Libby half-life of 5568 years. Ages in radiocarbon years (years BP) were converted to calendar years by using OxCal v4.2 calibration software (Bronk Ramsey, 2016) with the IntCal13 data set (Reimer et al., 2013). b
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
4
H. Goto / Quaternary International xxx (2017) 1e11
terrace surfaces, mainly based on the surface continuity and the degree of the surface incision. At both sites, the author established topographic profiles both parallel and perpendicular to the fault scarp and measured slip vector between the piercing points on the profiles. The author also measured the accumulated horizontal offsets after terrace formation on the detailed topographical map developed from the original 1-m grid DEM by aerial photogrammetry. The errors of the amount of offset were calculated from measured maximum and minimum offset taking into consideration the linearity of the references near the fault line. 4. Results 4.1. Ashiro site along the eastern Ikeda fault The Ikeda fault extends about 43 km ENE-WSW along a distinct topographic lineament in central Shikoku (Fig. 1B and C and 2A), making it the longest fault segment within the MTLAFZ (Okada, 1980). The fault's southern side is down-thrown east of Sakaime
pass, while on the west side; it is down to the north (Fig. 1B). The author focused on the geomorphology related to late Quaternary faulting near Ashiro, in the eastern part of the Ikeda fault along the southern foot of the Asan Mountains (Fig. 2). There are four terrace surfaces identified in this area, labeled M, L1, L2, and AL in descending order (Figs. 2 and 3, Plate 1). These classifications are same as previous studies (Mizuno et al., 1993; Morie et al., 2001), which divided surfaces based on the height difference of the terraces. However, in this study, the author revised the correlation of terraces to examine the reconstruction of reasonable geomorphological development by moving topographic anaglyphs left-laterally along the fault. The surfaces of the M, L1 and AL appear to be fan-shaped landforms deposited by tributaries near the fault traces of the Ikeda fault (Fig. 2B). Morie et al. (2001) reported that the M surface is overlain by the widespread 87.3 ka, Aso-4 tephra unit (Aoki, 2008) at Loc. 4 of Fig. 2. The L1 surfaces are the widest terrace surfaces in the study area, and are distributed along the Yoshino River and associated tributaries. The L1 surfaces were deeply incised by drainage channels,
Fig. 2. A: Distribution of active fault lines and fluvial terrace surfaces along the eastern part of the Ikeda fault. 10-m interval contours were processed using the 5-m and 10-m grid DEM of the Fundamental Geospatial Data issued by GSI. Solid red lines, dashed red lines, and dotted red lines indicate fault scarps on the terrace surfaces, indistinct fault lines, and buried active faults, respectively. Locations (Loc.) 1e4 indicate sites where tephra or carbon dating samples were collected in this and previous studies. White box indicates location of Fig. 2B. B: Enlarged geomorphological map of area B in Fig. 2A. 5-m interval contours were traced from a map issued by the town of Miyoshi (This town was disappeared several years ago due to the consolidation of municipality). The small black arrows by the labels W, X, Y, and Z indicate the offset of terrace risers and incised terrace stream. White box indicates location of Fig. 3A.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
Fig. 3. A: Topographical map produced from 1-m grid DEM. Blue dashed lines (A1 and A2) show location of topographical profiles across the Ikeda fault scarp (Fig. 3B). Red dashed lines (A3 and A4) are W-E sections parallel to the Ikeda fault scarp (Fig. 3C). B: Topographical profiles across the Ikeda fault based on the 1-m grid DEM. The measured lines are shown in Fig. 3A. C: Topographical profiles parallel to the Ikeda fault, apart from the fault scarp toward the north and south, based on the 1-m grid DEM. The measured lines are shown in Fig. 3A. Yellow lines indicate the projected topographical profiles on the fault trace (shifted for 1.8 m vertically and 30 m horizontally). White line with arrow indicates the displacement connecting the terrace inner edges as piercing points of the Ikeda fault.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
6
H. Goto / Quaternary International xxx (2017) 1e11
but its surfaces are preserved well. The longitudinal profiles show a much steeper slope of the L1 surface than that of the M surface (Fig. 3, Plate 1). These suggest that the L1 terraces were filled in the valley during the glacial period after the M terraces had been deeply incised. The development of the L1 surfaces is considered to be related to sea-level change associated with the global climate change which is commonly observed in the last glacial age in Japan (Hirakawa and Ono, 1974; Kaneko, 1977). Okada (1968) reported the age of wood fragment collected from the bottom of the L1 terrace deposit in Awa-Ikeda (Loc.1 in Fig. 2A) is about 30,000 cal years BP (Table 1). The Aira-Tanzawa tephra deposits (AT) (Smith et al., 2013) erupted during Marine Isotope Stage (MIS) 3 and are intercalated in the L1 terrace deposit at a point 5 km downstream of Fig. 2A along Yoshino River (Mizuno et al., 1993). Based on these ages, the timing of the fluvial abandonment from this surface cannot be determined, but limited to be after the age of the Aira-Tanzawa deposit (about 30,000 cal years BP). The elevation difference between the M and L1 surfaces is large (~45 m), thus they can be useful references for long-term displacement along the fault. The L2 surfaces are limited in the area near the L1 surfaces, and some scarps between L1 and L2 surfaces are indistinct. At Minota, the L1 surfaces distributed in both sides of the L2 surfaces extending from west to east (Fig. 2), and thus the L2 surfaces formed the strath surfaces below the L1 surfaces. The lower part of the topsoil on the L2 surface in Loc. 3 (Fig. 2A) (Mizuno et al., 1993) contains the Kikai-Akahoya tephra (K-Ah) (7165e7303 cal years BP) (Smith et al., 2013). The AL surfaces that overlie the M, L1, and L2 surfaces are the youngest alluvial surfaces except for the present riverbed and floodplain. The age of the M, L1, and L2 terraces are estimated to be MIS 5, MIS 3e2, and MIS 1, respectively. The M and L1 terrace surfaces on the northern side of the Yoshino River (Fig. 2) are lower toward to the south, suggesting formation by small streams flowing from the Asan Mountains. L1 terrace deposits are observed near the mouth of Aikurushidani stream (Loc. 2 in Fig. 2A), located about 4 km west of the Ashiro site. Here, large outcrops of the L1 terrace are exposed in profile from the construction of the bypass of national route No. 31. At this outcrop, there are fine deposits composed of interbedded silt to clay layers (Unit I in Fig. 4), from the land surface to a depth of 5.6 m, except for a 0.5 m-thick soil (Fig. 4). In the lower part of exposure (Unit II in Fig. 4), sub-angular gravel beds 2.3 m thick overlay the bedrock of the Izumi group (Fig. 4). The interbedded silt and clay layers (Unit I) were parallel to the top of sub-angular gravel beds (Unit II). The sub-angular gravels are composed of sandstone and mudstone only, suggesting that these were transported by the stream coming from the Asan Mountains, as is concordant with the topography of the L1 terrace surface. Wood fragments embedded in the humus of the Unit I 4.5 m below the land surface were dated at 17,122e16,639 cal years BP (Table 1). The Unit I covering terrace gravel beds (Unit II) is estimated to be overbank deposits, suggesting that the L1 surfaces formed during MIS3e2, and the age of the fluvial abandonment of the L1 terrace is estimated to be younger than 17,122e16,639 cal years BP. Okada and Tsutsumi (1990) reported that the humus embedded in the gravels of the lower terrace was dated at 20,412e18,871 cal years BP (Table 1) at about 11 km east (Loc. 5 in Fig. 1C) of the Ashiro site. This terrace is correlated with the L1 surface of this study area, and thus the estimated age of the fluvial abandonment of the L1 terrace is concordant with this age. The ENE-SSW trending Ikeda fault cuts across all of the terrace surfaces, forming a linear south-facing scarp (Fig. 3A). The ~2-kmlong Hashikura fault parallels the Ikeda fault to the north (Fig. 2A), also along the foot of the Asan Mountains (Mizuno et al., 1993). As the geomorphological features related to the right-lateral fault are
Fig. 4. The columns of the outcrop of the L1 deposit in Loc. 2 of Fig. 2A.
not observed along the Hashikura fault, the Hashikura fault is considered to accommodate very little slip of the MTLAFZ and thus to be a sub-fault of the Ikeda fault. The amount of offset for the most recent surface rupture along the Ikeda fault was measured at about 7 m at a point west of Kurogawaradani stream (Tsutsumi and Goto, 2006). Using topographic profiles across the fault scarps, the author measured about 43e45 m of vertical displacement of the M surface and ~7 m on the L1 surface (Fig. 3B), while a quite small vertical displacement of the M surfaces along the Hashikura fault were observed on the contours map (Fig. 3A). Several terrace-inner-edges between the M and L1 surfaces are offset right-laterally along the Ikeda fault. The author measured these offsets on the detailed topographical map as 110 ± 2 m (“W” in Figs. 2B, 3A) and 150 ± 5 m (X), and about 110 ± 5 m (Z), while a stream incising the L1 surface was only offset about 45 ± 15 m (Y). The terrace-inner-edge in the most west reference (W) is preserved on the south side of the fault scarp only (Figs. 2B and 3A). The height of terrace risers in the reference Z is small (~5 m), and the azimuth of this terrace riser is a little different between the both sides of the fault scarp (Fig. 2B). The amount of offset in streams incising into the terrace showed the minimum movement after the fluvial abandonment of its terrace. The most reliable reference point for accumulated slip after the fluvial abandonment of the L1 surface along the Ikeda fault should be “X” among these offsets (WZ), because of the similar azimuth between both sides of the fault scarp. ”W” and “X” are on the “trailing edge” side of the fault, which is being moved away from the stream that cut the inner edge, thus preventing post-faulting stream erosion of the inner edge. The same cannot be said of the inner edge on the eastern side of Umagidani Stream, which has been moved into the path of the stream and eroded to there is very little of its right-lateral offset left. To define the slip vector of the Ikeda fault, the author used the topographical profiles A3 and A4, that are parallel to the fault scarp (Fig. 3C). To project the topographical profiles just on the fault line,
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
these are shifted by 1.8 m vertically, and 30 m horizontally (Fig. 3C), taking into consideration following three geometrical positional relations. 1) The references “X” extend in a direction at about 45 against the strike of the fault line (Fig. 3A). 2) The inclination angle of the L1 surface near the reference “X” is about 3.4 (Fig. 3B). 3) Both topographical profiles are located at about 30 m away from the fault line near the piercing point (Fig. 3C). The slip vector can be derived from drawing a line from the inner lower edges of the terrace risers, which are the piercing points on either side of the fault. The vector shows that the A4 profile moved down 9 m, and 145 m to the west. The vertical component of displacement is about 6% of the horizontal component (Fig. 3C). 4.2. Kamigirai site along the central Chichio fault The Chichio fault extends for ~19 km trending ENE-WSW along the southern foot of the Asan Mountains, located to the north of the Yoshino River valley (Okada, 1970) (Fig. 1B and C). This fault shows mostly down to the south motion except for a north-facing scarp around its eastern end (Fig. 5). The slip rate of this fault is estimated to be about 6 mm/yr during the last 8000 years (Okada and Tsutsumi, 1997), but the slip rate during the late Pleistocene is not known. The author analyzed geomorphological features of late Pleistocene using the 1-m grid DEM produced from stereo-paired old aerial photographs along the west bank of the Higaidani stream, across the central portion of the Chichio fault where Okada and Tsutsumi (1997) already discussed the slip rate during Holocene. The author identified active fault traces and several fluvial terrace surfaces by the interpretation of both anaglyph images produced from the DEM (Plate 1) and aerial photographs. The terrace surfaces in the study area were divided into four levels, the
7
M, L1, L2, and AL surfaces, in descending order (Figs. 5 and 6). This classification and distribution are quite similar to the work of Okada and Tsutsumi (1997). The M terrace surfaces are only present near the mountains on the northern side of the fault trace. The L1 surfaces are well developed along Higaidani stream, but seem to be buried by AL deposits further east, southeast of the Kirihata hills (Fig. 5). The L1 surface near the fault scarp of Kamigirai sits ~4 m above modern stream grade and has a steeper slope than the AL and modern stream grade (Fig. 5). The L1 surfaces are subdivided at this location into two, L1-h and L1-l surfaces based on the differences in height and gradient (Fig. 6, Plate 1). A shallow valley about 140 m wide is observed within the L1-l surface along the terrace riser of the L1-h surfaces (Fig. 6). The longitudinal profiles of the L1-l surfaces are steeper than those of the L1-h surfaces, and thus, the height of the scarps between the L1-h and L1-l surfaces increases downstream (Fig. 7a). The alluvial plain of the Yoshino River distributed from the east of Kuzudani fan developed after a sea-level rise during Holocene (Yokoyama et al., 1990; Furuta, 1996). Thus, the development of the L1-h and L1-l surfaces are also considered to be related to sea-level change associated with the global climate change which is commonly observed in the last glacial age in Japan (Hirakawa and Ono, 1974; Kaneko, 1977). The AT tephra is contained in the upper part of the terrace deposit correlated to the L1-h in the east bank of Higaidani stream (Loc.6 in Fig. 5, Mizuno et al., 1993), and thus the L1-h terraces gravels were deposited during the MIS3 of the last glacial period. In the north bank of Yoshino river (Loc. 7 in Fig. 5), the humus in the land surface to a depth of 2 m embedded in the terrace gravels correlated to the L1-h was dated at 31,096e25,141 cal years BP (Okada, 1970, Table 1). Taking into consideration these age data and geomorphic features of the L1-h and L1-l surfaces, the L1-l terrace gravel is estimated to have
Fig. 5. Map showing the distribution of active fault and fluvial terrace surfaces along the eastern Chichio fault. The shade map and contours with 10-m interval are produced from the 5-m grid and 10-m grid DEM of the Fundamental Geospatial Data issued by the GSI. White box indicates location of Fig. 6.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
8
H. Goto / Quaternary International xxx (2017) 1e11
suggesting that the L1-l surfaces are correlated to the L1 surfaces in the Ashiro site of the Ikeda fault in this study. As the Kamigirai site is located about 35 km closer to the mouth of Yoshino River than the Ashiro site (Fig. 1C), the formation of the L1-l surfaces would finish earlier than those of the L1 at the Ashiro site. Therefore, the fluvial abandonment of the L1-l would be after ca. 18,000 cal years BP, before 17,122e16,639 cal years BP. Okada and Tsutsumi (1997) estimated that the age of the L2 surfaces is 8,000 cal years BP based on the height difference of the terrace riser (Okada, 1970). The AL surface is composed of the fan surfaces (ALf surface) and the fluvial plain (ALp surface) (Fig. 6). The ALf surfaces overlay the L1-h surfaces in the study area. A lot of abandoned rivers are observed on the ALp surfaces. A series of rice paddy dikes on the ALp surfaces at the Kamigirai site before the construction of Tokushima Expressway showed right-lateral offsets of about 7 m (Tsutsumi and Okada, 1996). An exposed gravel bed, shown in the archeological excavating survey related to the construction of the expressway, was also offset right-laterally approximately 6 m (Okada and Tsutsumi, 1997). Okada and Tsutsumi (1997) also estimated that the amount of right-lateral slip associated with the most recent faulting event on the Chichio fault was 6e7 m. The ENE-SSW trending fault scarp produced by the movement of the Chichio fault is observed across all surfaces in this area. As the Chichio fault extends as a single straight trace, and any other active faults cannot be identified in this area, the displacement of the Chichio fault derived from this fault scarp is representative for the central portion of the MTL active fault zone. The vertical amount of displacement on the L1-h, L1-l, and L2 surfaces is about 14.5 m, about 8.3 m, and about 4.5 m, respectively (Fig. 7A). A lot of small steps (~1 m) are observed on the surface of the L1-h, L1-l, and L2 terraces, and these are artificially modified topography to cultivate rice-fields and to build houses. However, the terrace risers and small cliffs on the terrace surface are continuous on the detailed topographical map (Fig. 6), suggesting the measured horizontal offset with small (~10 m) errors. The measured offset between the terrace inner edges on the L1-l surface across the fault scarp is 145 ± 5 m (Fig. 6). To reveal the slip vector of the Chichio fault, the author also made topographical profiles parallel to the fault scarp, to the north and south (Fig. 7B). To project the topographical profiles just on the fault line, these are shifted by 0.25 m vertically, taking into consideration following two geometrical positional relations. 1) The inclination angle of the L1-l surface is about 0.6 (Fig. 7A). 2) Both topographical profiles are located at about 25 m away from the fault line (Fig. 6B). For the horizontal component, it is not needed to shift the profile because the profiling lines are almost perpendicular to the fault trace. The slip vectors derived from the inner edge of the terraces as piercing points show that the vector direction is almost the same, and the vertical component of displacement is 2e6% of the horizontal component (Fig. 7C). 5. Discussion and conclusions
Fig. 6. A: Shaded-relief map with contour lines of 1-m interval induced from the 1-m grid DEM processed from the stereo-paired old aerial photographs in the Kamigirai site. The map is located in Fig. 5. B: Distribution of active fault traces and fluvial terrace surfaces at the Kamigirai site. The contours are based on the 1-m grid DEM processed from the stereo-paired old aerial photographs. The map is located in Fig. 5.
been deposited during the last glacial maximum period with the lowest sea-level (ca. 18,000 cal years BP) (Chen et al., 2017),
The horizontal slip of the Ikeda fault after the fluvial abandonment of the L1 surface is 145e155 m, measured from the displacement of the terrace's inner edge. The age of the terrace inner edge on the L1 surface is after the last glacial age maximum (18,000 cal years BP) and probably after 17,122e16,639 cal years BP based on the wood fragments embedded in the overbank deposits overlying the L1 gravel bed. The author calculated the net slip of 145e155 m based on the measured horizontal slip and the vector derived from the topographic profiles parallel to the fault (where the component of vertical displacement is 6% of the horizontal
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
9
Fig. 7. A: Topographical profiles across the Chichio fault, based on the 1-m grid DEM. The measured lines are shown in Fig. 6B. B: Topographical profiles parallel to the Chichio fault, apart from the fault scarp toward the north and south based on the 1-m grid DEM. The measured lines are shown in Fig. 6B. Yellow lines indicate the projected topographical profiles on the fault trace (shifted for 0.25 m vertically). White lines indicate the displacement connecting the terrace inner edges. C: Slip vectors of the Chichio fault induced from terrace edge displacement at the Kamigirai site. The vectors are based on the white lines in Fig. 7B.
displacement). The long-term slip rate of the Ikeda fault is at least 8.0 mm/yr and would be larger than 8.5 mm/yr. The horizontal slip of the Chichio fault after the fluvial abandonment of the L1-l surface is 140e150 m, measured from the displacement of terrace's inner edge. By the same calculation used above, in which the component of vertical displacement is 2e6% of the horizontal displacement, the net slip along the Chichio fault is calculated as 140e150 m. As the fluvial abandonment of the L1-l surface would be after 18,000 cal years BP, before 17,122e16,639 cal years BP, the long-term slip rate of the Chichio fault is 7.8e9.1 mm/yr. These long-term rates are a little faster than that of the Okamura
fault (5e8 mm/yr) in the central west portion of the MTLAFZ (Tsutsumi et al., 1991), and much faster than the geodetic slip rates of 5 mm/yr (Tabei et al., 2002), and ~5 mm (Aoki and Scholz, 2003) measured across the MTLAFZ. The amount of offset associated with the most recent surface rupture along the Ikeda fault was about 7 m (Tsutsumi and Goto, 2006). The recurrence interval was calculated as about less than 870 years, based on the measured amount of offset (Tsutsumi and Goto, 2006) and the calculated slip rate in this study (>8.0 mm/ yr), and this interval is much smaller than those in previous studies (1,000e1,600 years, Headquarters for Earthquake Research Promotion, 2011).
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
10
H. Goto / Quaternary International xxx (2017) 1e11
Plate 1. Topographical anaglyph induced from the 1-m grid DEM processed from the stereo-paired old aerial photographs in the Ashiro (A) and Kamigirai (B) sites.
For the Chichio fault, based on the amount of recent surface slip (6e7 m) revealed by Okada and Tsutsumi (1997) and the calculated right-lateral slip rate (7.8e9.1 mm/yr) in this study, the recurrence interval was estimated to be about 660e900 years. Okada and Tsutsumi (1997) estimated from a trenching survey that the latest event and the penultimate event occurred in the 16th century (AD 1596?), and around about 2,000 years ago, respectively. However, the timing of penultimate event is not surely constrained, because there is no information about the age of the layer overlying the fault of the penultimate event. Thus, the penultimate event is reinterpreted to have occurred after 2,000 years ago, which is concordant with the estimated recurrence interval (660e900 years) in this study. Approximately half of the recurrence interval on both the Ikeda and Chichio faults has passed since the latest faulting event occurred at the end of the Middle ages of Japan history (16th century or AD 1596) (Morino and Okada, 2002; Okada and Tsutsumi, 1997). This suggests that the next large earthquake along the
central portion of the MTLAFZ may not occur in the distant future, but sooner. The long-term seismic risk of the Ikeda and Chichio faults in the next 30 years according to the method of Headquarters for Earthquake Research Promotion (2001), was evaluated to be more than 0.4e1.9%, 0.1e3%, respectively, which is much larger than that of Headquarters for Earthquake Research Promotion (2011) (0e0.3%). Paleoseismological studies such as trenching surveys are required to confirm the results of this study. Acknowledgements This work was partly supported by MEXT KAKENHI grant numbers 16K01221, 25350428, and 22700855. The author is grateful to Professor Koji Okumura for providing valuable comments on my research, and Mr. Hiroaki Matsugi who supported my research in the field. The author also thanks the reviewers for careful reading of the manuscript and for giving fruitful comments and suggestions.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
References Aoki, K., 2008. Revised age and distribution of ca. 87 ka Aso-4 tephra based on new evidence from the northwest Pacific Ocean. Quat. Int. 178 (1), 100e118. Aoki, Y., Scholz, C.H., 2003. Interseismic deformation at the Nankai subduction zone and the median tectonic line, southwest Japan. J. Geophys. Res. 108 (B10), 1065e1070. Bronk Ramsey, C., 2016. OxCal Program, V. 4.2 Radiocarbon Accelerator Unit. University of Oxford, UK program accessible online at. http://c14.arch.ox.ac.uk/ oxcal.html. (Accessed October 2016). Chen, J., Liu, Y., Shi, X., Suk, B., Yao, Z., 2017. Climate and environmental changes for the past 44 ka clarified by pollen and algae composition in the Ulleung Basin, East Sea (Japan Sea). Quat. Int. 441, 162e173. Fitch, T.J., 1972. Plate convergence, transcurrent faults and internal deformation adjacent to southeast Asia and the western Pacific. J. Geophys. Res. 77 (23), 4432e4460. Furuta, N., 1996. Development history of the Yoshino and Agui River lowlands, northeastern part of Tokushima prefecture, Japan. Ritsumeikan Chirigaku 8, 61e72 (in Japanese). Goto, H., 1996. Fault morphology of the lyo Fault, the Median Tectonic Line active fault system. Act. Fault Res. 14, 28e38 (in Japanese, with English abstract). Goto, H., 2016. Extensive area of topographic anaglyphs covering inland and seafloor derived using a detailed digital elevation model for identifying broad tectonic deformations. In: Kamae, K. (Ed.), Earthquakes, Tsunamis and Nuclear Risks; Prediction and Assessment beyond the Fukushima Accident. Springer, Tokyo, Japan, pp. 65e74. Goto, H., Nakata, T., Tsutsumi, H., Okumura, K., Imaizumi, T., Nakamura, T., Watanabe, T., 2001. The latest surface-faulting events on the Median Tectonic Line in Shikoku, southwest Japan, based on mini-trenching and geoslicer studies. Zisin: J. Seismol. Soc. Jpn. 53, 205e219 (in Japanese, with English abstract). Goto, H., Tsutsumi, H., Toda, S., Kumahara, Y., 2017. Geomorphic features of surface ruptures associated with the 2016 Kumamoto earthquake in and around the downtown of Kumamoto City, and implications on triggered slip along active faults. Earth Planets Space 69, 26. Hashimoto, M., Kanmera, K., 1991. Pre-Neogene sedimentary and metamorphic rocks. In: Hashimoto, M. (Ed.), Geology of Japan. Terra Scientific Publishing Company, Tokyo, pp. 13e55. Headquarters for Earthquake Research Promotion, 2011. Evaluations of the Median Tectonic Line Active Fault System (in Japanese). http://www.jishin.go.jp/main/ chousa/katsudansou_pdf/81_83_85_86_89_chuo_2.pdf. (Accessed October 2016). Headquarters for Earthquake Research Promotion, 2001. Evaluation Method for Long-term Earthquake Occurrence Probability (in Japanese). http://jishin.go.jp/ main/choukihyoka/01b/chouki020326.pdf. (Accessed October 2016). Hirakawa, K., Ono, Y., 1974. The landform evolution of the Tokachi Plain. Geogr. Rev. Jpn. 47, 607e632 (in Japanese, with English abstract). Kaneko, S., 1977. Recent formations and their basal topography in and around Tokyo Bay, central Japan. Quat. Res. 8, 32e50. Kaneko, S., 1966. Transcurrent displacement along the median line, south-western Japan. N. Z. J. Geol. Geophys. 9 (1e2), 45e59. Katayanagi, Y., 2016. Simple DEM Viewer. http://www.jizoh.jp/english.html. (Accessed October 2016). Mizuno, K., Okada, A., Sangawa, A., Shimizu, F., 1993. Explanatory text of strip map of the Median Tectonic Line active fault system in Shikoku, Japan, scale 1: 25,000. Tectonic map series (8). Geol. Surv. Jpn. 63 (in Japanese, with English abstract). Morie, T., Ozawa, H., Okumura, K., 2001. Description and correlation of the Nagate Tephra found in the terrace deposits at the southern foot of Asan Range in Tokushima Prefecture, southwest Japan. Quat. Res. 40 (4), 331e336 (in Japanese, with English abstract). Morino, M., Okada, A., 2002. Faulting history of the Median Tectonic Line active fault
11
system in Shikoku, based on re-examination of trench survey results. In: Annual Report on Active Fault and Paleoearthquake Researches, vol. 2, pp. 153e182 (in Japanese, with English abstract). Nakata, T., Imaizumi, T., 2002. Digital Active Fault Map of Japan. University of Tokyo Press, Tokyo, Japan (in Japanese, with English abstract). Okada, A., 1968. Strike-slip faulting of late quaternary along the median dislocation line in the surroundings of Awa-Ikeda, northeastern Shikoku. Quat. Res. 7 (1), 15e26 (in Japanese, with English abstract). Okada, A., 1970. Fault topography and rate of faulting along the Median Tectonic Line in the drainage basin of the river Yoshino, northeastern Shikoku, Japan. Geogr. Rev. Jpn. 43 (1), 1e21 (in Japanese, with English abstract). Okada, A., 1980. Quaternary faulting along the median tectonic line of southwest Japan. Memoirs Geol. Soc. Jpn. 18, 79e108. Okada, A., Sangawa, A., 1978. Fault morphology and quaternary faulting along the median tectonic line in the southern part of the Izumi range. Geogr. Rev. Jpn. 51 (5), 385e405 (in Japanese, with English abstract). Okada, A., Tsutsumi, H., 1990. Date on fault exposures of the Median Tectonic Line and geopmorphic surface chronology in the central to eastern part of Shikoku, Southwest Japan. Act. Fault Res. 8, 31e47 (in Japanese). Okada, A., Tsutsumi, H., 1997. Holocene Activity of the Chichio fault, the median tectonic line Active fault system, southwest Japan, based on trenching studies. J. Geogr. 106 (5), 644e659 (in Japanese, with English abstract). Okayama, T., 1988. 1: 1,000,000 the Summit Level Map of Japan. Kokon Shoin, Tokyo (in Japanese). Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., , C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Haflidason, H., Hajdas, I., Hatte Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M., Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal BP. Radiocarbon 55 (4), 1869e1887. (revised edition). In: Research Group for Active Faults of Japan (Ed.), 1991. Active Faults in Japan: Sheet Maps and Inventories. University of Tokyo Press, Tokyo (in Japanese, with English abstract). Smith, V.C., Staff, R.A., Blockley, S.P.E., Bronk Ramsey, C., Nakagawa, T., Mark, D.F., Takemura, K., Danhara, T., Suigetsu 2006 Project Members, 2013. Identification and correlation of visible tephras in the Lake Suigetsu SG06 sedimentary archive, Japan: chronostratigraphic markers for synchronising of east Asian/ west Pacific palaeoclimatic records across the last 150 ka. Quat. Sci. Rev. 67, 121e137. Tabei, T., Hashimoto, M., Miyazaki, S., Hirahara, K., Kimata, F., Matsushima, T., Tanaka, T., Eguchi, Y., Takaya, T., Hoso, Y., Ohya, F., Kato, T., 2002. Subsurface structure and faulting of the Median Tectonic Line, southwest Japan inferred from GPS velocity field. Earth Planets Space 54, 1065e1070. Takahashi, M., 2006. Tectonic boundary between northeast and southwest Japan arcs during Japan sea opening. J. Geol. Soc. Jpn. 112 (1), 14e32 (in Japanese, with English abstract). Tsutsumi, H., Goto, H., 2006. Surface offsets associated with the most recent earthquakes along the Median Tectonic Line active fault zone in Shikoku, southwest Japan. Zisin J. Seismol. Soc. Jpn. 59 (2), 117e132 (in Japanese, with English abstract). Tsutsumi, H., Okada, A., 1996. Segmentation and Holocene surface faulting on the median tectonic line, southwest Japan. J. Geophys. Res. 101 (B3), 5855e5871. Tsutsumi, H., Okada, A., Nakata, T., Ando, M., Tsukuda, T., 1991. Timing and displacement of Holocene faulting on the median tectonic line in central Shikoku, southwest Japan. J. Struct. Geol. 13 (2), 227e233. Tsutsumi, H., Okada, A., Goto, H., Matsugi, H., 2000. Late Holocene faulting along the Kawakami fault, an active segment of the median tectonic line, southwest Japan. Act. Fault Res. 19, 77e86 (in Japanese, with English abstract). Yokoyama, T., Matsunami, S., Okumura, K., 1990. The formative process of alluvial deposits in the Tokushima plain. J. Geogr. 99, 775e789 (in Japanese, with English abstract).
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
Contents lists available at ScienceDirect
Quaternary International journal homepage: www.elsevier.com/locate/quaint
Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan Hideaki Goto Department of Geography, Hiroshima University, 1-2-3, Kagamiyama, Higashi-Hiroshima, Hiroshima, Japan
a r t i c l e i n f o
a b s t r a c t
Article history: Received 30 October 2016 Received in revised form 20 November 2017 Accepted 6 December 2017 Available online xxx
The Median Tectonic Line active fault zone (MTLAFZ) extends for about 190 km through Shikoku, southwest Japan. Though the MTLAFZ is the most significant onshore active tectonic feature in southwest Japan, its late Quaternary slip rate has been estimated at only a few locations with reasonable references and ages. Better information on this feature's recent slip rates is critical to understanding the ongoing tectonic processes in the region and evaluating the seismic risk of this fault zone. In this paper, new estimates of the late Quaternary slip rate are reported from the Ikeda and Chichio faults in the central portion of the MTLAFZ. The author mapped late Pleistocene fluvial terrace surfaces and used tephrochronology and radiocarbon dating to constrain the age, and measured offset of terrace risers. The slip vectors of both faults are similar, as derived from piercing points on the bottoms of the terrace risers. The vertical component of displacement is 2e6% of the horizontal component. Long-term slip rates during the late Quaternary were calculated at 7.8e9.1 mm/yr, which is more precise than those in the previous studies and represents the highest slip rate in the MTLAFZ. This rate is also much faster than previously reported shorter-term slip rates of geodetic study. The long-term seismic risk of large earthquakes (>M8) along the Ikeda and Chichio faults in the next 30 years are evaluated to be more than 0.4e1.9% and 0.1 e3%, respectively, much higher than a previous estimate of 0e0.3%. © 2017 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Keywords: Active fault Geomorphology Slip rate Digital elevation model Median tectonic line Southwest Japan
1. Introduction The Median Tectonic Line (MTL) is the most prominent geological boundary in southwest Japan (Fig. 1A). The Median Tectonic Line active fault zone (MTLAFZ) extends along the MTL for about 190 km through Shikoku (Research Group for Active Faults of Japan, 1991, Fig. 1). The MTLAFZ is an arc-parallel, right-lateral strike-slip fault related to the oblique subduction of the Philippine Sea plate beneath the Eurasian plate along the Nankai trough (Fitch, 1972), suggesting that the MTLAFZ is the most significant onshore active tectonic feature in southwest Japan. Though late Quaternary slip rate has been estimated at a few locations, these studies were conducted in 1960e90s based on the amount of offset measured on topographic map with a contour interval of 5e10 m. Improved constraints on the slip rate of this feature are critical to assessing the seismic hazard along this fault zone and to understanding the ongoing tectonic processes in the region. In this paper, new late Quaternary slip rate estimates are
E-mail address: [email protected].
reported for the Ikeda and Chichio faults along the central portion of the MTLAFZ in Shikoku, Japan. Towards this end, the author mapped fault traces and measured offsets within late Pleistocene fluvial terraces and dated terrace deposits using tephrochronology reported in the prior study and radiocarbon ages. Slip vectors and offset were measured from topographic profiles and maps derived from digital elevation models (DEM) developed from processing of aerial imagery taken in the 1975, before the construction of an expressway near the active fault traces.
2. Regional setting The MTL separates the low-pressure, high-temperature Ryoke metamorphic belt to the north and the high-pressure, low-temperature Sambagawa metamorphic belt to the south, along a line running from Shikoku to Kanto (Fig. 1A) (Hashimoto and Kanmera, 1991; Takahashi, 2006). The MTL is offset along the ItoigawaShizuoka Tectonic line (ISTL) in Chubu, suggesting that the MTL is older than the ISTL. The Upper Cretaceous Izumi group, composed of interbedded sandstone and mudstone, was deposited in the narrow half-graben to the north of the MTL, related to left-lateral
https://doi.org/10.1016/j.quaint.2017.12.013 1040-6182/© 2017 The Author. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
Fig. 1. A: Distribution of active faults and major tectonic lines in central Japan. The shaded relief map uses a 1-km grid DEM from SRTMplus Data provided by the USGS. White and red lines indicate active fault lines and tectonic lines, respectively, after Nakata and Imaizumi (2002) and Hashimoto and Kanmera (1991). The rectangle of dashed white line and purple line indicate location of Fig. 1B and C, respectively. Abbreviations are as follows: MTL, Median Tectonic Line; ISTL, Itoigawa-Shizuoka Tectonic Line. B: Active faults (red lines) along the MTL within the Shikoku region. Topographic map is after Okayama (1988). The stars with abbreviation indicates where the prior studies evaluated the long-term slip rates. Abbreviations show the site where the studies are conducted as follows: a: Goto (1996); b: Tsutsumi et al. (1991); c: Okada (1968); d: Okada (1970); e: Tsutsumi and Okada (1996). Abbreviations of the faults are as follows: I.F., Iyo fault; S.F., Shigenobu fault; Kw.F., Kawakami fault; O.F., Okamura fault; Iz.F., Ishizuchi fault; Ik.F, Ikeda fault; M.F., Mino fault; C.F., Chichio fault; Z.F., Zunden fault; It.F., Itano fault; N.F., Naruto fault. B: Enlarged geomorphological map of area C in Fig. 1A. The shaded relief map uses a 90-m grid DEM from SRTM Data provided by the USGS. Red lines indicate active fault lines, after Nakata and Imaizumi (2002). Abbreviations of the faults are same as Fig. 1B.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
faulting from the western part of the Kii Peninsula to Shikoku (Hashimoto and Kanmera, 1991). Evidence for late Quaternary faulting along the MTL was reported by Kaneko (1966) and Okada (1980), who observed fault scarps cutting fluvial terraces, and offset streams and terrace risers from the western part of Kii peninsula to Shikoku (Fig. 1A). The late Quaternary movement along the MTL has been found to be predominantly right-lateral strike-slip, with a vertical displacement less than one-tenth of horizontal displacement (Okada, 1980). The MTLAFZ is composed of several active faults, up to several tens of kilometers long and parallel to the MTL. Previous studies have estimated the long-term horizontal slip rate of the active faults along the MTLAFZ, based on the offset of channels that incised terraces and risers, as 1e3 mm/yr in Kii Peninsula (Okada and Sangawa, 1978), 5e10 mm/yr in the eastern to central part of Shikoku (Okada, 1968, 1970; Tsutsumi et al., 1991; Okada and Tsutsumi, 1997), and 1e2 mm/yr in the western part of Shikoku (Goto, 1996) (Fig. 1B). The slip rate in the eastern to central part of Shikoku is the highest, but is not well-constrained, because a few geomorphological studies was conducted in 1960e90s and the ages for the terrace surface of the offset references have not been clear. This study is focused on the Ikeda and Chichio faults which are the most prominent faults in eastern Shikoku (Fig. 1C). The recurrence intervals of faulting events in Shikoku during the late Holocene were revealed by paleoseismological studies to be about >1000 years (Okada and Tsutsumi, 1997; Morino and Okada, 2002). The most recent event was in Shikoku, concentrated at the end of the Middle ages in Japanese history (12-16th century) (Goto et al., 2001), probably corresponding to the 1596 Keicho earthquake recorded in historical documents from the Kinki district and Shikoku (Tsutsumi et al., 2000). Long-term seismic risk from this fault hosting a large earthquake (>M8) in the next 30 years is estimated to be 0e5% in Kii and 0e0.3% in Shikoku (Headquarters for Earthquake Research Promotion, 2011). Though the timing and slip amount of surface rupture associated with the most recent event are revealed in Shikoku (Goto et al., 2001; Tsutsumi and Goto, 2006), the long-term seismic risks are uncertain mainly because of unclear recurrence intervals (Morino and Okada, 2002). As the slip amount of the most recent surface faulting were detected in the previous studies, determining the long-term slip rate accurately is crucial for evaluating the earthquake risks along this fault zone. Based on the analysis of densely distributed campaign GPS measurements and block modeling results of Tabei et al. (2002), 5 mm/yr of contemporary dextral shear occurs across the MTL, when the fault dip angle is 35 and locked at 15 km from the earth surface. On the other hand, Aoki and Scholz (2003) analyzed the interplate coupling coefficient as the function of depth only, and the MTLAFZ slip rate was estimated to be quite variable, between 0.00 and 5.50 mm/yr assuming a vertical fault, and between 0.00 and 3.88 mm/yr assuming a 35 north-dipping fault. These studies have problems that the dip angle of the MTL and the MTLAFZ is uncertain
3
at depth, and the value of geodetic surveys may not be representative of the deformation within the intraplate zone because of the shortness of measuring periods. Thus, a long-term slip rate would be useful in providing insight to understanding the deformation through the Nankai trough to the intraplate region of southwest Japan. Two sites were selected along the Ikeda and Chichio faults where faults cut through several terraces, and laterally offset terrace risers. One is the Ashiro site on the Ikeda fault, and the other is the Kamigirai site on the Chichio fault (Fig. 1B and C). In the Kamigirai site, Okada and Tsutsumi (1997) discussed the slip rate based on the offset of Holocene terraces, and estimated that the slip rate was 6e7 mm/yr. However, the age of the terraces was determined only based on the height difference between the upper terrace developed during the last glacial age and alluvial plain near the fault scarps.
3. Material and methods To identify topographic features related to active faults, the author analyzed aerial photograph from the Geospatial Information Authority of Japan (GSI), taken in the 1960s and 1970s with the scale of 1: 8,000 to 1: 20,000. The author conducted the field survey to check the correlation of the terrace surfaces and observed the deposit of the fluvial terrace. The author also collected a wood sample buried in the terrace deposit for radiocarbon dating (Table 1). The ages of terrace formation were determined from radiocarbon age dating as well as from previous studies using tephrochronology (Mizuno et al., 1993; Morie et al., 2001). The original morphology in some of the study area was destroyed in the 1990s during construction of the Tokushima Expressway, and these vintage photos allow me to analyze the morphology before the disruption from construction. One pair of stereo-photos taken by GSI in 1975 was employed at each site (SI74-9 C15-31 and 32 in the Ashiro site, and SI-74-8 C11-4 and 5 in the Kamigirai site). To fix the DEM on the ground, we used the GCP points got from the topographical map with the scale of 1: 5,000 published from GSI in 1966 and 1967. After the photos were scanned with the resolution of 20 mm (1200 dpi), DEM were generated from these photos by aerial photogrammetry at the air survey company's laboratory. Topographic anaglyphs were then produced from the DEMs using Simple DEM Viewer software developed by Katayanagi (2016), and provide three-dimensional perspectives when viewed with red-cyan glasses Plate 1. The anaglyphs allow us to identify broad deformation as well as other small changes in topography (less than 1 m on the image from 1-m grid DEM), such as fault offsets and terrace risers (Goto, 2016; Goto et al., 2017). To determine the slip vector and the amount of faulting, the author examined pre-faulting geomorphology on these images by left-laterally realigning reference points along the fault using Adobe Photoshop software. The author divided and correlated
Table 1 Radiocarbon dates of the sample. Sample No.
Labo. No.
Method
Material
d13C PDB
14
C agea
Calibrated age rangeb
( ‰)
(years BP±1s)
(cal years BP ±2s)
References
e
IAAA-102836 TK-39
AMS beta
wood wood
27.16 ± 0.54 (-25)
13,940 ± 50 27,700 ± 800
17,122e16,639 31,781e28,653
This study Okada (1968)
e
TK-51
beta
humus
(-25)
25,200 ± 1400
31,096e25,141
Okada (1970)
e
Gak-6500
bata
humus
(-25)
17,830 ± 310
20,412e18,871
Okada and Tsutsumi (1990)
10101204
a 14
13
C ages were corrected by d C and calculated using Libby half-life of 5568 years. Ages in radiocarbon years (years BP) were converted to calendar years by using OxCal v4.2 calibration software (Bronk Ramsey, 2016) with the IntCal13 data set (Reimer et al., 2013). b
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
4
H. Goto / Quaternary International xxx (2017) 1e11
terrace surfaces, mainly based on the surface continuity and the degree of the surface incision. At both sites, the author established topographic profiles both parallel and perpendicular to the fault scarp and measured slip vector between the piercing points on the profiles. The author also measured the accumulated horizontal offsets after terrace formation on the detailed topographical map developed from the original 1-m grid DEM by aerial photogrammetry. The errors of the amount of offset were calculated from measured maximum and minimum offset taking into consideration the linearity of the references near the fault line. 4. Results 4.1. Ashiro site along the eastern Ikeda fault The Ikeda fault extends about 43 km ENE-WSW along a distinct topographic lineament in central Shikoku (Fig. 1B and C and 2A), making it the longest fault segment within the MTLAFZ (Okada, 1980). The fault's southern side is down-thrown east of Sakaime
pass, while on the west side; it is down to the north (Fig. 1B). The author focused on the geomorphology related to late Quaternary faulting near Ashiro, in the eastern part of the Ikeda fault along the southern foot of the Asan Mountains (Fig. 2). There are four terrace surfaces identified in this area, labeled M, L1, L2, and AL in descending order (Figs. 2 and 3, Plate 1). These classifications are same as previous studies (Mizuno et al., 1993; Morie et al., 2001), which divided surfaces based on the height difference of the terraces. However, in this study, the author revised the correlation of terraces to examine the reconstruction of reasonable geomorphological development by moving topographic anaglyphs left-laterally along the fault. The surfaces of the M, L1 and AL appear to be fan-shaped landforms deposited by tributaries near the fault traces of the Ikeda fault (Fig. 2B). Morie et al. (2001) reported that the M surface is overlain by the widespread 87.3 ka, Aso-4 tephra unit (Aoki, 2008) at Loc. 4 of Fig. 2. The L1 surfaces are the widest terrace surfaces in the study area, and are distributed along the Yoshino River and associated tributaries. The L1 surfaces were deeply incised by drainage channels,
Fig. 2. A: Distribution of active fault lines and fluvial terrace surfaces along the eastern part of the Ikeda fault. 10-m interval contours were processed using the 5-m and 10-m grid DEM of the Fundamental Geospatial Data issued by GSI. Solid red lines, dashed red lines, and dotted red lines indicate fault scarps on the terrace surfaces, indistinct fault lines, and buried active faults, respectively. Locations (Loc.) 1e4 indicate sites where tephra or carbon dating samples were collected in this and previous studies. White box indicates location of Fig. 2B. B: Enlarged geomorphological map of area B in Fig. 2A. 5-m interval contours were traced from a map issued by the town of Miyoshi (This town was disappeared several years ago due to the consolidation of municipality). The small black arrows by the labels W, X, Y, and Z indicate the offset of terrace risers and incised terrace stream. White box indicates location of Fig. 3A.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
Fig. 3. A: Topographical map produced from 1-m grid DEM. Blue dashed lines (A1 and A2) show location of topographical profiles across the Ikeda fault scarp (Fig. 3B). Red dashed lines (A3 and A4) are W-E sections parallel to the Ikeda fault scarp (Fig. 3C). B: Topographical profiles across the Ikeda fault based on the 1-m grid DEM. The measured lines are shown in Fig. 3A. C: Topographical profiles parallel to the Ikeda fault, apart from the fault scarp toward the north and south, based on the 1-m grid DEM. The measured lines are shown in Fig. 3A. Yellow lines indicate the projected topographical profiles on the fault trace (shifted for 1.8 m vertically and 30 m horizontally). White line with arrow indicates the displacement connecting the terrace inner edges as piercing points of the Ikeda fault.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
6
H. Goto / Quaternary International xxx (2017) 1e11
but its surfaces are preserved well. The longitudinal profiles show a much steeper slope of the L1 surface than that of the M surface (Fig. 3, Plate 1). These suggest that the L1 terraces were filled in the valley during the glacial period after the M terraces had been deeply incised. The development of the L1 surfaces is considered to be related to sea-level change associated with the global climate change which is commonly observed in the last glacial age in Japan (Hirakawa and Ono, 1974; Kaneko, 1977). Okada (1968) reported the age of wood fragment collected from the bottom of the L1 terrace deposit in Awa-Ikeda (Loc.1 in Fig. 2A) is about 30,000 cal years BP (Table 1). The Aira-Tanzawa tephra deposits (AT) (Smith et al., 2013) erupted during Marine Isotope Stage (MIS) 3 and are intercalated in the L1 terrace deposit at a point 5 km downstream of Fig. 2A along Yoshino River (Mizuno et al., 1993). Based on these ages, the timing of the fluvial abandonment from this surface cannot be determined, but limited to be after the age of the Aira-Tanzawa deposit (about 30,000 cal years BP). The elevation difference between the M and L1 surfaces is large (~45 m), thus they can be useful references for long-term displacement along the fault. The L2 surfaces are limited in the area near the L1 surfaces, and some scarps between L1 and L2 surfaces are indistinct. At Minota, the L1 surfaces distributed in both sides of the L2 surfaces extending from west to east (Fig. 2), and thus the L2 surfaces formed the strath surfaces below the L1 surfaces. The lower part of the topsoil on the L2 surface in Loc. 3 (Fig. 2A) (Mizuno et al., 1993) contains the Kikai-Akahoya tephra (K-Ah) (7165e7303 cal years BP) (Smith et al., 2013). The AL surfaces that overlie the M, L1, and L2 surfaces are the youngest alluvial surfaces except for the present riverbed and floodplain. The age of the M, L1, and L2 terraces are estimated to be MIS 5, MIS 3e2, and MIS 1, respectively. The M and L1 terrace surfaces on the northern side of the Yoshino River (Fig. 2) are lower toward to the south, suggesting formation by small streams flowing from the Asan Mountains. L1 terrace deposits are observed near the mouth of Aikurushidani stream (Loc. 2 in Fig. 2A), located about 4 km west of the Ashiro site. Here, large outcrops of the L1 terrace are exposed in profile from the construction of the bypass of national route No. 31. At this outcrop, there are fine deposits composed of interbedded silt to clay layers (Unit I in Fig. 4), from the land surface to a depth of 5.6 m, except for a 0.5 m-thick soil (Fig. 4). In the lower part of exposure (Unit II in Fig. 4), sub-angular gravel beds 2.3 m thick overlay the bedrock of the Izumi group (Fig. 4). The interbedded silt and clay layers (Unit I) were parallel to the top of sub-angular gravel beds (Unit II). The sub-angular gravels are composed of sandstone and mudstone only, suggesting that these were transported by the stream coming from the Asan Mountains, as is concordant with the topography of the L1 terrace surface. Wood fragments embedded in the humus of the Unit I 4.5 m below the land surface were dated at 17,122e16,639 cal years BP (Table 1). The Unit I covering terrace gravel beds (Unit II) is estimated to be overbank deposits, suggesting that the L1 surfaces formed during MIS3e2, and the age of the fluvial abandonment of the L1 terrace is estimated to be younger than 17,122e16,639 cal years BP. Okada and Tsutsumi (1990) reported that the humus embedded in the gravels of the lower terrace was dated at 20,412e18,871 cal years BP (Table 1) at about 11 km east (Loc. 5 in Fig. 1C) of the Ashiro site. This terrace is correlated with the L1 surface of this study area, and thus the estimated age of the fluvial abandonment of the L1 terrace is concordant with this age. The ENE-SSW trending Ikeda fault cuts across all of the terrace surfaces, forming a linear south-facing scarp (Fig. 3A). The ~2-kmlong Hashikura fault parallels the Ikeda fault to the north (Fig. 2A), also along the foot of the Asan Mountains (Mizuno et al., 1993). As the geomorphological features related to the right-lateral fault are
Fig. 4. The columns of the outcrop of the L1 deposit in Loc. 2 of Fig. 2A.
not observed along the Hashikura fault, the Hashikura fault is considered to accommodate very little slip of the MTLAFZ and thus to be a sub-fault of the Ikeda fault. The amount of offset for the most recent surface rupture along the Ikeda fault was measured at about 7 m at a point west of Kurogawaradani stream (Tsutsumi and Goto, 2006). Using topographic profiles across the fault scarps, the author measured about 43e45 m of vertical displacement of the M surface and ~7 m on the L1 surface (Fig. 3B), while a quite small vertical displacement of the M surfaces along the Hashikura fault were observed on the contours map (Fig. 3A). Several terrace-inner-edges between the M and L1 surfaces are offset right-laterally along the Ikeda fault. The author measured these offsets on the detailed topographical map as 110 ± 2 m (“W” in Figs. 2B, 3A) and 150 ± 5 m (X), and about 110 ± 5 m (Z), while a stream incising the L1 surface was only offset about 45 ± 15 m (Y). The terrace-inner-edge in the most west reference (W) is preserved on the south side of the fault scarp only (Figs. 2B and 3A). The height of terrace risers in the reference Z is small (~5 m), and the azimuth of this terrace riser is a little different between the both sides of the fault scarp (Fig. 2B). The amount of offset in streams incising into the terrace showed the minimum movement after the fluvial abandonment of its terrace. The most reliable reference point for accumulated slip after the fluvial abandonment of the L1 surface along the Ikeda fault should be “X” among these offsets (WZ), because of the similar azimuth between both sides of the fault scarp. ”W” and “X” are on the “trailing edge” side of the fault, which is being moved away from the stream that cut the inner edge, thus preventing post-faulting stream erosion of the inner edge. The same cannot be said of the inner edge on the eastern side of Umagidani Stream, which has been moved into the path of the stream and eroded to there is very little of its right-lateral offset left. To define the slip vector of the Ikeda fault, the author used the topographical profiles A3 and A4, that are parallel to the fault scarp (Fig. 3C). To project the topographical profiles just on the fault line,
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
these are shifted by 1.8 m vertically, and 30 m horizontally (Fig. 3C), taking into consideration following three geometrical positional relations. 1) The references “X” extend in a direction at about 45 against the strike of the fault line (Fig. 3A). 2) The inclination angle of the L1 surface near the reference “X” is about 3.4 (Fig. 3B). 3) Both topographical profiles are located at about 30 m away from the fault line near the piercing point (Fig. 3C). The slip vector can be derived from drawing a line from the inner lower edges of the terrace risers, which are the piercing points on either side of the fault. The vector shows that the A4 profile moved down 9 m, and 145 m to the west. The vertical component of displacement is about 6% of the horizontal component (Fig. 3C). 4.2. Kamigirai site along the central Chichio fault The Chichio fault extends for ~19 km trending ENE-WSW along the southern foot of the Asan Mountains, located to the north of the Yoshino River valley (Okada, 1970) (Fig. 1B and C). This fault shows mostly down to the south motion except for a north-facing scarp around its eastern end (Fig. 5). The slip rate of this fault is estimated to be about 6 mm/yr during the last 8000 years (Okada and Tsutsumi, 1997), but the slip rate during the late Pleistocene is not known. The author analyzed geomorphological features of late Pleistocene using the 1-m grid DEM produced from stereo-paired old aerial photographs along the west bank of the Higaidani stream, across the central portion of the Chichio fault where Okada and Tsutsumi (1997) already discussed the slip rate during Holocene. The author identified active fault traces and several fluvial terrace surfaces by the interpretation of both anaglyph images produced from the DEM (Plate 1) and aerial photographs. The terrace surfaces in the study area were divided into four levels, the
7
M, L1, L2, and AL surfaces, in descending order (Figs. 5 and 6). This classification and distribution are quite similar to the work of Okada and Tsutsumi (1997). The M terrace surfaces are only present near the mountains on the northern side of the fault trace. The L1 surfaces are well developed along Higaidani stream, but seem to be buried by AL deposits further east, southeast of the Kirihata hills (Fig. 5). The L1 surface near the fault scarp of Kamigirai sits ~4 m above modern stream grade and has a steeper slope than the AL and modern stream grade (Fig. 5). The L1 surfaces are subdivided at this location into two, L1-h and L1-l surfaces based on the differences in height and gradient (Fig. 6, Plate 1). A shallow valley about 140 m wide is observed within the L1-l surface along the terrace riser of the L1-h surfaces (Fig. 6). The longitudinal profiles of the L1-l surfaces are steeper than those of the L1-h surfaces, and thus, the height of the scarps between the L1-h and L1-l surfaces increases downstream (Fig. 7a). The alluvial plain of the Yoshino River distributed from the east of Kuzudani fan developed after a sea-level rise during Holocene (Yokoyama et al., 1990; Furuta, 1996). Thus, the development of the L1-h and L1-l surfaces are also considered to be related to sea-level change associated with the global climate change which is commonly observed in the last glacial age in Japan (Hirakawa and Ono, 1974; Kaneko, 1977). The AT tephra is contained in the upper part of the terrace deposit correlated to the L1-h in the east bank of Higaidani stream (Loc.6 in Fig. 5, Mizuno et al., 1993), and thus the L1-h terraces gravels were deposited during the MIS3 of the last glacial period. In the north bank of Yoshino river (Loc. 7 in Fig. 5), the humus in the land surface to a depth of 2 m embedded in the terrace gravels correlated to the L1-h was dated at 31,096e25,141 cal years BP (Okada, 1970, Table 1). Taking into consideration these age data and geomorphic features of the L1-h and L1-l surfaces, the L1-l terrace gravel is estimated to have
Fig. 5. Map showing the distribution of active fault and fluvial terrace surfaces along the eastern Chichio fault. The shade map and contours with 10-m interval are produced from the 5-m grid and 10-m grid DEM of the Fundamental Geospatial Data issued by the GSI. White box indicates location of Fig. 6.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
8
H. Goto / Quaternary International xxx (2017) 1e11
suggesting that the L1-l surfaces are correlated to the L1 surfaces in the Ashiro site of the Ikeda fault in this study. As the Kamigirai site is located about 35 km closer to the mouth of Yoshino River than the Ashiro site (Fig. 1C), the formation of the L1-l surfaces would finish earlier than those of the L1 at the Ashiro site. Therefore, the fluvial abandonment of the L1-l would be after ca. 18,000 cal years BP, before 17,122e16,639 cal years BP. Okada and Tsutsumi (1997) estimated that the age of the L2 surfaces is 8,000 cal years BP based on the height difference of the terrace riser (Okada, 1970). The AL surface is composed of the fan surfaces (ALf surface) and the fluvial plain (ALp surface) (Fig. 6). The ALf surfaces overlay the L1-h surfaces in the study area. A lot of abandoned rivers are observed on the ALp surfaces. A series of rice paddy dikes on the ALp surfaces at the Kamigirai site before the construction of Tokushima Expressway showed right-lateral offsets of about 7 m (Tsutsumi and Okada, 1996). An exposed gravel bed, shown in the archeological excavating survey related to the construction of the expressway, was also offset right-laterally approximately 6 m (Okada and Tsutsumi, 1997). Okada and Tsutsumi (1997) also estimated that the amount of right-lateral slip associated with the most recent faulting event on the Chichio fault was 6e7 m. The ENE-SSW trending fault scarp produced by the movement of the Chichio fault is observed across all surfaces in this area. As the Chichio fault extends as a single straight trace, and any other active faults cannot be identified in this area, the displacement of the Chichio fault derived from this fault scarp is representative for the central portion of the MTL active fault zone. The vertical amount of displacement on the L1-h, L1-l, and L2 surfaces is about 14.5 m, about 8.3 m, and about 4.5 m, respectively (Fig. 7A). A lot of small steps (~1 m) are observed on the surface of the L1-h, L1-l, and L2 terraces, and these are artificially modified topography to cultivate rice-fields and to build houses. However, the terrace risers and small cliffs on the terrace surface are continuous on the detailed topographical map (Fig. 6), suggesting the measured horizontal offset with small (~10 m) errors. The measured offset between the terrace inner edges on the L1-l surface across the fault scarp is 145 ± 5 m (Fig. 6). To reveal the slip vector of the Chichio fault, the author also made topographical profiles parallel to the fault scarp, to the north and south (Fig. 7B). To project the topographical profiles just on the fault line, these are shifted by 0.25 m vertically, taking into consideration following two geometrical positional relations. 1) The inclination angle of the L1-l surface is about 0.6 (Fig. 7A). 2) Both topographical profiles are located at about 25 m away from the fault line (Fig. 6B). For the horizontal component, it is not needed to shift the profile because the profiling lines are almost perpendicular to the fault trace. The slip vectors derived from the inner edge of the terraces as piercing points show that the vector direction is almost the same, and the vertical component of displacement is 2e6% of the horizontal component (Fig. 7C). 5. Discussion and conclusions
Fig. 6. A: Shaded-relief map with contour lines of 1-m interval induced from the 1-m grid DEM processed from the stereo-paired old aerial photographs in the Kamigirai site. The map is located in Fig. 5. B: Distribution of active fault traces and fluvial terrace surfaces at the Kamigirai site. The contours are based on the 1-m grid DEM processed from the stereo-paired old aerial photographs. The map is located in Fig. 5.
been deposited during the last glacial maximum period with the lowest sea-level (ca. 18,000 cal years BP) (Chen et al., 2017),
The horizontal slip of the Ikeda fault after the fluvial abandonment of the L1 surface is 145e155 m, measured from the displacement of the terrace's inner edge. The age of the terrace inner edge on the L1 surface is after the last glacial age maximum (18,000 cal years BP) and probably after 17,122e16,639 cal years BP based on the wood fragments embedded in the overbank deposits overlying the L1 gravel bed. The author calculated the net slip of 145e155 m based on the measured horizontal slip and the vector derived from the topographic profiles parallel to the fault (where the component of vertical displacement is 6% of the horizontal
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
9
Fig. 7. A: Topographical profiles across the Chichio fault, based on the 1-m grid DEM. The measured lines are shown in Fig. 6B. B: Topographical profiles parallel to the Chichio fault, apart from the fault scarp toward the north and south based on the 1-m grid DEM. The measured lines are shown in Fig. 6B. Yellow lines indicate the projected topographical profiles on the fault trace (shifted for 0.25 m vertically). White lines indicate the displacement connecting the terrace inner edges. C: Slip vectors of the Chichio fault induced from terrace edge displacement at the Kamigirai site. The vectors are based on the white lines in Fig. 7B.
displacement). The long-term slip rate of the Ikeda fault is at least 8.0 mm/yr and would be larger than 8.5 mm/yr. The horizontal slip of the Chichio fault after the fluvial abandonment of the L1-l surface is 140e150 m, measured from the displacement of terrace's inner edge. By the same calculation used above, in which the component of vertical displacement is 2e6% of the horizontal displacement, the net slip along the Chichio fault is calculated as 140e150 m. As the fluvial abandonment of the L1-l surface would be after 18,000 cal years BP, before 17,122e16,639 cal years BP, the long-term slip rate of the Chichio fault is 7.8e9.1 mm/yr. These long-term rates are a little faster than that of the Okamura
fault (5e8 mm/yr) in the central west portion of the MTLAFZ (Tsutsumi et al., 1991), and much faster than the geodetic slip rates of 5 mm/yr (Tabei et al., 2002), and ~5 mm (Aoki and Scholz, 2003) measured across the MTLAFZ. The amount of offset associated with the most recent surface rupture along the Ikeda fault was about 7 m (Tsutsumi and Goto, 2006). The recurrence interval was calculated as about less than 870 years, based on the measured amount of offset (Tsutsumi and Goto, 2006) and the calculated slip rate in this study (>8.0 mm/ yr), and this interval is much smaller than those in previous studies (1,000e1,600 years, Headquarters for Earthquake Research Promotion, 2011).
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
10
H. Goto / Quaternary International xxx (2017) 1e11
Plate 1. Topographical anaglyph induced from the 1-m grid DEM processed from the stereo-paired old aerial photographs in the Ashiro (A) and Kamigirai (B) sites.
For the Chichio fault, based on the amount of recent surface slip (6e7 m) revealed by Okada and Tsutsumi (1997) and the calculated right-lateral slip rate (7.8e9.1 mm/yr) in this study, the recurrence interval was estimated to be about 660e900 years. Okada and Tsutsumi (1997) estimated from a trenching survey that the latest event and the penultimate event occurred in the 16th century (AD 1596?), and around about 2,000 years ago, respectively. However, the timing of penultimate event is not surely constrained, because there is no information about the age of the layer overlying the fault of the penultimate event. Thus, the penultimate event is reinterpreted to have occurred after 2,000 years ago, which is concordant with the estimated recurrence interval (660e900 years) in this study. Approximately half of the recurrence interval on both the Ikeda and Chichio faults has passed since the latest faulting event occurred at the end of the Middle ages of Japan history (16th century or AD 1596) (Morino and Okada, 2002; Okada and Tsutsumi, 1997). This suggests that the next large earthquake along the
central portion of the MTLAFZ may not occur in the distant future, but sooner. The long-term seismic risk of the Ikeda and Chichio faults in the next 30 years according to the method of Headquarters for Earthquake Research Promotion (2001), was evaluated to be more than 0.4e1.9%, 0.1e3%, respectively, which is much larger than that of Headquarters for Earthquake Research Promotion (2011) (0e0.3%). Paleoseismological studies such as trenching surveys are required to confirm the results of this study. Acknowledgements This work was partly supported by MEXT KAKENHI grant numbers 16K01221, 25350428, and 22700855. The author is grateful to Professor Koji Okumura for providing valuable comments on my research, and Mr. Hiroaki Matsugi who supported my research in the field. The author also thanks the reviewers for careful reading of the manuscript and for giving fruitful comments and suggestions.
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013
H. Goto / Quaternary International xxx (2017) 1e11
References Aoki, K., 2008. Revised age and distribution of ca. 87 ka Aso-4 tephra based on new evidence from the northwest Pacific Ocean. Quat. Int. 178 (1), 100e118. Aoki, Y., Scholz, C.H., 2003. Interseismic deformation at the Nankai subduction zone and the median tectonic line, southwest Japan. J. Geophys. Res. 108 (B10), 1065e1070. Bronk Ramsey, C., 2016. OxCal Program, V. 4.2 Radiocarbon Accelerator Unit. University of Oxford, UK program accessible online at. http://c14.arch.ox.ac.uk/ oxcal.html. (Accessed October 2016). Chen, J., Liu, Y., Shi, X., Suk, B., Yao, Z., 2017. Climate and environmental changes for the past 44 ka clarified by pollen and algae composition in the Ulleung Basin, East Sea (Japan Sea). Quat. Int. 441, 162e173. Fitch, T.J., 1972. Plate convergence, transcurrent faults and internal deformation adjacent to southeast Asia and the western Pacific. J. Geophys. Res. 77 (23), 4432e4460. Furuta, N., 1996. Development history of the Yoshino and Agui River lowlands, northeastern part of Tokushima prefecture, Japan. Ritsumeikan Chirigaku 8, 61e72 (in Japanese). Goto, H., 1996. Fault morphology of the lyo Fault, the Median Tectonic Line active fault system. Act. Fault Res. 14, 28e38 (in Japanese, with English abstract). Goto, H., 2016. Extensive area of topographic anaglyphs covering inland and seafloor derived using a detailed digital elevation model for identifying broad tectonic deformations. In: Kamae, K. (Ed.), Earthquakes, Tsunamis and Nuclear Risks; Prediction and Assessment beyond the Fukushima Accident. Springer, Tokyo, Japan, pp. 65e74. Goto, H., Nakata, T., Tsutsumi, H., Okumura, K., Imaizumi, T., Nakamura, T., Watanabe, T., 2001. The latest surface-faulting events on the Median Tectonic Line in Shikoku, southwest Japan, based on mini-trenching and geoslicer studies. Zisin: J. Seismol. Soc. Jpn. 53, 205e219 (in Japanese, with English abstract). Goto, H., Tsutsumi, H., Toda, S., Kumahara, Y., 2017. Geomorphic features of surface ruptures associated with the 2016 Kumamoto earthquake in and around the downtown of Kumamoto City, and implications on triggered slip along active faults. Earth Planets Space 69, 26. Hashimoto, M., Kanmera, K., 1991. Pre-Neogene sedimentary and metamorphic rocks. In: Hashimoto, M. (Ed.), Geology of Japan. Terra Scientific Publishing Company, Tokyo, pp. 13e55. Headquarters for Earthquake Research Promotion, 2011. Evaluations of the Median Tectonic Line Active Fault System (in Japanese). http://www.jishin.go.jp/main/ chousa/katsudansou_pdf/81_83_85_86_89_chuo_2.pdf. (Accessed October 2016). Headquarters for Earthquake Research Promotion, 2001. Evaluation Method for Long-term Earthquake Occurrence Probability (in Japanese). http://jishin.go.jp/ main/choukihyoka/01b/chouki020326.pdf. (Accessed October 2016). Hirakawa, K., Ono, Y., 1974. The landform evolution of the Tokachi Plain. Geogr. Rev. Jpn. 47, 607e632 (in Japanese, with English abstract). Kaneko, S., 1977. Recent formations and their basal topography in and around Tokyo Bay, central Japan. Quat. Res. 8, 32e50. Kaneko, S., 1966. Transcurrent displacement along the median line, south-western Japan. N. Z. J. Geol. Geophys. 9 (1e2), 45e59. Katayanagi, Y., 2016. Simple DEM Viewer. http://www.jizoh.jp/english.html. (Accessed October 2016). Mizuno, K., Okada, A., Sangawa, A., Shimizu, F., 1993. Explanatory text of strip map of the Median Tectonic Line active fault system in Shikoku, Japan, scale 1: 25,000. Tectonic map series (8). Geol. Surv. Jpn. 63 (in Japanese, with English abstract). Morie, T., Ozawa, H., Okumura, K., 2001. Description and correlation of the Nagate Tephra found in the terrace deposits at the southern foot of Asan Range in Tokushima Prefecture, southwest Japan. Quat. Res. 40 (4), 331e336 (in Japanese, with English abstract). Morino, M., Okada, A., 2002. Faulting history of the Median Tectonic Line active fault
11
system in Shikoku, based on re-examination of trench survey results. In: Annual Report on Active Fault and Paleoearthquake Researches, vol. 2, pp. 153e182 (in Japanese, with English abstract). Nakata, T., Imaizumi, T., 2002. Digital Active Fault Map of Japan. University of Tokyo Press, Tokyo, Japan (in Japanese, with English abstract). Okada, A., 1968. Strike-slip faulting of late quaternary along the median dislocation line in the surroundings of Awa-Ikeda, northeastern Shikoku. Quat. Res. 7 (1), 15e26 (in Japanese, with English abstract). Okada, A., 1970. Fault topography and rate of faulting along the Median Tectonic Line in the drainage basin of the river Yoshino, northeastern Shikoku, Japan. Geogr. Rev. Jpn. 43 (1), 1e21 (in Japanese, with English abstract). Okada, A., 1980. Quaternary faulting along the median tectonic line of southwest Japan. Memoirs Geol. Soc. Jpn. 18, 79e108. Okada, A., Sangawa, A., 1978. Fault morphology and quaternary faulting along the median tectonic line in the southern part of the Izumi range. Geogr. Rev. Jpn. 51 (5), 385e405 (in Japanese, with English abstract). Okada, A., Tsutsumi, H., 1990. Date on fault exposures of the Median Tectonic Line and geopmorphic surface chronology in the central to eastern part of Shikoku, Southwest Japan. Act. Fault Res. 8, 31e47 (in Japanese). Okada, A., Tsutsumi, H., 1997. Holocene Activity of the Chichio fault, the median tectonic line Active fault system, southwest Japan, based on trenching studies. J. Geogr. 106 (5), 644e659 (in Japanese, with English abstract). Okayama, T., 1988. 1: 1,000,000 the Summit Level Map of Japan. Kokon Shoin, Tokyo (in Japanese). Reimer, P.J., Bard, E., Bayliss, A., Beck, J.W., Blackwell, P.G., Bronk Ramsey, C., Buck, C.E., Cheng, H., Edwards, R.L., Friedrich, M., Grootes, P.M., Guilderson, T.P., , C., Heaton, T.J., Hoffmann, D.L., Hogg, A.G., Haflidason, H., Hajdas, I., Hatte Hughen, K.A., Kaiser, K.F., Kromer, B., Manning, S.W., Niu, M., Reimer, R.W., Richards, D.A., Scott, E.M., Southon, J.R., Staff, R.A., Turney, C.S.M., Plicht, J., 2013. IntCal13 and Marine13 radiocarbon age calibration curves 0-50,000 years cal BP. Radiocarbon 55 (4), 1869e1887. (revised edition). In: Research Group for Active Faults of Japan (Ed.), 1991. Active Faults in Japan: Sheet Maps and Inventories. University of Tokyo Press, Tokyo (in Japanese, with English abstract). Smith, V.C., Staff, R.A., Blockley, S.P.E., Bronk Ramsey, C., Nakagawa, T., Mark, D.F., Takemura, K., Danhara, T., Suigetsu 2006 Project Members, 2013. Identification and correlation of visible tephras in the Lake Suigetsu SG06 sedimentary archive, Japan: chronostratigraphic markers for synchronising of east Asian/ west Pacific palaeoclimatic records across the last 150 ka. Quat. Sci. Rev. 67, 121e137. Tabei, T., Hashimoto, M., Miyazaki, S., Hirahara, K., Kimata, F., Matsushima, T., Tanaka, T., Eguchi, Y., Takaya, T., Hoso, Y., Ohya, F., Kato, T., 2002. Subsurface structure and faulting of the Median Tectonic Line, southwest Japan inferred from GPS velocity field. Earth Planets Space 54, 1065e1070. Takahashi, M., 2006. Tectonic boundary between northeast and southwest Japan arcs during Japan sea opening. J. Geol. Soc. Jpn. 112 (1), 14e32 (in Japanese, with English abstract). Tsutsumi, H., Goto, H., 2006. Surface offsets associated with the most recent earthquakes along the Median Tectonic Line active fault zone in Shikoku, southwest Japan. Zisin J. Seismol. Soc. Jpn. 59 (2), 117e132 (in Japanese, with English abstract). Tsutsumi, H., Okada, A., 1996. Segmentation and Holocene surface faulting on the median tectonic line, southwest Japan. J. Geophys. Res. 101 (B3), 5855e5871. Tsutsumi, H., Okada, A., Nakata, T., Ando, M., Tsukuda, T., 1991. Timing and displacement of Holocene faulting on the median tectonic line in central Shikoku, southwest Japan. J. Struct. Geol. 13 (2), 227e233. Tsutsumi, H., Okada, A., Goto, H., Matsugi, H., 2000. Late Holocene faulting along the Kawakami fault, an active segment of the median tectonic line, southwest Japan. Act. Fault Res. 19, 77e86 (in Japanese, with English abstract). Yokoyama, T., Matsunami, S., Okumura, K., 1990. The formative process of alluvial deposits in the Tokushima plain. J. Geogr. 99, 775e789 (in Japanese, with English abstract).
Please cite this article in press as: Goto, H., Late quaternary slip rates and vectors on the Median Tectonic Line active fault zone in eastern Shikoku, southwest Japan, Quaternary International (2017), https://doi.org/10.1016/j.quaint.2017.12.013