We proposed a new method on 3D geological modeling and visualization of strike-slip fault structure using previously reported three-dimensional trenching data. The trenching data was acquired from the survey on the Tanna fault of the Northern Izu active fault system, which generated the Kita-Izu earthquake in 1930 AD. The geological information exposed on the walls was recorded as the logs with boundaries of sedimentary units and faults in the field, and they are converted to raster graphics with RGB color information and geographical coordinates on the computer. The raster information was extracted as point cloud data in rectangular coordinate system, then 2-cm-grid depth data for sedimentary unit boundaries and fault planes were produced with complementary spline methods. Based on this grid data, surface modeling for both sedimentary unit and fault planes was carried out using ArcGIS and its extensions. In order to distribute these 3D geological models and animations for readers, we converted them to U3D format files and PDF files which have high coverage. As a result, we could successfully re-construct three-dimensional geological structure and re-confirmed the existence of penultimate paleoearthquake event prior to the 1930 Kita-Izu earthquake on the Tanna fault. Since the penultimate event occurred ca.500-600 y. B. P and had shorter recurrence interval, the event had been considered as low-reliability. However, the fact that we could re-identify the existence of this event arises a possibility of future surface deformation on the Tanna fault associated with its activity or with triggered slip by adjacent seismic sources. Because the tunnel of railway line penetrates into the Tanna fault, this possibility is not ignorable. Thus, our proposed method for modeling and visualization of three-dimensional trenches is useful for quantitative analysis and representation of more objective and practical geological data.

View full abstract

For investigating paleoearthquakes, dating is important. Generally the¹⁴C dating method is employed to date earthquakes that have occurred during the historical and/or the archaeological ages. The archaeomagnetic dating method can also be employed, which uses the remanent magnetization of objects and geomagnetic secular changes. The archaeomagnetic dating method is practical in Japan because detailed secular variations in the geomagnetic field during the past 2000 years have been clarified by Hirooka (1971), etc. In this study, we discuss the archaeomagnetic dating conducted in two regions: the Atotsugawa active fault region in Gifu Prefecture, where the sediment near the main fault zone was studied and the archaeological Ishizuka site of Takaoka city in Toyama Prefecture, where the sand boiling was studied.
In the Atotsugawa active fault region, archaeomagnetic dating was successfully adopted and it yielded an age of A. D.1850-1900, which corresponds to the earthquake of“ Ansei jishin” that occurred in A. D.1858. At the Ishizuka archaeological site, archaeomagnetic dating on sand boiling yielded an age of A. D.863the year in which“ Jogan jishin” occurred around Toyama Prefecture.
We conclude that archaeomagnetic dating is useful for the dating of active faults and further study on it is desirable.

View full abstract

We devised an easy-to-use, small, and light field mapping instrument named‘ Handy Station’, composed of a compact laser range finder fixed on a revolving base on a tripod. This device is essentially of the same function as total stations, and enables us to measure horizontal distance, vertical distance, and horizontal direction of objects. Handy Station costs one third to one tenth of general total stations.
The compact laser range finder used for the device is Impulse 200 or TruPulse 200, with precisions of 0.01m and 0.1m for horizontal and vertical distance, respectively. We read horizontal direction toward targets by 0.1 degree on 1-degree marks plotted around the revolving tables. Thus, the accuracy of measurements by Handy Station is expected much less than that of total stations with millimeter and minute precision. Handy Station also does not accurately measure distance beyond 100 meters, while total stations accurately measure distance over 1 kilometer. However, for topographic mapping in fields, we seldom need such precision, but we need more maneuverability, especially when we use such instruments in a remote place where we have to carry them for distance. Handy Station is very compact and is about 2.5 kilograms for Impulse 200, and 1.5 kilograms for TruPulse 200, while total stations with tripod are bulky and weigh more than 10 kilograms. We can use an inexpensive reflector bought at a dollar shop for Handy Station.
Total stations automatically record readings that can be processed by a computer for drawing maps. On the other hand, we take notes for readings when we use Handy Station. It is recommendable to make notes and sketch maps around targets, in order to avoid misunderstanding about locations.
We compared performance of Handy Station with that of total stations by mapping a part of fault scarp along the Atera fault, central Japan. Time for mapping did not differ much among the instruments. Results were processed by Excel to get coordinates from horizontal distance and direction, and then contour lines were drawn by Surfer. As a result, it is indicated that the measurement by Handy Station can produce accurate maps that are precise enough for maps at a scale of one-to-several hundreds.

View full abstract

In contrast to strike-slip faults, relationship among geometrical features and controls on fault rupture, growth, and segment linkage of reverse faults is poorly known although the detailed mapping of active faults and its data on GIS format enable us to analyze characteristics of their surface geometry. It is well known that reverse faults show sinuous surface trace. We analyzed geometric characteristics of 35 major reverse fault zones in Japan based on Digital Active Fault Map of Japan (Nakata and Imaizumi ed.,2002).
We measured distance between both ends on individual fault zone (L) and fault trace (1), and distance from the line connecting the ends to the furthest point on the arcuate zone (H) and the trace (h), and compare their curvature (H/L or h/1). Values of L, H,1 and h were distributed in range of 10∼ 80 km,2∼ 10 km,1∼ 10 km and 0.1 -.5 km respectively. The curvature (H/L) of the reverse fault zones is, regardless their length,0.1∼ 0.3, and that (h/l) of the fault traces is 0.02∼ 0.3, although it is 0.1∼ 0.3 among most of the traces. These results show a fractal-like hierarchy in a curvature of the reverse fault.

View full abstract

It implemented A-ray survey as the technique of the geophsical exploration which detects a fault and it detected fracture width. The following relation is admitted among the fracture width and the vertical displacement. The more vertical displacement (Δ H) is, the more fracture width (FW) gets widely. The relational expression is admitted by vertical displacement (Δ H) and fracture width (FW) Δ [H=3∼ FW]. After verifying the relational expression about the Emmyouji fault, the vertical displacement which is calculated from fracture width and the vertical displacement which was made clear in the boring exploration are same. The result of reviewing average slip rate from the relational expression with fracture width, the active fault number ratio which is extracted at present, the active fault number ratio which is calculated from the relational expression with fracture width and the earthquake occurring number ratio are 1: 3: 5. It is necessary to reconsider the activity standard (average slip rate) for the solution of the C class active faults problems about the number of the active fault existence.

View full abstract

In our previous paper, we used an eight-parameter formula to fit Kumamoto's data of fault length: “ Maxlen model” and “ Segment model”. We propose here a revised formula with four parameters determined by the maximum lilelihood estimation of α and β for the gamma distribution and C and D for the rate of fault segmentation. It is proved this formula fits well both the Maxlen and Segment models. The present study can be applied to the earthquake magnitude statistics through an empirical relation between magnitude and logarithmic fault length. We found that the magnitude distribution derived from the Maxlen model makes a good agreement with that of inland earthquakes (M>6.5)occuned in Japan in the past 400 years.

View full abstract

We compiled a database on fault slip distribution along the 98 major active fault zones in Japan selected for National Seismic Hazard Maps for Japan by The Earthquake Research Committee, the Headquarters for Earthquake Research Promotion, Data were collected from articles and reports on these faults published prior to 2001 as well as‘ Digital Active Fault Map of Japan’ (Nakata and Imaizumi, eds,2002), and then compiled as an Excel file and processed into GIS format. The database contains location, vertical and horizontal displacement and estimated slip rates together with method of measurement of fault-related features and their ages. Each feature is ranked A to D based on its reliability and accuracy. The database on several reverse faults in northeast Japan and strike-slip faults in southwest Japan are graphically represented by Mapinfo, using Shp file of active fault traces of‘ Digital Active Fault Map of Japan’, as well as historical earthquake faults.
This database is the first product of its kind, but is still far from substantial level, especially on slip rate data. However, we believe that this database is useful to roughly locate asperities on active faults for better estimation of strong ground motion.

View full abstract

bility to the BPT model. As a result, we conclude that case (c) is the most realistic model by the AIC procedure.
Also, we simulated the value by randomly selecting the event occurrence between the range of the recurrence interval and verified the variable estimated from the mean recurrence interval. As a result, the value 0.24 estimated from the four active faults or the value 0.42 estimated in this study is mostly smaller than the average value that can be practically estimated from data. Also, we show some case that is possible to estimate the variable more accurately by considering the probability density distribution of the event occurrence.

View full abstract

For more practical seismic hazard assessment based on the existing active faults data, we need to construct a better integrated earthquake source model. Previous assessments in Japan such as Headquarters for Earthquake Research Promotion (HERP,2006) were inadequate in estimation of future earthquake size (magnitude) distribution mainly due to lack of consideration on active faults less than 10 km long, and segmentation of extensive active fault systems. In order to cope with these issues, we propose a new method to generate an integrated source model by joining several existing active fault databases, and created a link-files to Digital Active Fault Map of Japan edited by Nakata and Imaizumi (2002) that provides most detailed active fault distribution nation-wide, by combining the attribute tables of “ Active Fault Map of Japan” (Active Fault Research Group,1991), that of HERP,2006), and that of“ Active Fault Research Center, National Institute of Advanced Industrial Science and Technology” (2005). A new linear source model for active faults we propose here for seismic hazard analyses and strong ground motion prediction, simplifies complicated fault traces into straight lines. The criteria for grouping and segmentation of fault traces, so-called 5-km gap threshold (Matsuda,1990) and 2-km gap threshold (Wesnousky,2006) were used in the Geographical Information System (GIS). Through these procedures approximately 400 active fault lines less than 10 km long are identified and listed in the new database. Magnitude-frequency distribution based on this database and the ellipsoidal model of magnitude estimation (Okino and Kumamoto,2006) better fits the Gutenberg-Richter relation than that based on the database of the HERP and the empirical magnitude estimation formula of Matsuda(1975), particularly for events of magnitude around 7.0.

View full abstract

The parameterization of active faults in seismic hazard prediction still involves many uncertainties. The aim of this study is to apply logic-tree analysis, which has been widely used in engineering research, to seismic hazard prediction, in order to refine earthquake magnitude estimation. In this study, earthquake magnitude estimation takes into account the results of the Headquarters for Earthquake Research Promotion (HERP) and focuses on the Itoigawa-Shizuoka tectonic line (ISTL) using a logic-tree consisting of four branches: geometric segmentation, a coupling scenario, an empirical magnitude estimation equation and large earthquake scaling. Depending on whether the branch deals with epistemic or aleatory uncertainty, branch weight was subjectively allocated by the Technical Facilitation Integrator (TFI) under the direction of the Technical Integrator (TI), a panel consisting of six specialists. An important outcome of this study is the realization of how quantitatively different specialists' opinions are with respect to the uncertainties. The logic-tree, including weights for the individual TIs and the IF1, calculates the 50-year conditional probability of an event with seismic intensity of 6.5 or higher (JMA scale) for the cities of Matsumoto and Kofu. The results are compared with those of the HERP. Although there is no significant difference between the results for Matsumoto, differences in the results for Kofu differ by a factor of 2.2 to 5.0 times. The main reasons for the differences are: (1) the evaluation of earthquake occurrence probability under some coupling scenarios for the Gofukuji fault; and (2) uncertainty whether the total length of the ISTL could rupture simultaneously. Finally, in order to assess the influence of epistemic uncertainties (geometric segmentation, empirical magnitude estimation equation and large earthquake scaling), the branches are compared by calculating average hazard curves for each branch. The results show that the most important of the branches is large earthquake scaling.

View full abstract

We intend to investigate the activities of the Hanawa higashi fault zone (eastern marginal faults of the Hanawa basin) at Northeast Japan. We obtained the displacement of the faults on terrace surfaces by the interpretations of aerial photographs, field survey and arrayed borings.
Fluvial terrace surfaces are divided into higher (H), and lower (L1-L8) surfaces in descending order of ages. Ages of the surfaces are estimated from tephra stratigraphy and¹⁴C dating. The L2 surface is the depositional surface of Towada - hachinohe pyroclastic flow deposits which include 15 ka carbonized woods. The L6 and the L7 surfaces are estimated to be 10 ka and 4 ka respectively, from¹⁴C age of terrace deposits. Thus the L3 - L5 surfaces are estimated to be 10-15 ka. The L8 surface is younger than 4 ka.
Two major faults, Fe 1 and Fe 2, comprise the Hanawa higashi fault zone. These faults dislocate the L2 surface (15 ka) and the L3 surface (10-15 ka). The Fe 2 does not show apparent dislocation on the L7 surface (4 ka). These show that the Fe 2 dislocated after 10-15 ka and the Fe 2 is estimated not to have dislocated since 4 ka.
Vertical slip rate of the Fe 1 is calculated to be ca.0.1 m/ky from the offset of the L2 surface. Vertical slip rate of the Fe 2 is calculated to be 0.3-0.5 m/ky from the offsets of the L2 and the L3 surfaces.

View full abstract

The Kushigata Range active fault zone is located at northeastern margin of the Echigo (Niigata) plain, central Japan, its length is 16 km and its mean vertical slip rate is about 0.2 to 0.4 m per 1000 years. We excavated two trenches across the Kajikawa fault, one of the Kushigata Range fault zone. Three faulting events are identified at both two trenches. The latest faulting occurred between BC1,200 and BC520, penultimate events was between BC4,850 and BC3,635, and the earlier event was between BC9,210 and BC7,330. Vertical slip of the latest and penultimate events are 0.7 m and about 1 m. Timing of the latest event is concordant to that of liquefaction event observed at the archaeological cites in the northern part of the Echigo (Niigata) plain.

View full abstract

On the Arakawa Alluvial fan, northwestern part of the Kanto Plain, late Quaternary fluvial terraces are widely distributed along the Arakawa river and a part of them are deformed by the Fukaya fault. The ages of the terraces have rarely been reported in this area, due to the scarcity of outcrops and lack of visible tephras. In this paper, we carried out a drilling survey on terraces to obtain samples of the overlying loamy soil and upper part of terrace deposits. From these samples, we extracted some well-known widespread volcanic ash, from which we were able to estimate the approximate age of the terraces and the vertical slip rate of the active faults. Late Quaternary fluvial terraces in the Arakawa fan are divided into 5 levels: from the Osato terrace to the Hanazono terrace in descending order. Widespread tephras such as the Hr-Hp (50ka) is contained in the lowest part of the loamy soil in the Kushibiki terrace and the As-BP (20-25ka) is present in the lower part of the loamy soil in the Miizugahara 1 terrace. From the ages and the vertical displacements of the fluvial terraces, the late Quaternary average vertical slip rates of the Fakaya fault are estimated to be 0.25-0.30 m/kyr (Kushibiki terrace, the minimum value), and 0.36-0.45 m/kyr (Miizugahara 1 terrace), respectively.

View full abstract

We conducted a tectonic geomorphological survey in the northeastern margin of the Suwa Basin (in the area between Okaya and Chino), along the middle part of the Itoigawa-Shizuoka Tectonic Line Active Fault System (ISTL). Procedure of this study is as follows; 1) aerial photo analysis and field survey,2) mapping of geomorphic surfaces and reconstruction of geomorphic evolution,3) mapping of tectonic landforms which explains the geomorphic evolution reasonably,4) measurement of vertical offset by construction of cross sections and calculation of vertical offset rates,5) measurement of horizontal offset of geomorphic markers (e. g., terrace scarps, small valleys) and calculation of horizontal offset rates. Vertical/horizontal offset rates were estimated based on offset of geomorphic markers (fluvial terraces) and ages of the terraces (H, older than 120 ka; Ml,90-100 ka; M2,40-65 ka; Li a,20 ka; Lib,10 ka; L2,4-7 ka; L3,1-2 ka). Although mapped faults in this study are similar to existing papers, different results were obtained in some areas as follows; 1) we judged fault traces in existing papers near the Suwa Taisha Harumiya which lies on the right bank of the Togawa River is not a tectonic landform but an erosional scarp,2) we judged that broad gentle slope at Suwa Taisha Akimiya which is interpreted as flexural scarp is not a tectonic landform but a depositional landform of landslide deposits,3) we judged that the (fault) scarp in Kuwabara is not a tectonic landform but an erosional scarp,4) fault scarps which lie along boundary between the Suwa Basin and mountainous area at Uehara are newly mapped,5) the fault scarp near the Chino station at the right bank of the Kamikawa River is newly mapped as certainly located fault,6) the fault scarp which deforms L3 surface in a valley bottom at Sakamuro is newly mapped. We constructed 41 cross sections and obtained 47 vertical/lateral offset rates. Based on obtained vertical/horizontal offset rate, faults in the study area can be classified into three or four segments, although we could obtain only minimum vertical offset rates in many areas. The most southeastern segment is interpreted as a main fault which is inferred to construct a pull-apart basin. The other fault segments are interpreted as secondary faults of the main faults, based on the fact that these fault segments have short length (3-4 km), and that the fault segments have gentle round shaped fault traces.

View full abstract

We conducted a tectonic geomorphological survey in the southern part of the Matsumoto Basin and the south coast of the Suwa Basin, along the middle part of the Itoigawa-Shizuoka Tectonic Line (ISTL) active fault system. Procedures of this study are as follows; 1) aerial photo analysis and field surveys,2) mapping of geomorphic surfaces and reconstruction of geomorphic evolution,3) mapping of tectonic landforms which explain the geomorphic evolution reasonably,4) photogrammetrical measurement of vertical offset by construction of cross sections and calculation of vertical offset rates,5) measurement of horizontal offset of geomorphic markers and calculation of horizontal rates. Vertical/lateral offset rates were estimated based on vertical/lateral offset of geomorphic markers, such as terraces and ages of them (H: older than 120 ka, M1: 100 ka, M2: 40-65 ka, Lla: 20 ka, Lib: 10 ka, L2: 4-7 ka, L3: 1-2 ka). LiDAR method is also used for measurement of lateral offset
Although mapped fault trace in the study area is similar to existing papers, different results were obtained in some areas. One of the remarkable results is that the Gofukuji fault in eastern margin of the Matsumoto Basin connects smoothly with the faults in northern end of the Suwa Basin. These faults have left lateral offset.
We obtained 108 vertical/ left lateral offset rates in the study area. The vertical and leftlateral slip rates are estimated to be 0.1-2.0 mm/yr and 0.6-8.5 mm/yr, respectively.

View full abstract

We analyzed the profiles of the ground-penetrating radar (GPR) performed before the trenching study across the Gero fault in the Atera fault system, and reviewed the reflection pattern of subsurface structure around the active fault. The GPR profile eliminated fake images by the migration process have enabled to interpret the reflection pattern more exactly. Comparing the GPR profile with the trench wall in the Mimayano site, it was proved that the active fault is detected as a sheared reflection or a linear gap. Besides, it was able to detect the sedimentary structure of the depression formed between two faults through the indeterminable depth by trench. Thus, GPR survey can detect the active fault in this area, and it is available for the method to get the additional information of trenching study.
The GPR profile performed above the outcrop of the Atera fault showed the geologic structure clearly, however, the active fault could not be detected according to a few snow cover of surface and anomalous landform of outcrop.
In the Norimasa site, where is able to estimate the place of the Gero fault from a landform, the GPR profiles showed a sheared reflection same as the Mimayano site at the estimated place. Besides, a complex structure confused from a landform was presumed from the profiles. The GPR survey is effective method to elucidate the subsurface structure around an active fault and support with trenching study.

View full abstract

The Yugamine fault,10 km in length, is in the north-central part of the Atera fault zone. We studied Holocene activity of the Yugamine fault through a trenching survey. We excavated a trench across the scarplet on the L 1 terrace at Obayashi district. Some humic layers and fluvial gravels are exposed on the trench walls. The Yugamine fault has cut through these layers to form a linear depression on its NE side. From the structural evidence along the fault plane and angular unconformity, we recognize at least four faulting events in this trench.

View full abstract

The Sanshukaido fault is located at the southern part of the Ina - Valley fault zone, trending in the NE - SW direction and accompanying late Quaternary activity. In this study, we excavated a trench across the Sanshukaido fault at Samuhara site in order to reveal the timing of recent faulting events. On the trench walls across the Sanshukaido fault, a layer of humic soil is displaced by a low angle reverse fault. This excavation revealed evidences for the latest and the penultimate events associated with at least two surface-ruptures. The latest event might be occurred sometimes after AD1408, and the penultimate event occurred between BC6587 and BC1753. Recurrence interval of this fault is estimated to be 3160-8590 years.

View full abstract

At Gerede Ardicli trench site on the North Anatolian Fault System (NAFS), ages of sedimentary beds cut by fault were estimated by archaeomagnetic method. The beds exposed on the trench walls were deposited under marsh conditions after 6th centuries (Okumura et al.,2003). The discrete samples were collected from 16 horizons analyzed after alternating field (AF) demagnetization for three components (declination, inclination, and intensity of remanent magnetization). As no geomagnetic secular variation data for archaeomagnetism was available in Turkey, the results were compared with those obtained in Bulgaria. C-14 ages (Okumura at al.,2002) and an archaeomagnetic age of the brick (Kanai at al.,2005) were used as the reference data to constrain the age ranges. Data for declination and normalized remanent magnetic intensities for seven horizons corresponds well to the Bulgarian data. Thus, age estimation based on archaeomagnetic study will provide us effective age constraint of paleoearthquakes.

View full abstract