This is a proposal of introducing the probabilistic indication system of maximum earthquake magnitude into a seismic zoning map. In the previous method of seisic zoning, we use the upper limit of magnitude presumed from fault lengths or historical records of seismic damages. In the present study, we convert the indication system from determination to probability. Practical applications of the new s y s tem are made to magnitudes transformed by M atsuda's formula from seismogenic fault lengths. Assigning Makjanic and Gumbel statistical models to the magnitudes such obtained, we express maximum magnitude in a form of, μ±σ, where μ is the statistically most plausible value of magnitude and σ is its deviation.
An extensive active fault system that obliquely cuts across the Himalaya extends from the Karakoram Fault in the northwest to the newly found strike-slip fault system within the Bengal Plain in the southeast. Along this fault system, right-lateral displacement is commonly recognized from offset of streams and ridges in the Himalayan range with vertical slip down to the north. Holocene activities along the fault are evidenced on the outcrops in the trench or in the field, confirming it as a potential source of future destructive earthquakes. It may be considered that this fault system was formed by the initial stage of the Indo-Eurasian plate collision, and has been modified by large-scale thrusting along the Himalayan front.
Several active and seismic faults are sporadically distributed in the San'in (northern Chugoku) District in the western part of Honshu, Japan. As the detailed surveys including trench excavations for these active faults have been recently carried out, common characteristics among these faults have clarified as follows: 1) The features of the fault morphology in this district are not so obvious, and cum u lative offsets are always smaller than a few hundreds of meters. 2) All active faults are strike-slip type, in w hich arcuate ENE-WNW trending ones are dextral, and NNWSSE trending are sinistral. These are distributed within the active moble zone between the north s i d e of the Chugoku Mountains and the southern Japan Sea fault zone (SJSFZ). 3) Each active fault (zone) is composed of a few to several fault s arranged with en echelon segment or accompanies the diverged branch. Although linear or slightly curving trace extends continuously, th e total length of each fault is shorter than 25km. 4) The fault shattered zone is usually not w i de and narrower than a few 10 cm. And these faults do not indicate the reactivity along the pre-existed wide shattered zone 5) The newly developed active fault system has been form e d under the stress field with maximum compressional axis of WNW-ESE direction since the middle Quaternary. The matuarity of ea c h fault is lain under the young stage. 6) As even indistinct lin e a ment or active fault has potential to produce large earthquake of M7 class in this district, the careful further more detailed surveys are required. 7) This centennial is the seismically active period in this district after the quiescent recurrence interval of about one thousand of years.
A boulder on the excavated trench wall just upper part of the Makigahora active fault, a part of the Takayama Oppara active fault zone, in Gifu Prefecture, Japan, is observed. The existence of boulders just upper part of, or several meters apart from, lateral active faults is sometimes observed on the wall of trench. When the sediments are composed of terrace deposit or fan deposit or debris flow, boulders are naturally found. In case of swamp deposit how boulders are derived? The author hypothesize that boulders a re concentrated by uprhrow in near field strong ground motion in earthquakes. The field survey is conducted along the part of the Makigahora active fault. As a result of the field survey, boulders are not accidentally concentrated n e a r the fault line, especially several tens meters near the fault line. Many surveys along surfacer uptures of earthquakes clarified that the distribution of upthrow of boulders corresponds to the area of near field strong ground motion. When boulders appear just upper part of, or s e v eral meters from, the active fault, especially within swamp deposit, we should treat them carefully as an evidence of the occurrence of strong earthquake. Plants under the boulders show the time of the strong earthquake. The distribution of boulders along active faults might show the strong ground motion area.
An active reverse fault zone, about 50 km long trending NW-SE, that is known by having caused the Zenkoji earthquake in 1847, exists at the northwestern boundary of Nagano basin, central Japan and the southwestern part of this fault zone, is named Komatsubara fault. As the result of precise air-photo interpretation of tectonic landforms, surface rupture traces which occurred during the 1847 Zenkoji earthquake were found along the Kamatsubara fault. The traces consist of east-facing small and fresh flexure scarps on lower terrace and alluvial plain and extend for about 3 km at least. Such evidence indicates that the Kamatsubara fault thrust up toward the east in 1874, and then the ground surface was displaced 2 m or more in vertical. In this area, we can recognize too flexure scarps formed by the fault event just before the 1847 Zenkoji earthquake.
Since Professor Tokihiko Matsuda had taken his post to Kyushu University, the active fault survey in Kyushu has progressed, and it has obtained much information on the active faults in Kyushu. Kamegawa fault located in the eastern part of central Kyushu has investigat e d on 1997. It was a first trench survey on active fault in Beppu-Haneyama fault zone. Then active fault survey of Oita Prefecture has started in the Beppu-Haneyama fault zone, and it is continuing now. These surveys are the important investigations of the dense normal fault zone in Japan. By these surveys, it is sure to clarify the late Quaternary tectonics of central Kyushu. In northern Kyushu, recent activity of Fukuchiyama fault inverted from the normal fault movement with mountainous area upheaval to the reverse-fault activity with upheaval of the plain. The trenching surveys of Nishiyama and Kego faults showed left strike-slip movements. In southern Kyushu, Hitoyoshi basin south e rn margin fault was newly found as the fault bordered between mountain and alluvial fan in Hitoyoshi basin. The latest activity of the Izumi fault with characteristically right-lateral displacement was also clarified to predominate the normal faulting.
Topographic deformation caused by the“Yufutu Fold and Thrust”was examined, using the large scale topographic maps. The Shikotu pyroclastic flow surface, ca.40-42 ka, has already been deformed to show upwarped landforms. This deformation should coincide with the most active zone of the so called“Hidaka collision zone”at present.
Active fold belts in Japan are likely to be underlain by detachment faults with flat-ramp geometry. These fold belts fall into two categories: (1) the tectonic inversion type and (2) the foreland fold-and-thrust belt type. The former includes the Uetsu and the Northern Fossa Magna fold belts, which are located in Miocene rift basins. The normal faults that formed the rift basins have rejuvenated as thrust faults due to regional compression since the Pliocene, resulting in folding in a thick pile of rift sediments. The latter category includes the Teshio-Ishikari and the Southern Fossa Magna fold belts. They are located in foreland basins that have been formed since Neogene time up to the present in association with arc-arc collision. The master faults underlying these fold-and-thrust belts are likely to extend down-dip to the plate boundary faults.
Professor Tokihiko Matsuda took an initiative in the compilation project of active faults map of Japan early 1970's, which eventually materialized as“Active Faults Map of Japan”in 1980. Mapping of active faults has significantly developed since then following scientific and technological progress in this research filed. In 2002,“Atlas of Quaternary Thrust Faults in Japan”and“Digital Active Faults Map of Japan”were published in the scale of 1: 25000. These maps provide us with detailed information regarding their location and characteristics of activities for basic data of other science fields and seismic disaster mitigation. It is, however, important to know that criteria as“active”for the active faults and aims for compilation of the maps vary each other as well as from previously published maps. Future active faults maps are expect e d to be GIS maps combining surface geomorphological and subsurface seismo-geological data, which enable us to better simulate future inland earthquakes.
While large-scale active fault maps are being published to present important informations for the earthquake disaster mitigation, it is necessary to estimate the accuracy of locality information shown on the maps. When the locations are digitized based on topographical maps of 1: 25,000, it is impossible to eliminate an error of about 20m which came from the accuracy of the based maps. In order to improve the accuracy of the data, we propose here a methodology for making digital locality information by use of photogrammetrical system. As it is necessary to observe original landforms deformed prior to land modifications, aerial photographs taken in 1940s' by the U. S. Armed Forces are often used for active fault mapping. However, the photographs have been much distorted, and the mapping by the photographs are often problematic and some countermeasures should be considered carefully. By utilizing the method presented here, the error of locality information can be minified within 5-7m horizontally.
The recent activities of the Kokura-higashi Fault, northeast Kyushu is investigated on the basis of the interpretations of aerial photographs, field surveys, leveling of topographic profiles and trenching surveys. The Kokura-higashi Fault extends with a NNE-SSW trend for about 17km and the fault tra c e is straight. The vertical component is upthrown on the west side: about 3m on the L surface younger than 25ka, and 5-10m on the H surface older than 140-150ka. These fluvial terrace surfaces show cumulative vertical offsets. The fault plane dips steeply westward and the striations are plunging to the south at an angle of 10-30°. The Kokura-higashi Fault has repeatedly moved in the late Quaternary with predominantly right-lateral movement. The average slip rate of the fault is over 0.1 m/ky (class B). The most recent faulting occurred around the early BC 2ndcentury. The timing of the penultimate faulting event is ca.10,000yBP. If the average recurrence interval is over 9,000yrs, it is not likely that the fault will generate a large earthquake in the near future. The ground surface might be vertically dislocated about 1.7m by reverse slip associated with the past faulting. The net-slip should be several times as large as the reverse slip, taking the right-lateral movement into account.
The Ms 8.1 Central Kunlun earthquake of 14 November 2001was triggered by the active Kunlun fault in the Central Kunlun mountain area, northern Tibet. A nearly 400-km-long co-seismic surface rupture zone occurred along the western segment of the Kunlun fault. Field investigations show that the surface ruptures are distributed in a zone with width ranging from few to several hundreds of meters, generally from 5m to 50m. The surface rupture zone is composed of distinct shear faults, extensional cracks, and mole tracks. The leftlateral offsets are measured by using the surface deformation markers such as present-day glacier, moraine, stream channel, gully, and road, which vary from few tens of cm to 16.3m, but generally from 4m to 8m. The maximum displacement up to 16.3m was observed across a rupture zone of 550m wide. Both the rupture length and maximum displacement are the largest among the co-seismic surface rupture zones ever reported worldwide. The co-seismic deformation characteristics of surface markers reveal that the earthquake had a purely strike-slip focal mechanism. Both the geological and geomorphological evidence indicates that the geometry of ruptures is controlled by the pre-existing Kunlun fault. The large amounts of strike-slip and rupture length produced by the earthquake support the hypothesis that the Kunlun fault plays an important role of strike-slip partitioning in the rapid eastward extrusion of Tibet to accommodate the continuing penetration of Indian plate into Euro-Asia plate. The Central Kunlun earthquake provides an exceptional opportunity to study the geometry of co-seismic rupture structures along a large strike-slip fault for further understanding the relationship between the fault geometry and rupturing process.
This paper shows the distribution of the Osaka-wan fault based on the existing multi-channel seismic profiles, and examines the validity of the estimation. According to the estimation, the length of the fault reaches about 35km, and the width of the flexure extends to about 1,000m. In addition, this result is in agreement with the distribution of the Osaka-wan fault by the single-channel seismic profiles.
The Yoro fault is an east vergent thrust that extends north-south for 30 km along the western margin of the Nobi Plain, central Japan. We conducted drilling investigation across the late Holocene fault-related landform at Shizu-Shobuhara site (Togo,2000) to define how this fault-related landform has developed during late Holocene time. Near-surface stratigraphic relations based on drillings of 21 cores up to 5 m long by Geoslicer and 15 cores up to 7 m long by percussion core sampler, age controls of strata based on radiocarbon dating, and structural analyses on stratal geometry indicate at least three episodes of active folding along the Yoro fault. Among them periods of two recent earthquakes postdate the deposition of A2-c (peat) layer (ca.2400-1200 yBP), and the most recent event occurred in the historical time. Reliable restoration of deformed strata was constructed by using readily identifiable beds with highly constraints on those ages. Syntectonic strata, uninterruptedly deposited to bury the preexisting topography associated with the former deformational event(s), were deformed during the penultimate earthquake. Unconformity above the peat suggests truncation of the upper half of older folded deposits and breaching of the fold due to the following erosional event. Most recent earthquake produced topographic relief on the top of the subsequent infilling above unconformity and completed the approximate shape of the present-day fault-related landform at this site. These descriptions demonstrate that evolution of the late Holocene fault-related landform at this site can be explained as convolution of discrete events of deformation associated with active faulting and of erosion and sedimentation during interseismic periods.