We have studied micro-seismicity in the northern part of Hokkaido (north above 44°N) from June to November in 1998 combining eleven temporal seismic stations with seven and five routine ones operated by Institute of Seismology and Volcanology (ISV) in Hokkaido University and Sapporo Meteorological Observatory, Japan Meteorological Agency (SMO), respectively.
Firstly we determined 91 hypocenters of local earthquakes in this period using the dense network. These hypocenters are about three times of the number of those from the routine network alone. One-dimensional P-wave velocity structure assuming four-layer model (assumed thickness of 2, 8, 10km and infinite) and station corrections were estimated using a P-wave travel time inversion method with 735 P-wave arrival time records of 81 events. The velocity of each layer was determined to be 2.83km/sec for the first layer, 5.32km/sec for the second one, 6.32km/sec for the third one and 6.69km/sec for the bottom half space. From the P-wave station corrections we obtained, this region can be classified into three zones parallel in the north-south direction; the western islands region in the Sea of Japan, the western part of mainland, and the eastern part of mainland. Each zone shows the value of less than-0.5sec, +0.1-+0.4sec, and-0.1--0.5sec, respectively.
Next, the hypocenters with the inverted velocity structure and the station corrections are relocated. As the results show, some hypocenters in the anomalous delayed station correction zone, i. e. the western part of mainland, are clearly located at the depth range from 20 to 25 km. Focal mechanism solutions of these deep events show normal fault type, while shallower events less than 20 km depth show strike-slip and reverse fault types. We also relocated 381 earthquake hypocenters which were routinely determined by ISV from October 1996 to December 2000. According to the relocated hypocenter distribution, a high seismic zone is shown in the western part of mainland with about 50km wide along a north-south direction. On the other hand, the eastern part of mainland is strongly characterized as aseismic zone. The boundary between the seismic and aseismic zones corresponds to the geological boundary between Kamuikotan metamorphic belt and Hidaka belt.

View full abstract

The Onikobe area is an active geothermal area situated in the Ou backbone range near the border between Akita and Miyagi Prefectures, northeastern Japan. There are Tertiary to Quaternary calderas and active volcanoes in this area. We calculated Q of coda waves (Qc) in this area by applying a single scattering model. Seismograms recorded at stations located within 10km of the epicentral distance for earthquakes which occurred at depths shallower than 10km were used in this analysis. The time window for the calculation of Qc ranges from twice the S-wave travel time to the arrival time of the wave reflected at the Moho discontinuity. We found a systematic variation of Qc with station location: Qc values were lower at stations in and around calderas than at other stations. The amplitude of the coda wave characterized by low Qc value decreased smoothly with the lapse time, while the decay rate of the coda wave with the high Qc value changed at around 6 s after the origin time. By taking these observational facts into account, we proposed a model in which the scattering coefficient was very large in the lower crust and a small and strong attenuation zone was embedded in the upper crust beneath a caldera. The synthetic coda envelopes explained the difference in the gross feature of observed Qc and decay curves.

View full abstract

When the empirical Green's function method is applied to seismograms recorded at a site which is located on soft soil layers, it is necessary to take into account the nonlinear behavior of the soft soil layers. In practice, it has often been assumed that the seismic wave is affected by soil nonlinearity only after it's incidence to the local soft soil layers. If we consider a seismic ray connecting the source and the site, however, it might be true that the ray crosses the soft soil layers several times. Therefore, it is reasonable to assume that the seismic wave is affected by soil nonlinearity several times during the transmission from the source to the receiver. This potential phenomena is referred to as “multiple nonlinear effects” in this article. Taking into account these effects in the empirical Green's function method, the authors propose “nonlinear parameters” which represent the deviation of material properties of the sediments due to soil nonlinearity from linear status and which are used to modify Green's functions. The parameters are applied to the simulation of the 2000 Tottori-ken Seibu and the 1995 Hyogo-ken Nanbu earthquake ground motions. It is found that, although the new parameters are quite simple, they greatly contribute to the improvement of the simulation results.

View full abstract

We determine the prior distributions of b-value in Gutenberg-Richter law from 81 aftershock sequences accompanying with shallow earthquakes of M7.5 or larger in the world and 67 sequences related to large events in and around Japan. We suppose that b-value in the prior population follows gamma distribution, Γ (φ, ζ), where φ and ζ are shape and scale parameters, respectively. The observed b-value for each sequence corresponds to a random sample from the population but has statistical estimation error. In this paper, we estimate φ and the mean of population distribution, b_avg=φζ, using the maximum likelihood method. Estimated φ and b_avg are 55 and 1.13 for sequences of aftershocks of M5.0 or larger in the world, and 28 and 1.22 for ones of aftershocks of D<2.55, respectively, where D is the magnitude difference between the main shock and aftershock. For sequences in Japan, φ=28 and b_avg=0.97 are obtained. Numerical tests of Monte Carlo method show that estimated b_avg is fairly close to the input value. However φ is apt to become larger than the input. Then smaller values, such as φ=26 to 31 for sequences in the world and φ=20 to 25 in Japan, might be better than the estimates by the maximum likelihood method. On the assumption that the prior distribution of b-value is Γ (φ, bavg/φ), the maximum likelihood estimate, b, for a set of N earthquakes is giveR with b= (N+φ-1) / (N/b_U+φ/b_avg), where b_U is a estimate of the maximum likelihood method without prior distribution proposed by Utsu. The estimate b is more stable and reliable than b_U. In particular, the method proposed here is quite valuable when few event data are available.

View full abstract

A new seismotectonic province map of the Japanese Islands and the adjacent areas, which carries maximum magnitudes of earthquake (M_max) expected for the individual provinces, has been prepared as a revised edition of Kakimi et al. (1994). The major part of the mapped region constitutes an island arc-trench system, which is surrounded by Northwest Pacific Basin (1), Shikoku Basin (2), Philippine Basin (3), Kurile Basin (4), Japan Sea Basins (5), and Korean Peninsula and Tonhai Continental Shelf (6). All of the peripheral provinces have too low seismicity to be given M_max. The island arc-trench system is subdivided into the following constituent arcs: Kurile Arc (7), Northeast Honshu Arc (8), Izu-Bonin Arc (9), Southwest Honshu Arc (10), Ryukyu Arc (11), Sakhalin Arc (12), and the Tectonic Belt along the Eastern Margin of Japan Sea (13). While the constituent arcs 7 to 11 are divided into three tectonic belts, which remarkably differ from each other in tectonic, seismic, and volcanic activities, from the trench to the inland: Continental Slope on the Trench Side (A), Non-volcanic Outer Belt (B), and Volcanic Inner Belt (C), the constituent arc 10 alone has additionally the Continental Slope on the Marginal Sea Side (D). Province 12 started developing in Late Mesozoic and functioned as a collision belt between the North American Plate (NA) and the Eurasian Plate (EUR) in Late Cenozoic, whereas province 13 is considered to form a current collision belt between NA and EUR plates. Province 11X, Okinawa Trough, is defined as a current rift zone developing between the Tonhai Continental Shelf and the Ryukyu Arc. Some of the provinces are further divided into subprovinces in response to local differences in active faults, seismicity, M_max etc.
All the active faults on land are grouped into seismogenic faults (Matsuda, 1990), which are considered to generate characteristic earthquakes. The magnitudes of earthquake expected for the seismogenic faults (ML_max) are estimated by the equation: log L=0.6ML-2.9 (Matsuda, 1975), where L is the length of the faults in kilometers. The maximum magnitude of earthquake expected for seismogenic faults (ML_max) and the maximum one for historical shallow earthquakes (Mh_max) are compared in each province to choose the larger one as the expected maximum earthquake magnitude (M_max) for the province. Since no method to decide a seismogenic unit from offshore active faults has been established, Mh_max is tentatively adopted as the M_max representing the province. Extraordinarily long faults found in inland provinces, which are called the designated faults (Matsuda, 1990), are excluded from estimation of the M_max. None of the magnitudes of earthquake expected for the designated faults is shown here, because they should be individually estimated. All of the information, such as tectonic geomorphology and geology, characteristics of active (seismogenic) faults, historical earthquakes, modern seismicity, and other, is put into a table to facilitate the identification of a seismotectonic province and the determination of the M_max and the designated faults. The details of the boundaries between seismotectonic provinces are shown in another table.

View full abstract

The amount of slip per event is one of the most important parameters establishing a recurrence model of large earthquakes. In Japan, previous works have usually exposed only two trench walls, and recognition of faulting events and estimation of holizontal slip have some uncertainty. This study tried to estimate accurate slip per event and identify faulting events precisely using threedimensional trenching and Geoslicer techniques. The Tashiro basin was selected as the study site along the Tanna fault which ruptured during the 1930 Northern Izu earthquake (M_s 7.3); the trench site is covered with continuous alluvial deposits and is expected to contain numerous piercing-points of offsets across the strike-slip fault.
At the Tashiro site, the four earthquake events are recognized in the last 3, 000 years. The ages of the four events are: event 1: after AD 1442, event 2: AD 1610-1296, event 3: AD 1398-685, event 4: 1671-ca. 2800 y. B. P. This result is consistent to the one obtained at the 1982 Tanna-Myoga site. Event 1 can be correlated to the 1930 Northern Izu earthquake, and a lot of geological sections revealed right-stepping en echelon cracks associated with the 1930 earthquake at Tashiro site. Event 2 is correlated to a possible event estimated by previous works and event 2 may have ruptured the surface from Tashiro site to Osawa-ike site to the south. Event 3 is probably correlated with the 841 Izu-koku large earthquake according to historic records and previous paleoseismological works. The recurrence intervals of the last four events are estimated as follows: 500-630 years between events 1 and 2, 460-590 years between events 2 and 3, assuming the correlation of event 3 with the AD841 event, c. a. 500-1600 years between event 3 and 4. Average recurrence interval during the past 3000 years ranges from 650 to 750 years. This average recurrence interval is significantly shorter than that of previous work at Myoga site, 700 to 1000 years in the last 7, 000 years.
Reconstruction of offset channel-fill gravel led us to idendify accurate 50cm left-lateral and 15 to 20cm vertical slip along the 1930 rupture. Another offset channel allowed us to estimate cumulative 60-80cm slip by the 1930 event and the penultimate event. Slips of the last two faulting events are estimated at 50cm and 10-30cm, respectively. Cumulative vertical slip associated with 4 faulting events is estimated at 1.7 to 1.8m. These data suggest that in the last 3, 000 years the Tanna fault has possibly not moved with characteristic slip during individual seismic events. Taking short reccurence intervals of the Tanna fault into consideration, recent activity of the Tanna fault may not support slip-predictable reccurence model, because intervals between events 1-2 and 2-3 are not proportional to the amount of slip of the following faulting events, events 1 and 2.

View full abstract