We present co-seismic displacements of the Hyogo-ken Nanbu earthquake of January 17, 1995, detected by continuous GPS (Global Positioning System) observation, campaign type GPS survey and leveling. Continuous GPS observation gives a consistent pattern of displacements with those expected from a right lateral slip on a NE-SW trending vertical fault in far field: stations about 50 km east and west of the epicenter moved toward the epicenter by about 4 cm, while stations north and south moved away from the epicenter. By comparing with line lengths obtained by geodolite about 10 years ago, the campaign type GPS revealed most control points on Awaji Island moved to the southwest or south, which may be attributed to the movement of the Nojima Fault which cut the surface. On the other hand, control points northwest of the Rokko fault system moved toward the northeast and those on the other side moved slightly to the west, in and around Kobe. Leveling data revealed uplift of 19 cm on the northwestern side of the Suma Fault, a member of the Rokko fault system, and subsidence of 7 cm just east of this fault. Furthermore, uplift of about 5 cm was observed in the central part of Kobe City, and subsidence of 5 cm was detected east of Kobe. There is no significant gap in horizontal and vertical displacements around the northern extension of the Nojima Fault, which implies a complicated rupture process of this event. Leveling on the east coast of Awaji Island revealed a significant uplift of about 20 cm with slight subsidence at both edges of this uplift region during the past 20 years. By fitting the above geodetic data, we searched for an optimal set of parameters of a dislocation model. We assumed six nearly vertical faults trending NE-SW from Kobe to Awaji Island on the basis of aftershock distribution and focal mechanism. About 250 cm of the right lateral slip for the fault on Awaji Island is derived from large horizontal displacements near the Nojima Fault. The fault in Kobe may be divided into two segments with 100-200 cm slip by a slip-free zone which roughly corresponds to the cluster of aftershocks. The southern part of the Nojima Fault, segments near the Akashi Strait, and north of central Kobe may have significantly large thrust components of 100 cm. We also examined the possibility of buried faults beneath the zones of severe damage. Since this model cannot explain the observed geodetic data and the estimated slips are inconsistent with focal mechanism, these possible buried faults may not play a significant role, if any.

View full abstract

A large M=7.2 earthquake occurred just below the western part of the Osaka-Kobe megalopolis, southwest Japan on January 17, 1995. The Japanese University Consortium for GPS Research conducted extensive local GPS measurements in and around the hypocentral region to collect near-field data on co-seismic and post-seismic crustal deformation. Adopting old GPS data collected before the earthquake, co-seismic displacement vectors have been obtained at five sites. Horizontal displacements are larger than 0.4 m in the vicinity of the earthquake fault and decay steeply with distance from the fault. The horizontal deformation pattern represents a typical right-lateral slip motion along the fault. In contrast, vertical displacements are rather difficult to interpret probably because of the low precision of old GPS measurements and the complex local site condition.

View full abstract

Within 1 to 4 days following the 1995 Hyogo-ken Nanbu earthquake (M 7.2) that occurred on January 17, 1995, the Japanese University Consortium for GPS Research deployed more than 30 GPS receivers in the hypocentral area to investigate post-seismic activity. Of these stations, we analyzed data from 8 sites that were placed near the hypocentral area; six in Kobe and two in Awaji. Two methods of analysis were employed: 1) change of absolute geocentric coordinates were estimated by assuming Usuda global site, which is about 300 km northeast of the hypocentral area, as fixed, 2) analyses for short baselines were conducted within the hypocentral area using the automatic processing technique. Results suggest that 1) rapid displacements toward the southeast of more than 4cm/year were found at all of the analyzed sites, which might be part of a post-seismic deformation process, 2) largest of which was found at Rokko Farm site, at which displacement amounted to about 3 cm until the end of February, though its cause were not specified, and 3) local straining seems to be taking place at both Kobe and Awaji areas.

View full abstract

Post-seismic deformations in the focal region of the 1995 Hyogo-ken Nanbu earthquake (M=7.2), in southwest Japan, which occurred on January 17 (JST), were observed with GPS observations. We deployed more than 30 GPS stations in and around the area of aftershock activity within several days after the occurrence of the earthquake. First the data were stored and manually collected at the observation sites, but by the middle of February the data at most of sites were started to be telemetered through public phone lines to data centers. GPS observations continued until the end of March, except for 4 sites left active for monitoring long-term deformations. We used two types of GPS observations, usual static and real-time kinematic GPS observations. Static day-by-day analyses, which are based on 24-h data sampled every 30 s, show the deformation associated with post-seismic readjustment with a relaxation time of about 70 days and with an amplitude of 1-2 cm in the area close to the epicenter. Also, rapid movements were observed just after the occurrence of the earthquake. However, since our analyses are only preliminary, further analyses are needed to clarify these observations and to investigate models of post-seismic deformations. Real-time kinematic GPS observations, which is a new type of GPS observations, were made every 1 s from February 9 to March 22 for detecting possible rapid relative movements between two sites across the surface fault trace on Awaji Island and across the other fault, the Arima-Takatsuki Tectonic Fault on the Kobe side. Present real-time kinematic GPS observations show scatters with the amplitude of 0.5-3.0 cm depending on components and baseline lengths. Any motions with amplitudes larger than these noise levels could not be detected during the observation period.

View full abstract

A disastrous earthquake with a magnitude of 7.2 hit the southern part of Hyogo Prefecture on January 17, 1995. The mainshock was located on an active fault of the Arima-Takatsuki-Rokko fault system. Its focal mechanism was consistent with a right-lateral strike-slip fault trending N40°E. Three days after the occurrence of the mainshock, we started to install a highly dense seismic array in and around the fault area of the quake. Two permanent regional seismic networks of more than 30 stations covered the entire area of 200 km × 200 km. The temporarily installed array of 27 stations spanned the fault area of 15 km × 50 km. All data were telemetered to a temporary observation center at Uji. We located about 3, 100 aftershocks in real time for 1 month using an Internet connection between seismic networks. The aftershock area extended 70 km trending northeast to southwest. Hypocenters determined by the network were delivered automatically through the Internet. We found that the aftershock distribution was heterogeneous in space and time: seven clusters of hypocenters were identified and temporary variation in the rate of occurrence of aftershocks had a periodic component with periods of a half of day, 1 day, and 3 days, that are superimposed on decaying of the rate following Omori's law.

View full abstract

On January 17, 1995, the Hyogo-ken Nanbu (Kobe) earthquake of magnitude 7.2 occurred in the Kansai district near Kobe, Japan. After the quake, the Net-Hyogo group set temporal seismic stations in and around Kobe and Awaji Island. Using P and S arrival data from temporal observations, we determined the aftershock distribution of the Kobe earthquake. We used the JHD (joint hypocenter determination) method which eliminates mislocation of hypocenters caused by errors of the simplified model of 1-dimension velocity structure and heterogeneity near stations. From the analysis, we obtained final results of the aftershock distribution. First, the hypocenters thus obtained form a thin flat zone in the Kobe area but form a complex feature in the Awaji area. The aftershock zone strikes N51°E and the total area of the aftershock zone is 60 km wide and 15 km deep. Second, in its northeastern half (Kobe and central area), the aftershock zone forms nearly a simple plane. Rupture did not appear at the surface in the Kobe area while a number of aftershocks were located underneath the surface trace of pre-existing active faults. Third, in the southwestern half (Awaji area), it shows complicated features. The aftershock zone dips northwestward near the northern tip of Awaji Island but southeastward in the central part of the Awaji area. Although the rupture trace appeared on the surface along the Nojima Fault which runs on the western coast of the island, no aftershocks have yet been found to occur immediately beneath the surface rupture. The aftershock zone is located 3 km to the east of the Nojima Fault. It indicates that the fault geometry at the surface does not correspond to the distribution of aftershocks in the Awaji area. These results show that surface fault traces do not necessarily connect directly with the underground fault geometry estimated by aftershock distribution if the fault dips vertical. Pre-shocks occurred several hours before the main shock in the central part of the fault region where the initial rupture of the main shock started at the a depth of 17 km. Hypocenters of the pre-shocks and the initial rupture are located at the lower edge of the aftershock zone. This implicates that the precursory rupture process of the Hyogo-ken Nanbu earthquake began at the base of the seismogenic zone.

View full abstract

We carried out a seismic marine survey around Akashi Strait employing reflection and side-scan sonar imaging methods and an ocean bottom seismograph (OBS) array in order to closely examine the aftershocks in the immediate vicinity of the rupture initiation area of the 1995 Hyogo-ken Nanbu earthquake. A total of 71 aftershocks and some mechanisms were determined using OBS data for a period of 1 month with land temporary station data starting 12 days after the main shock. We identified a number of seismologically characteristic clusters to understand the heterogeneity that seem to exist around Akashi Strait. We determined that two major faults, the Nojima Fault and the Rokko fault system, are right-stepping within the strait cause en-echelon faultings in between. The Nojima Fault seems to terminate by deflecting westward in the strait. Focal mechanisms of these events are generally concordant with an E-W compression stress field, which is the tectonic trend. Events beneath the northern part of Awaji Island occur as a cluster dipping eastward, but mechanisms vary.

View full abstract

To evaluate the temporal variation of seismicity in the Inner Zone of Southwest Japan before and after the Nankai trough events, we introduced a statistical model and estimated the value of the model parameters. We used the data of disastrous earthquakes to estimate them. Because of the lack of spatially sufficient data, we used data from the 9th century in the case of the northern Kinki region, and in the case of the whole Inner Zone of Southwest Japan only data after the 17th century. The results show that for the northern Kinki region the seismicity has a peak before the Nankai trough events, although there is no significant change before them in the whole Inner Zone of Southwest Japan. The seismicity in the Inner Zone of Southwest Japan increases just after the Nankai trough events. We compared the obtained intensity functions with the recent JMA data from 1885 to 1995. The seismicity seems to have increased in the last 30 years. Using data from this period, we estimated the occurrence time of the next Nankai trough event. Our results show that it will occur in the 2030'S.

View full abstract

We carried out dense gravity measurements with the Fast Static GPS positioning along five survey lines across the Rokko fault system. We calculated the Bouguer anomaly to estimate the underground structure relevant to the 1995 Hyogo-ken Nanbu earthquake. We applied the two-dimensional Talwani's method to the modeling, assuming that the density structure is similar for all the five survey lines. All the Bouguer anomaly profiles are characterized by flat areas on the Rokko Mountains and Kobe Plain, with steep gradient between them. The gradient is most likely due to the fault scarps between the Rokko granite and the sedimentary layer of the Osaka and Kobe groups. It also indicates that the faults are vertical or reverse ones, at least just below their surface traces. This finding is consistent with the tectonic background of the Rokko Mountains that have upheaved under the east-west compression in the Quaternary. Moreover, we found the southwestward extension of the Koyo Fault underneath the sedimentary layer. The extension runs on the edge of the earthquake disaster zone. We inferred that the thickness of the sedimentary layer is 1-2 km near the sea, decreasing gradually toward the mountain side. The wedge-like structure and the hidden fault under the Kobe Plain may have served as a focusing lens of seismic rays during the earthquake.

View full abstract

There have been many reports on earthquake-induced groundwater anomalies, but the mechanism of these anomalies is not very clear. The 1995 Hyogo-ken Nanbu earthquake, which occurred on January 17, 1995, was one of the greatest shallow inland earthquakes in Japan since the modern Japanese observation system for monitoring seismicity was established. Many groundwater anomalies have been observed following this earthquake. Our purpose was to study these anomalies systematically and to gain some knowledge of their mechanisms. We initiated a questionnaire survey in and around the source region on March 8, 1995. We also investigated the discharging waters which appeared on Awaji Island after the Hyogo-ken Nanbu earthquake. These results were compared with the volumetric strain changes estimated from fault models for the earthquake which stretch in a SW-NE direction. Many post-seismic rises in the level or discharge of groundwater as well as a few post-seismic drops were detected by the questionnaire and the investigation. However, their distribution did not coincide with the distribution of the volumetric strain changes, which are of a quadrantial type because the fault models are of a right-lateral strike-slip type. Several precursory groundwater anomalies were also reported, although few pre-seismic crustal strain changes were observed. Therefore, the volumetric strain changes do not always explain the groundwater anomalies that accompanied the 1955 Hyogo-ken Nanbu earthquake.

View full abstract

A post-earthquake investigation revealed statistically significant changes in the chemical composition of groundwater pumped from wells located about 20 km east of the epicenter within the aftershock region, preceding the disastrous M7.2 Hyogo-ken Nanbu earthquake in southwestern Japan on January 17, 1995. The chloride ion concentration had increased steadily after August 1994, compared to the constant level since July 1990, and reached an excess of about 10% just before the earthquake. Groundwater sampled after the earthquake showed much higher Cl^- concentrations. Together with the increase in Cl^-, sulfate and alkaline earth metal ions (Mg²⁺, Ca²⁺, Sr²⁺, Ba²⁺) showed similar increases in their concentrations, while sodium ion and silica contents showed decreases. These changes may be the result of shallow groundwater inflow, enriched in chloride, sulfate and alkaline earth metal ions, to the artesian layer of the wells, due to permeability enhancement in the shallow stratum. The observed changes in groundwater chemistry before and after the earthquake are useful clues to understanding the mechanism of the present earthquake.

View full abstract

This article reports a preliminary result of seismic-related radon changes in groundwater observed at a continuous monitoring station in the Tokai district. Immediately after the M7.2 Hyogo-ken Nanbu earthquake which occurred on January 17, 1995, the radon concentration of a water well at a hypocentral distance of 260 km decreased by about 5%. It is meaningful that a drop in the radon concentration was observed, while the water level showed no significant change.

View full abstract

e made observations of the electric potential at three sites located near the Nojima Fault on Awaji Island, aiming at examining whether any anomalous changes appear in association with aftershocks of the 1995 Hyogo-ken Nanbu earthquake of M7.2. Unfortunately however, the observation area turned out to be highly contaminated by noise currents arising from the remote DC-operated railway systems. In order to avoid noise contamination, we attempted to apply the BAYTAP-G analysis method to the data obtained during the aftershock activity and found a gradual change which appeared several days before an aftershock of M4.8. We also show time changes in the preferential direction of noise currents, due to the inhomogeneous resistivity structure in the vicinity of the fault, and suggest that they are likely to reflect changes in the electrical property of the fault.

View full abstract

Before and after the 1995 Hyogo-ken Nanbu earthquake on January 17, three kinds of phenomena of electromagnetic radiations were recorded at Uji Station, Kyoto Prefecture, Japan. They were: 1) A series of increases in number of electromagnetic noises starting from 12 days before the main shock to 2 days after the main shock, 2) A clear increase of electromagnetic noises from 5 o'clock on January 17, and 3) A co-seismic change of electric field intensity in the VLF range observed at the time of arrival of seismic waves with large amplitude, then the area around Uji Station was strongly shaken. An investigation by questionnaire about the anomalous electromagnetic phenomena before and after the earthquake was also executed after the earthquake by mailing to amateur radio operators around the focal region, i.e. those in Takarazuka, Itami and Sumoto Cities. The daily variations of the numbers of anomalous electromagnetic phenomena reached a peak just before and just after the earthquake.

View full abstract

The sub-ionospheric VLF Omega signal transmitted from Tsushima, Japan (geographic coordinates: 34°37'N, 129°27'E) is continuously received at Inubo (35°42'N, 140°52'E). The signal's propagation characteristics (especially phase) exhibited abnormal behavior (especially around the sunrise and sunset local times) a few days before the main shock of the 1995 Hyogo-ken Nanbu earthquake. We found from the computer simulation of VLF signal propagation that the observed effect can be explained by a decrease of about 0.7 km of VLF reflection height at the atmosphere-ionosphere boundary.

View full abstract

Surface waves excited by the January 17, 1995 Hyogo-ken Nanbu earthquake recorded by the newly upgraded Taiwan short-period seismic network were used to perform a dispersion study. It is shown that for large and shallow earthquakes, recordings from short-period seismic networks can be effectively used for an intermediate period band (10 to 30 s) surface-wave analysis. The small station spacing of the Taiwan Seismic Network allows the application of array processing to extract high resolution phase velocities for both Love and Rayleigh waves, and to better handle the problem of surface-wave velocity determination for wave approach deviating from the great circle path. From the dispersion data so derived, a better definition of the Taiwan crustal structure is obtained. Since the general direction of the surface-wave arrival is nearly parallel to the major structural grain of Taiwan, this provides an opportunity to delineate the crustal lateral heterogeneities for different geological provinces in that region. The S-wave crustal structures from this surface-wave analysis are consistent with P-wave crustal structures from recent tomographic studies. Our results further indicate the absence of a root for the Central Range of Taiwan which rises to a height of 4, 000 m. This lack of a mountain root is also required by the recently completed Taiwan Bouguer anomaly map.

View full abstract