Preliminary Observations on the Niigata Ken Chuetsu, Japan, Earthquake of October 23, 2004
Total Page：16
File Type：pdf, Size：1020Kb
EERI Special Earthquake Report — January 2005 Learning from Earthquakes Preliminary Observations on the Niigata Ken Chuetsu, Japan, Earthquake of October 23, 2004 EERI organized a field investigation significant earthquake to affect Japan more than 100,000 people into tem- team led by Charles Scawthorn of since the 1995 Kobe earthquake. porary shelters, and as many as Kyoto University, which included Forty people were killed, almost 10,000 will be displaced from their Scott Ashford, University of Califor- 3,000 were injured, and numerous upland homes for several years, if nia San Diego; Jean-Pierre Bardet, landslides destroyed entire upland not permanently. Total damages are University of Southern California; villages. Landslides were of all types; estimated by Japanese authorities Charles Huyck, ImageCat Inc.; Rob- some dammed streams, creating at US$40 billion, making this the ert Kayen, U.S. Geological Survey; new lakes likely to overtop their new second most costly disaster in his- Scott Kieffer, Colorado School of embankments at any moment and tory, after the 1995 Kobe earth- Mines; Yohsuke Kawamata, Univer- cause flash floods and mudslides. quake. sity of California San Diego; and Landslides and permanent ground Rob Olshansky, University of Illinois, deformations damaged roads, rail The epicenter was in northwestern Urbana/Champaign and Visiting lines and other lifelines, resulting in Honshu, about 80 km south of Professor, Kyoto University. Paul major economic disruption. The nu- Niigata City (population 500,000), Somerville, URS Corporation, and merous landslides resulted, in part, well-known as the place liquefac- Jim Mori, Kyoto University, covered from heavy rain associated with tion was first systematically stud- the seismological aspects, but were Typhoon Tokage. At Nagaoka City, ied following the 1964 M7.5 earth- not part of the field investigation. there had been 100 mm (4 inches) on quake. The epicenter was beneath Bardet and Kayen are members October 20 and 13 mm (.5 inch) on the Uonuma Hills along the Shin- of the Geotechnical Engineering October 21. The earthquake forced anogawa Lowland (Figure 1), an Earthquake Reconnaissance (GEER) Working Group, which organized a second team to focus on landslides, supported by an EERI Beyond Reconnaissance Grant. It was led by Ellen Rathje, University of Texas, and included James Bay, Utah State University; Randy Jibson, U.S. Geological Survey; Keith Kelson, William Lettis and Associates; and Robert Pack, Utah State University. Dr. N. Kishi of AIR participated with the EERI team and contributed to this report. The EERI team arrived on October 30, and was in the field through No- vember 4. The GEER team was in the field November 15-19. The re- search of both teams and the publi- cation of this report are funded by EERI’s Learning from Earthquakes Program, under National Science Foundation grant #CMS-0347055. Introduction Figure 1. Aftershock pattern and mechanism. The Unonoma Hills are approxi- The Mw 6.6 earthquake that struck mately outlined by dashed ellipse, and the Shinanogawa Lowland is the white Niigata Prefecture on the evening area to the west (adapted from original figure by Group for Joint Observation, of October 23, 2004, was the most Kyoto and Kyushu Universities). 1 EERI Special Earthquake Report — January 2005 EERI Special Earthquake Report — January 2005 alluvial plain that is bisected by the shows almost pure reverse faulting pectively, located 16 km north of the Shinano River as it flows toward on a fault striking 30 degrees east of epicenter. The peak acceleration re- the Sea of Japan. While the 2004 north, dipping down to the northwest corded at Niigata City, at an epicen- earthquake was felt in Niigata, at 54 degrees. The source area is 20- tral distance of 70 km, was 0.11 g. neither significant effects nor lique- 30 km long, with down-dip widths of For comparison, the Mw 7.5 1964 faction was observed there. South about 14 to 25 km. Niigata earthquake, which caused (or upriver) of Niigata, toward the the extensive liquefaction, was epicentral area, there are a number The earthquake occurred near the located offshore at an epicentral of smaller cities and towns on the Sea of Japan in an area of relatively distance of about 55 km north of plain and some industrial sites; the high seismicity. Within 100 km, there the city, and had a peak value of remainder of the plain is covered by have been five other earthquakes 0.16 g, with a longer duration in wet rice farmland (i.e. rice paddies). greater than Mjma6 since 1941, includ- Niigata City. Nagaoka (population 194,000) was ing the 1964 Mjma7.5 Niigata earth- the largest city significantly affected quake. All the large earthquakes Aftershock hypocenters are distrib- by the earthquake. It suffered mini- of the present sequence occurred uted over a region that is about 30 mal damage, but had highway and beneath the mountainous region east x 20 km at depths ranging from 15 train service disruptions, school of Ojiya, where there are mapped km to the surface, which roughly closures, and an influx of refugees Neocene-Quaternary folds. Across coincides with the area of the main- from surrounding areas. More the region, there are many shock and larger aftershock rupture seriously damaged areas were the mapped short lineaments city of Ojiya (population 41,000) (1-10 km) trending north- and surrounding rural areas, most east-southwest, along with notably the village of Yamakoshi the recognized Muika-Machi- (population 2,222) and the town Bonchi-Seien fault (Geologi- of Kawaguchi (2000 population cal Survey of Japan). 5,748), which sustained JMA intensity 7, the maximum on that The hypocenter was in the scale (approximately equivalent to central or lower region of the MMI XI-XII). fault, so the rupture propa- gated updip toward the east The Joetsu Shinkansen crosses and along strike in the north- this region in a north/south direc- east and southwest direc- tion and terminates in Niigata. One tions, with a duration of of its trains was derailed by the about 10-15 sec. The lar- earthquake. gest amount of slip was in an area about 10 km x 10 km in Seismology and Strong the region of the hypocenter. Ground Motion Strong hanging wall effects were manifested in the large The mainshock at 17:56:00 Japan peak accelerations recorded time (08:56:00 UT) on October 23 above and to the west of the had a moment magnitude Mw 6.6. fault at locations such as The epicenter (lat. 37.30, long. Ojiya, where 1.33 g PGA 138.84) was located 4 km east of and 128.7 cm/sec PGV the city of Ojiya (Figures 1 and 3). were recorded 4 km west of The mainshock was followed by the hypocenter (Figure 2). an aftershock sequence with four The largest recorded peak events of JMA magnitude 6 or lar- acceleration, 1.75 g (and ger, and 12 events of magnitude 53.1 cm/sec PGV), was re- 5 to 5.9 (as of November 28), far corded at Tohkamachi more than usual for an event of (NIG021), located about this magnitude. There was no sur- 20 km south-southwest of face rupture observed with the the epicenter. Peak acceler- earthquake, and the mainshock ations of 0.48 g and 0.89 g Figure 2. Ojiya K-NET Seismograph NIG019. is currently not associated with were recorded at Nagaoka (top) PGA 1.33 g EW, 1.17 g NS, and 0.83 g any mapped fault. The focal and Nagaoka-Shisho, res- UD, and (bottom) 5% damped response spec- mechanism of the main shock tra for two horizontal directions (source: NIED). 2 3 EERI Special Earthquake Report — January 2005 EERI Special Earthquake Report — January 2005 areas. The locations show a complex pattern of westward and eastward dipping structures that are associated with the mainshock and larger after- shocks. Geotechnical Aspects Landslides, liquefaction, and perman- ent ground displacements accounted for most of the physical and economic effects of the earthquake. Figure 3 shows the distribution of landslides caused by the earthquake, as well as areas where liquefaction was gen- erally observed. The map also shows other effects discussed below. Landslides: Niigata Prefecture docu- ments 442 individual landslides in this earthquake, although on the ground the number appeared even larger. Heavy rainfall in the days prior to the earthquake likely had a fundamental impact on the distribution and scale of landsliding. Bedrock throughout the area of major landsliding consists of a folded sequence of very weak and friable claystone and siltstone that is fairly nondurable, with interbedded sandstone and minor conglomerate. For the argillaceous bedrock materi- als, high antecedent moisture would tend to cause softening and strength reduction, as well as elevated pore or joint water pressures and seepage forces. Figure 3. Map of affected area showing main effects of mainshock Observed modes of slope failure in- (adapted from KKC aerial photos). cluded translational soil slips, deep- seated rotational slumps, debris flows, slump-flow complexes, and large block The most widespread type of ob- Another commonly observed slides. In some cases, the failures served failure consisted of trans- failure mode was slump-flow were clearly controlled by bedrock lational slip of the regolith and complexes within colluvium-filled structure, while other deep-seated highly weathered bedrock mate- swales and in areas of deep bedrock failures did not have an rials on very steep slopes flanking regolith. If antecedent moisture apparent relation to the bedrock floodplains and along incised riv- conditions had not been so severe, structure. The limited scope of the er and creek channels. This type it is probable that the significant reconnaissance mission did not allow of failure generally involves the flow characteristics commonly for the team to determine how many upper 5-10 feet of soil, and slip exhibited by these failures would of the landslides were reactivated appears generally to be concen- not have been so prominent.