Available online at www.sciencedirect.com ScienceDirect

Procedia and Planetary Science 16 ( 2016 ) 61 – 70

The Fourth Italian Workshop on Landslides Initial Pore Pressure Ratio in the Triggered Large-scale Landslide near Aratozawa in Miyagi Prefecture, Japan

Hendy Setiawana*, Kyoji Sassab, Kaoru Takaraa, Toyohiko Miyagic, Hiroshi Fukuokad

aDisaster Prevention Research Institute (DPRI), Kyoto University, Uji 611-0011, Japan bInternational Consortium on Landslides (ICL), Kyoto 606-8226, Japan cTohoku-Gakuin University, Sendai, Miyagi 981-3193, Japan dResearch Institute for Natural and Recovery, Niigata University, Niigata 950-2181, Japan

Abstract

The Iwate-Miyagi inland earthquake with magnitude of 7.2 on 14 June 2008 in Tohoku region of Japan resulted in more than 4,000 landslides. The deep large-scale landslide near Aratozawa Dam in Ohu of Miyagi Prefecture occurred shortly after the earthquake. In this paper, landslide geotechnical simulation with high normal stress and pore-water pressure measurement in undrained conditions is explained through laboratory experiment by means of ring shear tests. We used the newest version of undrained dynamic loading ring shear apparatus to test volcanic tuff samples from the collapsed zone and flank side of the Aratozawa landslide. The simulation results of combined triggering factors using ring shear apparatus implied that the initiation mechanism of Aratozawa landslide was influenced by initial pore pressure before being triggered by the earthquake. © 2016 The The Authors. Authors. Published Published by Elsevierby Elsevier B.V. B.V.This is an open access article under the CC BY-NC-ND license Peer-review(http://creativecommons.org/licenses/by-nc-nd/4.0/ under responsibility of the organizing). committee of IWL 2015 Peer-review under responsibility of the organizing committee of IWL 2015 Keywords:Iwate-Miyagi earthquake; deep landslide; ring shear tests; initial pore pressure ratio

1. Introduction

The Iwate-Miyagi Nairiku (inland) earthquake with magnitude of 7.2 occurred on 14 June 2008 at 8:43 JST in Tohoku region of Japan with the epicenter located at 39°01.7’ of north latitude and 140.9°52.8’ of east longitude1,10,22(Fig.1). The depth of the earthquake was reported of about 8 km with the source area beneath the Ohu Mountains5. The Ohu Mountains stretches from north to south of Tohoku region across Iwate and Miyagi Prefecture.

* Corresponding author. Tel.: +81-774-38-4129; fax: +81-774-38-4130. E-mail address: hendy@.dpri.kyoto-u.ac.jp

1878-5220 © 2016 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). Peer-review under responsibility of the organizing committee of IWL 2015 doi: 10.1016/j.proeps.2016.10.007 62 Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70

The type of this earthquake was a reverse faultwith the surface of found from east of the epicenter to the southwest besides Mount Kurikoma1,5.The 2008 Iwate-Miyagi earthquake resulted inmore than 4,000 landslides, mainly in the form of shallow slide and deep-seated slide21. Those thousands of landslides destroyed network or blocked the rivers and reservoirs, with landslides distribution up to 15 kilometers from the epicenter3, 10. Along the fault line, most of the large-scale landslides were found on the hanging wall side where Mount Kurikomais located21,7,5,10. There are three dam reservoirswith their upstream coming from Mount Kurikoma: Hanayama dam which store the water of Ichihasama River, Aratozawa dam for Nihasama River and Kurikoma dam for Sanhasama River. These three main rivers (Hasama Rivers) from Mount Kurikoma are part of Kitakami river basin which is the biggest river basin in Tohoku region. During the 2008 Iwate-Miyagi Nairiku earthquake, numerous slope failures, landslides and debris flows took places on the Hasama Rivers. For example, theDozousawa shallow landslide generated long-traveling near the top-east of Mount Kurikomaand hit the Komanoyu hot spa hotel and Gyoja with the flow distance more than 10 km on the upstream of Sanhasama River5,6,21. Other large-scale landslide was occurred near the Aratozawa Dam reservoir on the upstream of Nihasama River, which is the main concern in this study.

Fig.1. Distribution of peak ground accelerationof the 2008 Iwate-Miyagi Nairiku earthquake (K-Net and KiK-Net data from NIED, Japan)

The deep large-scale landslide near Aratozawa Dam reservoir on the southeastern part of Mount Kurikoma of Miyagi Prefecture occurred shortly after the earthquake (Fig.2). Aratozawa landslide is located 14 km south- southwest of the epicenter, experienced peak ground acceleration of more than 1,000 gal7. In addition, this landslide hada gentle gradient varied for about 2-4 degree with height of the head scarp of 50-150 m which moved about 300 m and resulted inmassive blocks of 1,300 m in length, 900 m in width and more than 100 m deep3,5,7. Such aunique deep large-scale landslide is importantto be studied in detail particularly for the mechanism of landslide initiation. The Aratozawa landslide is located just upstream of the reservoir. Although the 2008 Iwate-Miyagi earthquake as a triggering factor has been reported based on a geological site investigation and analysis7,8,10,22, the mechanism of landslide initiation and its causal factors have not yet been clarified in detail. Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70 63

Fig.2. The deep large-scale landslide near Aratozawa Dam reservoir (photo by H. Fukuoka)

In this paper, geotechnical simulation of the Aratozawa landslide with high normal stress and pore-water pressure measurement in undrained conditions is explained through laboratory experiment by means of ring shear tests. We used the newest version of undrained dynamic loading ring shear apparatus to test volcanic tuff samples from the collapse zone and flank side of the Aratozawa landslide. In order to observe generation, this paper focuses on the 2008 Iwate-Miyagi earthquake loading given to the Aratozawa samples within high normal stress, normally consolidated and in undrained condition.

2. Materials and Methods

2.1. Samples from the Aratozawa landslide

Mount Kurikomaislocated in the border between Iwate and Miyagi Prefecture. Geological feature in this area mainly consists of tertiary and quaternary volcanic rocks. Geological structures in Aratozawa area which is near to Mount Kurikomaare composed of lapilli tuff, black mudstone, alternating beds of sandstone and silty mudstone, very thick layer of pumice tuff, thin layers of mudstone, silty and sandy clay, columnar joints layer of welded tuff and debris flow deposit on the surface layer7,8. The debris flow deposit on the surface layer was believed to come from the pyroclastic material of Mount Kurikoma. From the interpretation of major landform7, the Aratozawa landslide was occurred in the transition area between foot of Mount Kurikoma and hilly slope of the upstream Nihasama River where Aratozawa dam reservoir islocated. We conducted site investigation to observe the Aratozawa landslide, particularly near the head scarp, huge landslide blocks and ridges, debris material at the flank side and depression area at the toe part. The quaternary layer was slightly visible during investigation near the head scarp (Fig.3). The wall of head scarp clearly shows the upper thick pumice tuff layer, though it is believed that the rest of quaternary layer, interface with sliding surface and down to the layer was buried by the collapsed materials of debris flow deposit and welded tuff. To conduct laboratory test using ring shear apparatus for landslide initiation mechanism, we took disturbed samples from the exposed pumice tuff debris in two locations (Fig.4). First, sampling of squeezed tuff along cracks within the major landslide block in collapse zone was carried out. Another sample was taken from the flank side. We found a mound of pumice tuff on the flank side of landslide body. Professor Miyagi of Tohoku-Gakuin University estimated that the pumice tuff was squeezed out from the bottom of the major landslide block to the surfaceand formed as a mound on the flank side.We generally believed that samples in thesetwo areasare the key to study the initiation mechanism of deep large-scale landslide event in 2008 with suspected sliding surface in the depth of more than 100 m. 64 Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70

Fig.3. Head scarp of the Aratozawa landslide (photo from site investigation)

Grain size distribution of samples from the Aratozawa landslide is given in Table 1and Fig.4below. The average particle size of the Aratozawa samples in two locations are less than 0.1 mm, with the effective particle size of less than 0.02 mm and the coefficient of gradation more than 1.00. Although samples in two locations show a typical result in grain size distribution, the sample in collapse zone has the finer material than the flank side.

Table 1.Samples of the Aratozawa landslide for ring shear tests.

Specification Collapse zone Flank side Specific gravity 2.06 2.07 Average particle size, D50 (mm) 0.022 0.070 Effective particle size, D10 (mm) 0.0038 0.0131 Particle size of D30 (mm) 0.0014 0.044 Coefficient of uniformity, Cu 6.61 6.35 Coefficient of curvature, Cc 2.16 1.85

Fig.4. Samples location and grain size distribution of the Aratozawa landslide (photo by H. Fukuoka)

2.2. Undrained dynamic loading ring shear apparatus

The fundamental concept of ring shear apparatus is to test soil samples from the landslide source, sliding surface or deposited landslide mass area. The test itself is carried out by applying normal stresses due to gravity, static shear Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70 65 stress from the view of original slope angle, preexisting pore pressure in the specimens and seismic loading taken from the earthquake seismic record. The development of ring shear apparatus in the Disaster Prevention Research Institute (DPRI) of Kyoto University has been conducted since 1984 by Sassa and other colleagues13,14,15. The main concern in landslide simulation using ring shear apparatus is to address the undrained shear test with pore-water pressure generation monitoring and shear resistance observation at a large shear displacement during initiation of failure, mobilized-rotating stage as a post-failure motion and steady state condition. Since 2010, the development of ring shear apparatus by Sassa and other colleagues has moved from the pillar-beam based, to the single central axis based for applyingthe normal stress11,12. This change has some beneficial processes such as smaller size of apparatus and reducing the oscillation which influences the servo-control system of normal stress in undrained conditions and sudden change of stresses during cyclic and seismic loading. For testing the Aratozawa samples, we used the undrained dynamic loading ring shear apparatus ICL-2 version (ICL stands for the International Consortium on Landslides). This latest version was developed by Sassa and other colleagues in the ICL as a part of SATREPS (Science and Technology Research Partnership for Sustainable Development)Project between Japan and Vietnam which is started in 2012, and also utilized for the case of deep landslide in Unzen-Mayuyama of Japan11. The ICL-2 has high performances to apply the stress-pressure up to 3,000 kPa that could represent the deep landslide tests that need high stress and pressure in undrained conditions (Table 2). Similar to the previous DPRI version, the successor series of ring shear apparatus in the ICL-2 could perform drained and undrained consolidation tests, monotonic shear and speed-controlled tests, pore-water pressure fluctuation controlled test, undrained dynamic, continuous cyclic and earthquake loading tests.

Table 2.Features of the undrained dynamic loading ring shear apparatus ICL-2.

Specification Unit Capacity Shear box: Inner diameter mm 100 Outer diameter mm 142 Maximum height of sample mm 52 Ratio of maximum height/width - 2.48 Shear area of sample cm2 79.79 Maximum shear speed cm/s 50 Maximum normal stress kPa 3,000 Resolution of gap control system mm 0.001 Maximum data acquisition rate per second 1,000 Maximum frequency of loading Hz 20

The ring shear apparatus ICL-2 basically has five components11,12: (1) Undrained ring shear box and loading system; (2) Main control unit; (3) Monitoring control unit; (4) Separated servo generator with back pressure unit; and (5) Vacuum pump de-aired water system. The upper part of the undrained ring shear box consists of inner and outer rings (Fig.5). The inner ring shear box has a cylinder shape which is connected to the central axis (CR), while the pore pressure transducer (P) is installed to the outer box which is connected to the circumferential gutter inside shear box that is facing directly to the samples (S). The single central axis in the ICL-2 is pulled by piston which is controlled by servo-valve (SV) using feedback signal from load cell (N) to servo amplifier (SA) and back to load cell through servo motor (SM). The back pressure was installed integrated with the servo generator which is controlled with oil pump, similar with the normal stress loading system. 66 Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70

Fig.5. Structure of the ring shear apparatus ICL-2 and its loading system (Sassa et al. 2014)

2.3. Testing procedures

The Aratozawa sample is composed of volcanic pumice tuff material which is very fine and soft. In the ring shear tests, all samples were prepared carefully in fully saturated conditions by adding the de-aired water to the oven-dried samples. The immersed samples with de-aired water are then left in a vacuum tank9. For Aratozawa landslide, the samples preparation was taken in two days before the test was carried out. After the installation of upper inner and outer ring shear box, the gap adjustment was conducted. The gap adjustment through gap sensor (GS) aims to avoid water leakage during the high stress while minimizing the rubber edge between inner-outer ring shear box and rubber edges at the lower rotatable part of the apparatus16,17. The adjusted gap value is kept constant during rotation and loaded stress in the samples so data acquisition of the shear displacement and pore pressure generation could be monitored with no leakage occurred. Before pouring the immersed samples into the box, the CO2 and de-aired water circulation were used to avoid the entrapped air inside the shear box. The degree of saturation was checked using the parameter of saturation degree in the direct shear state, called BD value15,16,18,19. The BD value is the ratio between the increment of generated pore water pressure (u) and the increment of normal stress () in the samples which are loaded in the undrained condition, formulated as BD = u/. The sample inside the ring shear box is nearly in the saturated condition when the BD value reaches more than 0.9515,20. After installing the loading plate which is connected to the central axis (CR) and shear stress load cell (S1, S2), the pore pressure transducer (P) is then plugged in to the valve on the outer of ring shear box (Fig.5). The drainage valve exists on the upper of loading cell while water supply valve is located on the lower of rotatable ring shear part close to the shear displacement sensor (SD). Another valve for water-pressure is located near the valve of pore-water pressure transducer on the outer of ring shear box. This valve is used during the pore-water pressure controlled test. The samples are then consolidated in drained conditions under normal stress and static shear stress. The values of stresses were assumed based on the landslide geometry (depth, slope angle) and soil properties. For Aratozawa samples, the depth of suspected sliding surface was represented around 3,000 kPa of normal stress in the test as the maximum performance capacity of ICL-2. Meanwhile, the static shear stresses were given differently about 1,100 kPa which represents initiation of block failure near the head scarp and 500 kPa of a gentle slope. Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70 67

3. Testing Results and Discussion

3.1. The 2008 Iwate-Miyagi Nairiku earthquake

The Iwate-Miyagi Nairiku inland earthquake observed data were obtained from the National Research Institute for Earth Science and Disaster Prevention (NIED).In this paper, we used the data from Tsukidate station (MYG004) with peak ground acceleration (PGA) of 740 gal in North-South direction and about 678 gal in East-West direction (Fig.6).

NS

EW

UD

Fig.6. The PGA of Iwate-Miyagi Nairiku earthquake in 2008 at MYG004 (NIED, Japan)

To simplify the Aratozawa landslide simulation using ring shear apparatus ICL-2, we used the PGA with the direction perpendicular to the body slide moving direction. The earthquake record of NS direction on MYG004 with maximum acceleration of 740 gal was chosen. This earthquake record data were inserted (as an additional shear stress) into the apparatus with 5 times slower speed in order to precisely monitor the pore pressure generation behavior that fits to the servo-motor capacity. The additional shear stress as a manifestation of the earthquake was obtained by using the relationship between gravitational and seismic acceleration11.The gradient of sliding surface was about 2-4 degrees with type of landslide as translational block glide3,7. However, to ease the shear and pore-pressure measurement during tests using the ICL-2, the gradient of sliding surface was assumed a bit higher of about 9 degrees on the gentle slope and 20 degrees on the initial block failure. Through this assumption, the static shear stress was applied up to 500 kPa and 1,100 kPa. The total normal stress was given in 3,000 kPa. The landslide simulation by directly applying earthquake loading has generated only 270 kPa of pore pressure (Fig.7).

Fig.7. Earthquake test using ring shear apparatus of ICL-2 to the collapse zone of the Aratozawa landslide 68 Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70

Within very small increment of pore pressures, shear displacements were less than 3 mm in all tested samples which means almost no moving failure appeared. The test results indicated that earthquake loading cannot cause landslide motion if any pore pressure was not working on the sliding surface at the time of earthquake.

3.2. Pore-water pressure generation

The earthquake loading simulation on Aratozawa samples that conducted directly after drained static loading of normal stress and shear stress imply that the initial pore pressure was not zero. The water level of the Aratozawa reservoir before the earthquake was 268.5 m and when the landslide occurred was 270.9 m (Miyagi Prefecture Government) due to landslide debris flows into the reservoir. In addition to this, the suspected sliding surface was estimated in the range of 270-280 m. To find out the critical pore pressure in the failure state which triggers significant shear displacement, pore-pressure tests were carried out. Pore pressure controlled test was conducted by means of giving the additional water pressure to the shear box in undrained conditions after predefined static normal stress and shear stress were applied as normal consolidation. This test represents pore pressure rise on the site above the sliding surface. The additional water pressure was given with the rate about 1-2 kPa/s until the samples had failures. The pore pressure test results are shown in Fig.8.

Fig.8. Initial shear displacement and failure in pore pressure tests

The initial shear displacement and failure in pore pressure controlled tests were initiated by sample from collapse zone, followed by sample from flank side of the landslide body (Fig.8). About 2.5 minutes of data acquisition after initial shear displacement, the sample from collapse zone reached more than 8 m of shear displacement, while sample of flank side approached 5 m of shear displacement. During pore pressure test, pore pressure ratio (ru) could be calculated as the ratio between the increment of pore 2,4,16 water pressure (u) during undrained loading and the given total normal stress () . The values of (ru) are not greater than 1.0, but the liquefaction state in the landslide sliding surface occurred if the pore pressure ratio approached to 1.0. The Aratozawa landslide had a potential for generation of excess pore-water pressures where liquefaction might have occurred in the pumice tuff layer during the earthquake3. However, results from ring shear tests above indicated that the initial pore pressure might exist before landslide. In pore pressure controlled tests, the critical pore pressure ratio in Aratozawa samples reached 0.61-0.63 in the normal stress of 3,000 kPa and static shear stress of 1,100 kPa. It means that landslide motion will not occur until pore pressure ratio will reach up to 0.61-0.63, if any earthquake had not produced additional seismic stress (Fig.9). According to drilled hole near upper ridge for inspection by Tohoku Regional Forest Office in 2009, water pressure after landslide was experienced in the height of 70 m above suspected sliding surface. It is not possible to estimate the initial pore water pressure at the time of earthquake, so we assumed that the initial water Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70 69 pressure is possibly around 100 m above sliding surface. With the normal stress of 3,000 kPa given in the samples and unit weight of water is 9.8 kN/m3, we could use the maximum initial pore water pressure of 0.33.

Fig.9. Critical pore pressure ratio in collapse zone and flank side of the Aratozawa landslide

By knowing the critical pore pressure ratio and groundwater condition above, we applied the initial pore pressure ratio of 0.33 or a half from critical pore pressure ratio before the earthquake loading. The test results are shown in Fig.10.The ring shear simulation results presented that the Aratozawa landslide initiation mechanism was triggered by seismic shear-stress loading which was calculated from the Iwate-Miyagi Nairiku earthquake monitoring record under the initial pore water pressure of 0.33.

Fig.10. Ring shear test of earthquake triggered the Aratozawa landslide with predefined initial pore pressure

4. Conclusions

Landslide simulation in Aratozawa samples by directly applying earthquake loading through ring shear apparatus ICL-2 shows that pore pressure generated only up to 270 kPa in the condition of normal stress 3,000 kPa (Fig.7). We found that the critical pore pressure ratio of the samples reached 0.61-0.63 in pore pressure controlled test with the same normal stress rate. To observe the failure condition in combined factors, we applied initial pore pressure ratio of 0.33 followed by the Iwate-Miyagi earthquake loading. The test results show rapid increase of pore pressure during the main shock of earthquake loading with decrease in progressive failure motion (Fig.10). It is indicated that to achieve the failure condition due to Iwate-Miyagi earthquake, Aratozawa area experienced the initial pore pressure from groundwater level about 100 m in maximum above sliding surface. With high initial pore pressure ratio, the motion of large-scale Aratozawa landslide in the distance up to 300 m during Iwate-Miyagi earthquake in 2008 mightbe possible to occur3,7. 70 Hendy Setiawan et al. / Procedia Earth and Planetary Science 16 ( 2016 ) 61 – 70

Acknowledgements

The author would like to thank Kawanami Akiko from Ministry of Agriculture, and Fisheries of Japan who is in charge of the recovery works of Aratozawa dam and reservoir area, for kind assistance during two times of site investigations and sampling. The help from Japan Conservation Engineering and Japan Forestry Agency during investigation, and technical team from Marui Co., Ltd for the ring shear apparatus are gratefully acknowledged. This landslide research is supported by the Global Survivability Studies Program of Kyoto University (A Leading Graduate School Program of MEXT) and SATREPS (Science and Technology Research Partnership for Sustainable Development) Program of the Government of Japan. The undrained ring shear apparatus ICL-2 version is developed by the International Consortium on Landslides (ICL) as a part of SATREPS-Vietnam project in 2012-2017.

References

1. Aoi S, Kunugi T, Fujiwara H, Morikawa N, Suzuki W. Strong motions of the 2008 Iwate-Miyagi Nairiku earthquake-asymmetric ground motion and trampoline effect. NIED Major Disaster Investigation 2010; 43: 19-29 (in Japanese). 2. Fukuoka H, Sassa K, Wang G. Influence of shear speed and normal stress on the shear behaviour and shear zone structure of granular materials in naturally drained ring shear tests. Landslides 2007;4:63-74. 3. Gratchev I, Towhata I. Geotechnical characteristics of volcanic soil from seismically induced Aratozawa landslide, Japan. Landslides 2010;7:503-510 4. JurkoJ, Sassa K, Fukuoka H. Study on seismic behaviour of nonplastic by means of ring-shear apparatus. Landslides 2008;5:189-201. 5. Kazama M, Kataoka S, Uzuoka R. Volcanic area disaster caused by the Iwate-Miyagi Nairiku earthquake of 2008, Japan. and Foundations 2012;52(1):168-184. 6. Kazama M, Uzuoka R, Mori T. Earthquake induced debris flow and landslide in the Iwate-Miyagi Nairiku earthquake in 2008, Japan. The Fifth International Conference on Earthquake , , Chile; 2011. 7. Miyagi T, Yamashina S, Esaka F, Abe S. Massive landslide triggered by 2008 Iwate-Miyagi inland earthquake in the Aratozawa Dam area, Tohoku, Japan. Landslides 2011; 8: 99-108. 8. Moriya H, Abe S, Ogita S, Higaki D. Structure of the large-scale landslide at the upstream area of Aratozawa dam induced by the Iwate- Miyagi earthquake in 2008. Journal of the Japan Landslide Society 2010; 47(2): 77-83 (in Japanese). 9. Ostric M, Ljutic K, Krkac M, Setiawan H, He B, Sassa K. Undrained ring shear tests performed on samples from Kostanjek and Grohovo landslide. The 10th anniversary of ICL, Kyoto 2012. 10. Sammori T, Tada Y, Daimaru H, Murakami W, Yasuda Y, Noguchi S. Characteristic of sediment by the Iwate-Miyagi Nairiku earthquake in 2008. Bulletin of Forestry and Forest Product Research Institute, Tsukuba 2012; 11(3): 97-120 (in Japanese). 11. Sassa K, Dang K, He B, Takara K, Inoue K, Nagai O. A new high-stress undrained ring-shear apparatus and its application to the 1792 Unzen-Mayuyamamegaslide in Japan.Landslides 2014;11:827-842. 12. Sassa K, He B, Dang K, Nagai O, Takara K. Progress in landslide dynamics. In Sassa et al., editors. Landslide science for a safer geoenvironment, Vol.1.Springer International Publishing Switzerland; 2014.p.37-67. 13. Sassa K, He B, Miyagi T, Strasser M, Konagai K, Ostric M, Setiawan H, Takara K, Nagai O, Yamashiki K, Tutumi S. A hypothesis of the Senoumi submarine megaslide in Suruga Bay in Japan-based on the undrained dynamic-loading ring shear tests and computer simulation. Landslides 2012;9:439-455. 14. Sassa K, Nagai O, Solidum R, Yamazaki Y, Ohta H. An integrated model simulating the initiation and motion of earthquake and induced rapid landslides and its application to the 2006 Leyte landslide. Landslides 2010;7:219-236. 15. Sassa K, Fukuoka H, Wang GH, Ishikawa N. Undrained dynamic-loading ring-shear apparatus and its application to landslide dynamics. Landslides 2004; 1: 7-19. 16. Sassa K, Wang G, Fukuoka H. Performing undrained shear tests on saturated in a new intelligent type of ring shear apparatus. Geotechnical Testing Journal 2003;26(3):257-265. 17. Setiawan H, Sassa K, Takara K, Miyagi T, Fukuoka H, He B. The simulation of a deep large-scale landslide near Aratozawa dam using a 3.0 MPa undrained dynamic loading ring shear apparatus. In Sassa et al., editors. Landslide science for a safer geoenvironment, Vol.1.Springer International Publishing Switzerland; 2014.p.459-465. 18. Wafid A, Sassa K, Fukuoka H, Wang G. Evolution of shear-zone structure in undrained ring-shear tests. Landslides 2004;1:101-112. 19. Wang F, Sassa K, Wang G. Mechanism of a long-runout landslide triggered by the August 1998 heavy rainfall in Fukushima prefecture, Japan. Engineering 2002;63:169-185. 20. Wang G, Zhang D, Furuya G, Yang J. Pore-pressure generation and fluidization in a landslide triggered by the 1920 Haiyuan earthquake, China: a case study. 2014;174:36-45. 21. Yagi H, Sato G, Higaki D, Yamamoto M, Yamasaki T. Distribution and characteristics of landslides induced by the Iwate-Miyagi earthquake in 2008 in Tohoku District, Northeast Japan. Landslides 2009;6:335-344. 22. Yamashina S, Yamazaki T, Hashimoto J, Kasai S, Agatsuma T, Shibuya K. Aratozawa landslide triggered by the Iwate-Miyagi Nairiku earthquake in 2008. Journal of the Japan Landslide Society 2009; 45(5): 392-397 (in Japanese).