Soils and Foundations 2006年
出版

soils and Foundations

表示件数

10 30 50 100
タイトル 表紙(Soils and Foundations)
著者
出版 soils and Foundations
ページ -〜- 発行 2006/12/15 文書ID 20953
内容
ログイン
タイトル Contents(Soils and Foundations)
著者
出版 soils and Foundations
ページ -〜- 発行 2006/12/15 文書ID 20954
内容
ログイン
タイトル Foreword
著者 Ikuo Towhata
出版 soils and Foundations
ページ 発行 2006/12/15 文書ID 20955
内容 表示
ログイン
タイトル The 2004 Sediment Disasters in Quezon Province, Philippines Triggered by Heavy Rainfall
著者 R. P. Orense・S. E. Sapuay・E. B. Billedo・Kiyokata Matsuoka
出版 soils and Foundations
ページ 701〜711 発行 2006/12/15 文書ID 20956
内容 表示
ログイン
タイトル Evaluation of Run-out Distances of Slope Failures during 2004 Niigata-Ken Chuetsu Earthquake
著者 Yoshimichi Tsukamoto・Kenji Ishihara・Yasuhide Kobari
出版 soils and Foundations
ページ 713〜725 発行 2006/12/15 文書ID 20957
内容 表示
ログイン
タイトル Evaluation of Natural Slope Failures Induced by the 2004 Niigata-Ken Chuetsu Earthquake
著者 Hirofumi Toyota・J.Wang・Kouichi Nakamura・Naoki Sakai
出版 soils and Foundations
ページ 727〜738 発行 2006/12/15 文書ID 20958
内容 表示
ログイン
タイトル Damage to Earth Structures for National Highways by the 2004 Niigata-Ken Chuetsu Earthquake
著者 Junichi Koseki・Tetsuya Sasaki・Nichiro Wada・Junichi Hida・Masaki Endo・Yukika Tsutsumi
出版 soils and Foundations
ページ 739〜750 発行 2006/12/15 文書ID 20959
内容 表示
ログイン
タイトル Investigation and Restoration Countermeasure Work for the Slope Disaster Induced by the 2005 West off Fukuoka Earthquake
著者 G. Chen・Kouki Zen・Hideo Nagase・Kenichi Sato・Kiyoshi Omine・Taizo Kobayashi・Hidefumi Sato
出版 soils and Foundations
ページ 777〜791 発行 2006/12/15 文書ID 20960
内容 表示
ログイン
タイトル Damage to Residential Retaining Walls at the Genkai-Jima Island Induced by the 2005 Fukuoka-Ken Seiho-oki Earthquake
著者 Taizo Kobayashi・Kouki Zen・Noriyuki Yasufuku・Hideo Nagase・G. Chen・Kiyonobu Kasama・Akihiko Hirooka・Hiromu Wada・Yuji Onoyam
出版 soils and Foundations
ページ 793〜804 発行 2006/12/15 文書ID 20961
内容 表示
ログイン
タイトル slope Disasters Caused by Typhoon No. 14 of 2005 in Yamaguchi Prefecture
著者 Yoichi Sehara・Motoyuki Suzuki・Tetsuro Yamamoto・Takashi Terayama・Tomohiro Tomokiyo・Yoshifumi Kochi
出版 soils and Foundations
ページ 817〜830 発行 2006/12/15 文書ID 20962
内容 表示
ログイン
タイトル Remote Sensing Observations of Landslides and Ground Deformation from the 2004 Niigata Ken Chuetsu Earthquake
著者 E. Rathje・R. Kayen・K.-S. Woo
出版 soils and Foundations
ページ 831〜842 発行 2006/12/15 文書ID 20963
内容 表示
ログイン
タイトル Examination of Slope Hazard Assessment by Using Case Studies of Earthquake-and Rainfall-Induced Landslides
著者 Masanori Mizuhashi・Ikuo Towhata・Junichi Sato・Takashi Tsujimura
出版 soils and Foundations
ページ 843〜853 発行 2006/12/15 文書ID 20964
内容 表示
ログイン
タイトル Ground Damage Resulting from Torrential Rains in Fukui July 2004
著者 The Japanese Geotechnical Society: Emergency Survey Team for Ground Damage Resulting From Torrential Rains in Fukui, Jul
出版 soils and Foundations
ページ 869〜884 発行 2006/12/15 文書ID 20965
内容 表示
ログイン
タイトル Uplift of Sewage Manholes and Pipes during the 2004 Niigataken-Chuetsu Earthquake
著者 susumu Yasuda・Hiroyoshi Kiku
出版 soils and Foundations
ページ 885〜894 発行 2006/12/15 文書ID 20966
内容 表示
ログイン
タイトル slope Failures at Yokowatashi and Nagaoka College of Technology due To the 2004 Niigata-Ken Chuetsu Earthquake and Their Analytical Considerations
著者 Atsuo Onoue・Akihiko Wakai・Keizo Ugai・Kunihiro Higuchi・Kiyoshi Fukutake・Hiroyuki Hotta・Seiichiro Kuroda・Hideaki Nagai
出版 soils and Foundations
ページ 751〜764 発行 2006/12/15 文書ID 20967
内容 表示
ログイン
タイトル Effects of Nonlinear Properties of Surface Soils on Strong Ground Motions Recorded in Ojiya during 2004 Mid Niigata Prefecture Earthquake
著者 Kohji Tokimatsu・Toru Sekiguchi
出版 soils and Foundations
ページ 765〜775 発行 2006/12/15 文書ID 20968
内容 表示
ログイン
タイトル Zoning for Liquefaction and Damage to Port and Harbor Facilities and Others during the 2005 Fukuoka-Ken Seiho-oki Earthquake
著者 Hideo Nagase・Kouki Zen・Akihiko Hirooka・Noriyuki Yasufuku・Kiyonobu Kasama・Taizo Kobayashi・Yoshito Maeda・Kiyoshi Uno・Kenji
出版 soils and Foundations
ページ 805〜816 発行 2006/12/15 文書ID 20969
内容 表示
ログイン
タイトル Geodisasters in Kyushu Area Caused by Typhoon No.14 in September 2005
著者 Mitsuhiro Sezaki・Ryosuke Kitamura・Noriyuki Yasufuku・Akihiro Hirooka・Hidetoshi Ochiai・Hiroshi Yokota・Sigeki Yokoyama・Hiro
出版 soils and Foundations
ページ 855〜867 発行 2006/12/15 文書ID 20970
内容 表示
ログイン
タイトル JGS NEWS
著者
出版 soils and Foundations
ページ I〜II 発行 2006/12/15 文書ID 20971
内容 表示
ログイン
タイトル Contents and Indexes for Volume 46 (February 2006 to December 2006),List of Key Words
著者
出版 soils and Foundations
ページ III〜XI 発行 2006/12/15 文書ID 20972
内容 表示
ログイン
  • タイトル
  • 表紙(Soils and Foundations)
  • 著者
  • 出版
  • soils and Foundations
  • ページ
  • -〜-
  • 発行
  • 2006/12/15
  • 文書ID
  • 20953
  • 内容
  • ログイン
  • タイトル
  • Contents(Soils and Foundations)
  • 著者
  • 出版
  • soils and Foundations
  • ページ
  • -〜-
  • 発行
  • 2006/12/15
  • 文書ID
  • 20954
  • 内容
  • ログイン
  • タイトル
  • Foreword
  • 著者
  • Ikuo Towhata
  • 出版
  • soils and Foundations
  • ページ
  • 発行
  • 2006/12/15
  • 文書ID
  • 20955
  • 内容
  • FOREWORDIKUO TO vHATAi)Editor-in-Chief of Soils and Foundations JournalIn the year's of 2004 to 2006, the editing committee of this journal sa¥v many geotechnicai natural disasters to occurin Japan as ¥vell as in the world. While their' speciai attention ¥vas paid most significantly to those caused by theNiigata-Chuetsu earthquake on October 23r'd, ,_004, rainf'all-induced landslides ¥vhich were caused by many typhoonsin the summer of 2004 were remarkable as lvell. It is certainly necessary, moreover, to mention the slope instabilities atthe time of Fukuoka-Seiho-Oki earthquake on March ,-Oth, 2005. With these in mind, the editing committee decided toorganize an issue of Soils and Foundations Journal vhich is devoted to technical reports on those disasters.Since the theme of this issue includes many disasters, the editing committee expected numerous manuscripts to besubmitted. To accept as many articles as possible in one issue of the journal, a page limitation of 12 ¥vas decided. Af'telthe submitted manuscripts ¥vere revie ved with the same standards as those f'or regular issues of the journal, it ¥vasfound difficult in some cases to reduce the number of pages to 1'_. In this situation, the committee made a discussion toallo v a fe¥v extra pages to some articles, ¥vhile the rule for extr'a payment for extra pages was applied.One of the points of concern in this natural-disaster issue ¥vas the use of colored photographs. Every committeemember agreed the importance of colored photographs for correct understanding of geotechnical natural disasters,The printing of photos with colors is, ho¥vever', too costly for the administration of the Japanese Geotechnical Societyto afford. It is not a reasonable idea to char'ge the color-printing cost to authors or subscribers. Understandinga thissituation, the editing committee decided to use the web site of the journal. Although the photographs are printed inthis journal issue in black-and-¥vhite manner's, the same photographs lvith colors can be do¥¥'nloaded by readers free of'charge.URL of the site for downloading photographs: http://¥v¥vw.jiban.or.jp/e/sf/sf.htmlOnly those photographs that the author agreed are a¥'ailable in this site.This downloading site will be open to the public for a fe v years after this publication, and readers are encouragedto collect impor'tant information through this visual mediurn. Being combined ¥vith the written articles, thosephotographs vill help readers understand the facts of natural disaster's in more details.It is true that there are still many important natural disasters ¥vhich are missing in this issue. They are, for exampie,the one caused by the 2004 Indian Ocean earthquake and the 2005 Kashmir' earthquake in Pakistan. Although theformer' is kno¥vn widelyvith tsunami tragedies, it ¥vas actually associated ¥vit.h geotechnical damage caused by tsunarnierosion as ¥vell (Photo 1). The earthquake in Pakistan caused slope pr'oblems at many places (Photo 2), leading toproblems in transportation and evacuation. Similar problems ¥vill be repeated from no¥v on. Moreover, developrnentof infrastructures ¥vill shed lights on different types of geotechnical hazar'ds which have not so far been studied. Theediting committee ¥vill ¥velcome submission of papers concerning those events.- :;** ;'+';;" ; 1Photo 1. Erosion of ground due to tsunami action (Bandn Aceh,Sumatra, Indonesia)i} professor, Schooi of Engineering, The University of Tokyo, Japan,Photo 2. Slope failure neaF Muzaffarabad, Pakistan
  • ログイン
  • タイトル
  • The 2004 Sediment Disasters in Quezon Province, Philippines Triggered by Heavy Rainfall
  • 著者
  • R. P. Orense・S. E. Sapuay・E. B. Billedo・Kiyokata Matsuoka
  • 出版
  • soils and Foundations
  • ページ
  • 701〜711
  • 発行
  • 2006/12/15
  • 文書ID
  • 20956
  • 内容
  • SOILS AND FOUi¥,DATIONSNo46,¥ro I .Dec. 20066. 701-71 1 ,Japanese Geolechnical Societ}THE 2004 SEDIMENT DISASTERS IN QUEZON PROVINCE,PHILIPPINES TRIGGERED BY HEAVY RAINFALLROLANDO P. ORENSEi), SAh,IUEL E. SAPUAYii}, EL*¥"IER B. BILLEDoi i) and KIYOKATA MATSUOKAi+)ABSTRACTSevere rainfall from mid-November to early December 2004 triggered numerous sedirnent disasters and flashfloodson the eastern coast of Luzon Island in the Philippines, resulting in extensive damage to life and property. Amongthe affected areas, the coastal tolvns of Quezon Province sufi red the most dama*'e, ¥¥'ith numerous landslides occurring in the town of Real and debris fio¥¥'s bur'ying the to¥ 'ns of Infanta and General Nakar. This paper outlines theobservations f'rom reconnaissance ¥vorks conducted at the affected sites follo¥ving the disaster, with emphasis on thehydro-geological aspects of the disaster and its impact on civil engineering structures and other infrastructures.Moreover, the results of unsaturated-saturated seepage analyses conducted to investigate the mechanism of thecollapse of one of the slopes ¥vhich resulted in death of more than 100 people is presented.Key words: debris flo v, Iandslide, rainfall, site investigation, slope (IGC: B3 IC4)S:INTRODUCTIONSii o' e. 2.2:'; e.Severe rainfall from mid-No¥'ember to early December2004 caused by four successive typhoons and tropicalcyclones triggered numerous landslides and flashfloodson the eastern coast of Luzon Island in the Philippines(see Fig. l). The nurnber of deaths and missing peopleexceeded 1600 and the cost of' damage was estimated atUSS78.2 million. The province of Quezon, Iocated on theeastern side of the Sierra Madre mountain range, sufferedthe most damage due to landslides, debris flows and fiashS .I; j';:srl er'.i ///;; {Ar/; F'fS '::"ii - tl;i*" F: )'worstCapftalaffeetedAreas} !s':sed'-!T Sf Wefst'ajfected Provinces:,S: : E] 'State uf Cala titr dedafed''/ti: l:,i2/a r'ft e j;'; ' ;;i^f!! ¥;X" "I/;: t'! '* sl 1': _c : ; J'S sfloods induced by the heavy rainfall. Hardest hit ¥vere the;f:s!than half of the fatalities occurring in these three t.owns.:;'?' ;*' " .*. r ;eF 'lE!ss!towns of Real, Infanta and General Nakar, ¥vith moreif*'ar; #;{ /'set' 'E* ;-- Er'* 'T?teSf:e; es i ;,_' J I'/ i; -'*;L' - ;i )_ v'_. . t' :'t,j. .. * ;. L '/ 's' s /In addition, damage to civil engineering structures wasalso extensive, with numerous houses, several bridges andslope protection measures (e.g., riprap, retaining wallsand revetments) washed a¥vay by the moving debris ande SJ ; :;' : !: ;'1't$:Is 1 si'Ss's' '' "!i" ' ' /s.;f. ,' . :' es!ti : ,;'T:'flash floods.In this paper, the observations derived from postFig. 1. Areas afrected by the heavy rainfall lvhich occurred on ¥. 'ovem-disaster investigation undertaken in the three towns arereported, and the causes of slope failures and damage tocivil engineering structures are discussed. Furthermore,berDecember 2004 (modified from OCHA, 2004)the results of saturated/unsaturated seepage analysesTOPOGRAPHIC AND GEOLOGIC SE1'TINGperformed to simulate the failure of one of the slopes arepresented. The shear strength parameters back-calculatedfrom analyses can be used to evaluate the stability ofQuezon is an elongated province on the eastern coast ofLuzon, the Philippines' main island. The norther'n partsimilar' slopes in the area for the purpose of preventingof the province is located bet¥veen the Sierra Madrefuture disasters.mountain range and the Philippine Sea. Several br'anchesAssociate Professor. Depar lTlent of Civil Engineering. Yamaguchi University, Ube City, Japan (orense( :.,yamaguchi-u,ac.jp).::l::)iY,President, Infra-tech Systems Consultants, Pasig City. PhilippinesC hief Geologist, DENR l¥"lines and Geosciences Bureau, lvlanila. Philippines,Engineer. M.K. C*onsultant (, o.. Ltd., Fukuoka, JapanThe manuscript for ihis paper vas received for revie v on ivlay l, 2006; appro¥'ed on October 2, 2006,¥Vritten discussions on lhis paper should be submitted before July I , 2007 to the Japanese Geotechrrical Society, 4-3S-2. Sengoku, Bunkyo-ku,Tok_vo 1 12-001 1. Japan. Upon request the closing date ma_v be extended one month.f O1 TO_-ORENSE ET AL_of the 1,300 km-long Philippine fault system passthrou h the eastern side of the mountain rang:e in thepro¥'ince, running in the NS direction.Real is a coastal to vn of Quezon Pr'ovince located atthe eastern foothills of the Sierra Madre. Natural ve*'etation in the ar'ea is predominantly second-gro¥vth forests.Primary forests in the area have significantly declinedduring the logging boom in the 1960s. A narr'o¥v strip offlat lo¥vland exists bet¥veen the mountain and sea, for'ming the coastal area of Real. On the other hand, the to¥vnsof Infanta and General Nakar' are located north of Realin the coastal plains at the foot of the Sierr'a Madre. Thepresent active channel of Agos River, one of the manyr'iver systems that drain the mountain range, traver'sesthrough these to¥vns and cirains to the sea. Thus, compared to Reai ¥vhere steep mountains r'ise directly fromthe shoreline, the eastem parts of these t¥vo to¥vns arelocated in the allu¥'ial fan formed by the deposition oferoded sediments transported by Agos River from thehighlands of the Sierra Madre.Hill slopes are relatively steep, ¥vith some locationshaving :radients as high as 30-500/0. Basaltic volcanicsare the primary geologic formation in the upland regions¥vhile surface soils are general ¥veathered deposits.Sprin*'s are prevalent along the mountain roads and areused by residents for ¥vater supplies. Coastal plains have400300- 200c:o(O 100o1 O!30 1 11!4 1/i4 1 1 !19 1 1!24 1 1 /29 1 2/41 i 19Fig. 2. Daih., ainfall data in Infanta station observed in November2004 (data from PAGASA)attacked Luzon on 02 December, ¥vith 220-'_40 km/hr¥vinds, adding more rain to the already inundated region.Records obtained by a PAGASA rain gauge in Infantastation (located In Quez,on Province) for the month ofNovember' 2004 sho¥vs that the average daily rainfall lvasabout 80 mm durin*" the passa_ e of typhoon Unding(Fig. '-). Moreover, on the morning of 29 Novemberduring the passage of storm Winnie, 342 mm of rain fellduring a 9-hr period at Infanta station. Unfortunately,the flashfioods caused by this heavy rainfall ¥vashed a¥vaythick alluvial deposits consistin*" of unconsolidated sedi-the rain au e. The total rainfall recorded in Infantastation for the month of November ¥vas 979 mm lvhile forments (clay, sand and gravel),the period bet¥veen 1629 November, rainfall ¥vas 614mm.PRECIPITATION CHARACTERISTICSThe Tropical Rainfall Measuring lvlission (TRlvlM)based in NASA Goddard Space Fli_9:ht Center monitorsOtlt!ille of Weather Cor7c!itionDue to its archipela_9:o characteristics and geographicallocation, the Philippines has a tropical maritime climate.rainfall over the global tropics. Near-real time Multisatellite Precipitation Analysis (lvIPA) rainfall totalsmonitored in the Philippines for the period bet¥veen 16Much of the rainfall comes from tropical cyclones, whichdevelop over the Pacific Ocean and move ¥vestlvard acrossNovember-03 December indicated that along thethe Philippines into the South China Sea. These areprimarily. due to Violeta. Winnie and Yoyon_g: (TRMM,lo¥v atmospheric pressure areas of tropical origin charac-2004).terized by strong lvinds and normally accompanied byrainfall. Depending on their lvind speed, tropical cy-abruptly terminated, the recorded rainfall data ¥ver'e morec]ones are classified by., the country's ¥veather bureauor less consistent ¥vith those monitored by TRMM.eastern side of Luz,on, total rainfall exceeded 1 100 mm,Considering that the PAGASA rainfall record ¥vasPAGASA (Philippine Atmospherlc Geophysical andAstronomical Services Administration) as tropicaldepression (maxirnum lvinds bet¥veen 35 and 64 km/hr),Col77pa/'ison with Previous Rainfa!! Recol'clsEvery year, Infanta is usually the first port of call fortropical storm (bet¥veen 65 and 118 km/hr), typhoontropical cyclones that enter the Philippines. A summary(bet¥veen 1 18 and 200 km/hr), and super-typhoon (maxirnum ¥vinds exceeding 200 km/hr). Among the countriesin the lvorld, the Philippines has the most number ofpassage of tropical cyclones, avera_g:in*"_O tropical cyclones per year (PACJASA ¥vebsite)^of the annual precipitation data from 1951-'_002 obtained by PAGASA is sho vn in Fig. 3(a). It can bePrecipitation during Noven7ber-Decelllber 2004Initially, typhoon "Unding" struck the nor'th-centralPhilippines in mid-November ?-004. The ¥vorst hit area¥vas the pro¥*ince of Mindoro Oriental ¥vhere more than2000 houses ¥vere destroyed. Follo¥ving this typhoon,observed that almost every year, the region is subjected to¥'ery high precipitatlon, ¥vith an annual a¥'erage of 4016mm, probably the highest mean intensity among allPAGASA stations in the country.The average monthly precipitation from 1951-'_OO'_,together ¥vith the maximum precipitation observed foreach month ¥vithin the same time frame, are sho¥vn inNovember. Next another tropical storm "¥x rjnnie" struckFig. 3(b). The average monthly precipitation during therainy season (October to December) is about 610 mm. Onthe other hand, during periods of lo¥v rainfall (March toMay), the average is about '-OO mm. This indicates thatLuz,on on '_9 No¥*ember. Finally, typhoon "Yoyong"the region has practically no dry season, ¥vith almost eventropical storm "Violeta" passed over Luzon on )_2 SEDllvIENT DISASTERS IN THE PHILIPPINESin the area lvithin this 18-day period.7000(a)The National Water Information Net vork (NWIN), aeooocomputer-based netlvork system that electronically linksvarious databases of the ¥vater data collection agencies inthe Philippines, has been analyzing rainfall data obtained5000a(_4000by PAGASA. The maximum monthly 1-day rainfallrecorded in Infanta station from 19612000 is sho¥vn in< 3000Fig. 4(a) and yearly 1-day maxima are sho¥vn in Fig. 4(b).The hi**hest recorded 1-day rainfall in the station is 33920001 95019eo1 970 i 9eOi 9902000Year(b)703mm. This is less than the 342 rnm of rainfall recordedduring the 9-hr period on 29 November 2004 before therain gau*"e ¥vas lvashed a¥vay by a flashflood.2000Assuming that rainfall during the passage of tropicalstorm Winnle on '_9 No¥'ember '_004, continued at theE 1500Esame intensity for the next 3 hrs, a 12-hr rainfall intensityIOOoequivalent to 456 rnm ¥vould have been recol'ded. This500¥'alue is almost the same as the half-day (12-hr) intensityof' 462 mm for a rainfall event in the region with a return:::period of 100 years, as calculated by the NWIN. Thus, itcan be concluded that the precipitation during theoJan Feb M r Apr May Jvn Jui Aug Sep Oct Nov DeeMonthpassage of the series of tropical cyclones in '-004extraordinary event, oneFig. 3. Rainfall tlata in lrfanta station from 1951-2002: (a) annualrainfaH and (b) average and maximum monthty rainfaU (tlata fromPAGASA)vas anvith a return period of about100 years.DAMAGE CHARACTERISTICS(a) 400Dalnage Out!ineThe combined impact of the floods and landslidesE:_ 300caused significant loss of life and damage to property,severe disr'uption to daily life and damage to infrastruc-ture in the country. Disaster reports (ADRC, '_004;e: 200>NDCC*, '_004) indicate that more than 1600 people ¥verekilled or missing, 102,3 ¥vere injured and 880,000 ¥vere1,r!: iOO:displaced. Moreover, 38,000 houses were totally destroyed while another 134,000 houses ¥vere damaged. Ato(b)Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Decleast li road net¥vorks ¥vere disrupted ¥ 'hile 25 bridgesMonthlver'e rendered impassable due to landslides, scoured/washed-out bridge approaches, dama*・ed bailey bridges,4cocollapsed spans and scoured brid**e piers. Damage tos50EE'1:_c:a:C(crops, fishing and other infrastructures (including schoolbuildings and health f'acilities) exceeded USS78.2 million.3002so200Danlage in Rea! ToTvn, Quezon ProvinceT:1 50xlooNumerous landslides of all sizes ¥vere observed in thesoutheastern part of Sierra Madre, especially along thecoastal road of Real. These rainfall-induced landslideshcl'c!'50occurred preferentially on moderate to steep slopes,o1 9eo1 97019801 99020OoYea rFig. 4. Maximum 1・day rainfall observed in Infanta station fromdrainage divides, ¥'alley heads and along road cuts 'iththeir cro¥vn coinciding vith ridge crests. The landslides¥vere mainly soil and rock slides or r'otational slumps1961-2000: (a) monthh.' variation and (b) annilal variation (tlatainitiated by high pore-water pressuresfrom PAGASA as analyzed b, N¥VIN, 2002)underlying bedrock. Some of these slides transformeddistribution of rainfall throughout the year.into debris flows as the failed slope materials underwentrnixing hvith additional ¥vater and became fluidized duringSince TRMM measured a rainfall intensity of aboutvithin the soil andthe do¥vnslope movement.1 100 mm bet¥veen 16 November-03 Decernber 2004, it canMost of the landslides ¥vere shallolv, 0.5-1.0 m deep,be surmised that about 300/0 of the average annuallargely composed of' surficial soils. The r'elatively deeperprecipitation in Infanta and approximately t¥vice thelandslides (2-3 m deep) were cornposed of residual soilsder'ived from the highly ¥veathered and highly fracturedaverage monthly precipitation during the rainy season fell ORENSE ET AL.704** ;,.,・"'=:l""f'"_'i"i'" _" '"' -'_-**"i_* ; :'i: *r;'' 'i{;{: {ii.i."" ii'-' '-+* *S' -i : S'* - ' s " "; '*-"*""*"i /*+ ":s ' _:::1:, ::::;s ;"-iii: ;' 'f/"'__{";:: ::: {; ; :1'j'^il:: ';:.is,:.S=;; - - "-- 's -'-::';i- ::*:]j :;:::'--'c 5; . -;:' '_ ;-. ;';{? :* : i ;; ' s;=' _ ' *l}'r; :' -- '** t 'l-=*'_**' ' _' '- * ;' ' ;' *S ; : * '" *' '; '' - ":; i-" : i:"'... f; S:S#;SS i'-S {"Fig. 5. Damagetl bridge railing due to the impact of logs andsediment*laden floodlvater;:: .' ' - iS"' ,iSi#';#;#;:" ;:.';-'. ;- "Fio*. 7. View of landslide in Real tolvn lvhere more tl]an 100 peoplelvere killed******, ""**;"""'1{;:"":::s=""';"f,****** ::**,*S;SS*=. .; :iS;+,i; - + '';i ; *--_ ;' j /* - *:Fig. 6. A house in Quezon buried bl_' iogs and mud (Photo bl.' N.Resplandor)* ""'" ' s I '*-S '*^'i*!""i'*'*+;:;::;/::;:*i:' 1* , .;;i i+;""*.: ' * ; . __"' it;i;;:・"+"*'#'.f i;i'.:'*.,i "* '"" '.i;;:;,i *: ::***:t:_ *?- + 'Fig. 8. Debris and iogs covering the coasta roa:d of Rea : ¥^ ,ote thepresence of big boulders amon"* the debris (Photo by N. Respland o r)rocks. Slope collapse mainly occurred along the boundary of the un¥veathered zone.During these storms, areas ¥vith good vegetation coveralso experienced landsliding. Trees ¥vere uprooted fromthese mountains and slid do¥vn the slopes together ¥vithLogs and debris blocked major portions of the nationalhigh¥vay, renderin_9: Real and adjacent to¥vns inaccessibleto traffic for sever'al ¥veeks (Fig. 8). At least four bridgesin the town were destroyed by the sediment- and debrisladen flood¥vaters. Real Bridge ¥vas ¥vashed out ¥vhenlarge masses of soil. Dama_ e ¥vas most frequently causedby drifting logs striking bridges (Fig. 5) or drift¥vooddeposited in residential areas (Fig. 6). In some areas,dressed logs ¥vere transported from the mountains ¥vithhigh-¥'elocity flood¥vaters rushin_g do¥vn fr'om the mountain scoured the pier foundations, Ieading to the collapseof the bridge (see Fig. 9). At brid**es, debris and logsthe cascadingpiled up and jammed the side of the piers, forcing thevater, hintin*' the possibility of log'*in_",_¥vhether legal or ille*'al, as the main culprit of the disaster.Some of the slides plo¥ved through r'esidential areassituated at the foot of the mountain. In one of the ¥vorsthit areas, about 100 persons ¥vere trapped ¥vhen a threestorey buildlng being used as an evacuation site collapsedunder the impact of one landslide (Fig. 7). A high¥vayseparated the building and the foot of the mountain. Thesaturated mass of soil and debris, ¥vith volume estimatedat about 20,000-30,000 m3, came do¥vn rushing from themountain, crossed the road and smashed into the building, as it plo¥ved to¥vard the sea. The total distancetraveled by the moving mass ¥vas about 100 m. Continuous bad ¥1'eather and lack of equipment hampered rescueefforts. Eleven days after the disaster, four survivors ¥veremiraculously pulled out of the collapsed building.moving current to seek aiternative flo¥v paths.Moreover, a portion of Tanauan Road collapsed ¥vhenhigh-pressure ¥vater from the mountain scoured the outletof the drainage culverts located just belo¥v the road(Fig. 10). Also, the force of the ¥vater cascading do¥vnfrom the mountain added to this failure. Efforts to reachthe to¥vns by rescue organizations ¥vere further hampered¥vhen typhoon Yoyon*・ hit the region three days afterstorm Winnie. Although Winnie ¥vas .just a tropical stormand not a strong typhoon, the destructive impact ¥vasmagnified by the amount of rain that fell over the area.Most of the victims of the landslides in Real ¥vere lo¥v-income peoplevho ¥vere obliged to li¥'e on highlyhaz,ardous land, typically along the road through verysteeply sloping areas of the mountain fronts that rise SEDli¥,IEINT DISASTERS IN THE PHILIPPINESFrg 9. Real Bridge washed out by t le flash floods705Frg ll. Vielv of the heavi y-silted Agos River flowing from SierraMadre mountain tolvards Infanta and Geueral Nakarfrom surrounding areas evacuated to the second fioor of aschool building. By 0300H the next day, the ¥vater levelbegan to drop.Some portions of Infanta, especially those near theriver, ¥vere covered by silt-laden fiood¥vater, necessitatin*'the evacuation of the affected area. Mud, Iogs and debrisfilled houses, roads and farm lands. Fi**ure 13 shows thecondition of the houses in the area. Several of the housesdestroyed 1'ere located along the banks of major waterways .Although residents of these t¥vo towns were ac-Fig. 10. Slope failure along Tanauan Road in Real 'Town: Drainageculverts are exposed at the collapsed section of t re roadcustomed to typhoons and have experienced floods in thepast (especially in the 1980s), the scale and resultin*"impact of the 2004 disaster ¥vas the vorst in memory.lvlonths after the disaster, some remote villages inGeneral Nakar ¥vere still co¥'ered by mud rendering ¥'astdirectly from the sea. Fortunately, the density of thetracts of farmlands useless. The volume of rnud andhouses in Real ¥ 'as lo¥v, and therefore, the probability ofdebris that covered the two to¥vns ¥vas estimated at nearly20 million m3.these structures being struck by falling earth blocks or'landslides was relatively low. Nevertheless, darnage to lifeand property ¥vas still extensive.Dalnage in lilfallta and Genera! Nakar TownsDue to the heavy rainfall, slope failures occurred onthe flanks of the mountain ranges. Debris flolvs were thenmobilized at different locations in the rnountain, and asthey continued flo¥ving down hills and through channeis,One of the spans of the brid*・e connectin*' Infanta andGener'al Nakar' collapsed due to the impact of thefloodwater' and driftwood (Fig. 14). The channel of AgosRiver is shaped like "L" near the location of the bridge,cr'eating an obstacle to floods. The rushing fioodwater'fr'om the mountains, carrying many logs and debris, ¥vasforced to make a turn near the bridge, as evidenced bythe scars on the slopes near the north abutment. Thet.hey converged in Agos River (Fig. 11). The volume ofabundant logs caused logjams along the course of river,debr'is flo¥vs increased ¥vith the addition of water, sand,especially where the debris abutted against the piers of thetrees and other' materials scraped along the ¥vay. Thesebridge. As a result, the on-rushin*' water lvas forced toseek other possible routes, causing the inundation of thearea south of the bridge.quickly inundated and overtopped the banks of AgosRi¥'er. As the debris flo¥vs reached the flatter grounds ofGeneral Nakar and Infanta, the materials spread o¥'er thealluvial f'an, burying the t¥vo lo¥v-lying to¥1'ns. Theestimated area covered by flo¥vs is about 40 kml (Fig. 12).The alluvial fan advanced ofiLshore by about 50 m due tothis debris flo¥v.Inter¥'ie¥vs ¥vith local r'esidents indicate that rain fell theentire day of' 29 November as storm Winnie passed. Withno ¥varning, flood¥vater began to rise, and by 2100H, itGEOTECHNICAL CHARACTERISTICS OF FAILEDMASSESDuring the ¥'isit, soil samples ¥ 'ere obtained at threelocations in or'der to investigate the geotechnical featuresof the surface soils, an important factor in analyzing thehad risen to about the frst fioor level (roughly 3 m). Themechanism of slope failure and transport pr'operties ofthe debris fio¥v. One surface soil sample was obtainedflashfioods carried huge drif'ted logs and mud. Peoplefrom the slope adjacent to the one that failed in Real (see ORENSE ET AL.706{) ' : {"__i,_t '-:;' f ;; -'}!1' '4'; s i - '-:'" ' i''S- ':j' ; r L ; ' _ '' 8s' - 1N'i:'Li:; - "'"i * ;''"l- '* s"'!'P_'$-!#;;! ;'i i :i' ? r #;r '#ts?';'';" -;;;'t ;;:- '.**.., iLEGEN0=;'' { i -{!'fi' i( :-i;': '; ' # ' :i _- I s ' _ _ _i-; ' 1;--,:':i;' -' ; ; " l! ' :f -*' '_ }{ "I'_ r#'__ ' _ii :i.;:-_ " ;i; :!i ":.; i"i::'}"'jii';i:;'-:;j; :i "' r :i"Lt;! :f- j : . _! i1;"j;t'e:1 "!' fP r_ ! r' ;-' ; ;'s/ -- -i;; ;i 'i;i;i:::;:::;ii:::i: :'ii:ii- !Is:Sa ":ee' f;e 'ar'f i ;Uneemenf(Aaja Pho(o interpf taj l)/¥S}1 oeeft5flew- +7f_ ' '"'::';'JF{ s p #! f ! : _h"i ;;;t:"De is-i/ /' ' _ _' /f ';1:':_ ': ;;?};';b/'e ' -;'1'1/ Ii?:i 's4} -?t':- : T'ii,= : : Sss'}!i '?ss{'r'¥:: iei '-j ' :-;{; " 7it;j ;;'' ' F1'':"i I sd i- '-.'{' ';i;.¥'"T $';h- "'1'; faota. t"il' -"'t; .:; ' :' ¥'- i.:_"i--;f-'!"' '1' t:'l; _ '!'T i st; '¥;'¥;I ; ''- "!-^!-. i'':1t¥";'"Y''; ;?i' / :{"I Y/tt "I"*i:;¥' '.i i"¥J; i';5-;;t ' w'r " "' t;; :' ' ' t t #'1 :'r J dl¥'i 'l IiS ' ; :;-':" i,' , ,.?L,;; 4 i' ' 's'! !;{ ' ' - '' ;:!;t_ Y^l oif!5 ; :sai;;2;i i':';;; :iSf:: :: !;^::: ) i : :;11 ;' ' d' ' '- j¥-!'= A ts'__Rivet'il'ss,i:4 ' r '_;" '*'s* *";j!/: " ;" "; ; ; :!; ; : ;4'" S' ' 'i' 1?:"!:"i 1e ::' ';;T::}* ! ^**:s+*s'#"!S-'";$ :; -$+'"/+' *"'ii""'-;'Si' ";i*/ +i*';j?' ';t * :s_ { - ";(:i:':-i ;;1* { " "/ "#S "' #!'; i'rf!'*';";*(!;;; ;:f:;' i ;;;{'s SF;'t;i ;:;/;;;_;::'i!c;;:::;'!_ _' )'_ "__ '" ' =s;;*:"' ;_S ;_ '"_;'-' - -/ ";r': t:''?; ;'': -f ' l;ii "'-" 1# I '; s'';{ ' ;i'-f;;it 'sst;-';*' #- :';' * * *: .;* .;#='')* :t '*- .*' '- =* * *: *' S' '-''SS'' {'w;' "':';-- *-" * * *;*:i*''.i ;. S" i{*t;.j:::-"- -:;**/:;; (__ _ 'S#i '{{:i' ""** "***: ;: ;t{!:'; "'";(;i '+"'fii;I (':;;:1:i:;i: ;/?*"!/{^ i/j"'.'.-*....; !' '!' ' :!';# " '' ';;' ; li;!f' 1!'.'!"F'i j-":;;; ;i:;'!::{'i;i ;;'* *'lj'; 'i ;' s*''r//::::i{;*'; *' : ' ;ti "{i: i::i::1::i: ***S'? # ;''t:_'t#{_'l' _; " :* _/'' + ;s'_ '_ii# { { 'I; :}j SS#.;' t:;::;:;:;f!(;:.jSWi;'i'js i**:;:::{:::{:i:{::}:{:;:;'j ! '+;1; : .'.; SIs: ;: ;t "t i ;:;";;;/ti ; ;'t* *"'- ;t;' '" !i;i;:;i;i;:;:!:;;f't;:;;:;;''t'; ' -*'- ! ';;'+$#; : ii;' '- '' '-*f"';'_i+ts*'i*_i''-;s.#' $:''¥ - . ' - '*- ; ''s' * - Tts_'"*ss' " 's+ i/'_''" - ;^_: :SSS_* " ';'s;s* "+' - * - i#"/I _.;:1- ' """ '* i! ' ;"'*'!' ' ^ ; "=; :f(;;{: 'if # ''##;';:*;' ''* t# l -'-f ' ¥+L #- '* s's" ;:_- *-+-'' i' i -''#- = s¥;I'; =Si '_i2s=;/' '- ''s;'i -t-rF'_:ITT":ii ; .':*,t. _ . 1'=?:: ':1l' ':1l''T'; "*sf';'_?':s{ 'r/'}h t f:. ':¥J s ' ' :s ' - - _f r ' i: ']'! i !i ': ! ;1'- f ' ;;' t**' "'tll "'l':;* -i " _;} ' z^*' ''"s_- ' :*:'1 : ;'It! : i i'' i;1';_';i ; : j i ':s;!:::;;"'^ti ' __i tii ';;i :;rlii ,+v tiirect ts/. .r r?1'ssa:'#: s; "' " 1 -I 'j' i' s'-' ;F f'-; 1;''1i :'iT-s#I;='_ ;:i'-i:irf-i7_ ii?i)':'-s "!# -' "- : ' ;{ 1iii1:1T;r'' "'" -_ -_f'F ts *'- 1i !ts! l;'_1;"r' '1:x"+x S"F 'A'';d""' 'tiS ' ' ;';T : t ;t' __"- ?"'-'- --"' ;'P-'-F' '-' "*'sls' r'i :;ir?l-']';i +"''e d^'x ; " _s;s"F 1r ; 1 ';' ' : ';/ ;"'sf:'i ;l:- -"4iLaSd$! S/;i I; -rf"itrI: "t'::;:::::: :i 't" 'j i"; s';j!f!Sf lhtt; -_'_- : : i ' : i :iS ;:! ;lrJ¥iie lej:e ::'/; js;S:j.;j ''t f-Fig. 12. Areas affected bl.' debris flol" i 1 Infanta and Gen. Nak tr, Quezon Province_:]l ;}:" ;.;・・・・・-i*; ;a *,f.:' s:s;i."'., "Fig. 14. One of the spans of the bridge connecting Infamta a ld Nakarlvas lvashed*out by debris-laden fioodlvaterFig. 13. Mud, Iogs and debris covering the residential a ea of Infanta:Note the difference betlveen the traces of mud in the houses and theorigil]a leve] of riverFig. 7). A second sample lvas collected upstream of theAgos Ri¥'er, about I km to the ¥vest of the collapsedbridge linking Infanta and General Nakar. Finally, thethird vas sampled do¥vnstream of the allu¥'ial fan nearthe eastern coast of General Nakar.The physical properties of the samples are shown inTable I ¥'hile the grain siz,e distribution curves areillustrated in Fig. 15. The sample obtained in Real is lightbro¥ 'n silty clay, consistent ¥vith the geological description of mountain soil. Due to the small eff ctive siz,e ofthe particles, it is presumed that the coefiicient ofpermeability is very lo¥v. Considerin_g: the intensity andduration of rainfall preceding the failure and the poor SEDllvIFNT DISASTERS IN THE PHILIPPINESrable 1.Ph .・sical properties of soil samples taken from disastcr sitesPro pertySpecific gravity, G*LandslideDebris fio¥vDebris fl0 1(Real)(Upstream)(Dowr stream)2.422 702 50693Clay content, C* (Of6)llo8.6O. 157lvlean grain size, D*o (mm)O.O 1 52.634Effective size, Dlo (mm)o 004o.222Uniformity coeficient, C5.6617.60Cur¥'ature coef ciem, C*o 951 .765.1717.61NPNP- Landslide (Real)- Debris Flow (Upstre8m)- Debris Flow (Downstream)1 OoCiay SiltSandT¥vo conditions must be met for rainfall to cause a60debris fio v: the hillslope must be susceptible to failurea)c,)tQunder saturated conditions, and the rainfall must ha¥'e40the intensity and duration necessary to saturate theQ'c:a)larger grains compose most of mass of debris fiows andthat silt and clay size grains commonly constitute lessthan 200/0 of the mass (Daido, 1971; Takahashi, 1991).CAUSES OF SEDIMENT DISASTERS80a)c:so 00823 . 87Gralei:>223Fines coment, F* ( /o)Plas icity index, Ip:,707ground to a sufficient depth. Hillslope steepness has the20e,Q'oO OOI O 1 10 iOOO Oi1Particie Size (mm)most influence on debris-flolv likelihood, followed bythickness and texture of the loose material that mantlesthe slope, and the mechanical and hydrological propertiesof the underlying rock. Moreover, it is clear thatdraina*"e condition, it is possible that the surface soilbecame fully saturated during the passa*'e of the tropical¥vhatever the mechanism, the production of' debris flowsrequires intense rainfall, sustained for at least a briefperiod of' time. These conditions were satisfied by theslopes in Quezon Province during the passage of the 2004tropical cyclones.In the aftermath of the disaster, go¥'ernment officialscyclones and pore water pressure built-up, resulting in¥vere quick in blaming logging activities in thefailure.C*omparison of grain size dist.ribution between samplessoutheastern portion of Sierra Madre mountain ranges asthe main cause of landslides and floods in the affectedobtained upstrearn and do¥vnstream of the debris flo¥vareas. Holvever, aerial photos sho¥ved that in thearea shows that the upstream deposits are coarser,consisting generally of gravel and sand. Do¥vnstreamsamples, on the other hand, are silty sands. Thispreferential grading is consistent with the mode ofsouth¥vestern portion of the mountain ranges, ¥vhich havealrnost similar geolo*'ic and topographic features, veryfew landslides occurred although forest cover ¥vas verylittle (Foronda, 2005). Further inspection reveals thatlandslides ¥vere prevalent on steep slopes, whether forested or not (Fig. 16). This observation confirms the notionthat the steeper the slope, the higher' the probability ofFig. 15. Grain size distribution curves of soil samples obtained at thesite of disasterdeposition and the transport distance. Note that the *"rainsizes pr'esented herein do not include lar*"er-sized grains,such as cobbles and boulders, because they are difficult tosample. Nevertheless, it ¥vas clear that heavier particles,such as gravel and sand, tend to settle earlier' due togra¥'ity ¥vhile li**hter materials, such as silt and clay, arecarried further from the source by the flo¥vin_ : water.It is worthy to mention that the particle size distributions of the soil samples taken from the debris fio¥v sitesare consistent with the range of sizes observed in previousdebris flo¥vs which occurred in different parts of the world(e.g., Hutchinson, 1988). Although the presence ofcobbles and boulders, Ivhich are difficult to sample, arei**nored, it is nevertheless clear that sand, **ravel andfailure. The results of preliminar'y investigation stron*"Iysuggest that hea¥'y rainfall preceding the disaster is theimmediate cause of the slides. The torrential do¥ 'npoursaturated the slopes, Ieading to loss in shear strength andresulted in failures.Note that before rainfall can increase pore ¥vaterpressure, the slope materials must already containenough soil rnoisture to neutralize the soil suction(negative pore-¥vater pressure) in dry soils. Analysis ofhistorical data by usin*' the relationship between rainfallintensity/duration and debris flow initiation (Guadaqno, ORENSE ET AL.70s'l10m;ft sndin9 surf:ace:IT-11 :" 1'h suffacet tl-":;:::::1:¥ ;; 1 - soil; :iom BedrockFig. 17. Finite element modelapsed slope in ReaiToseE: : the_l'used in the seepage amat,.・sis of the col-slopes in Real to¥vn. Due to space limitation, details ofthe study are not presented herein and readers areFig. 16. Aerial vie v of Sierra Madre mountain detineating locationsof slope fajlure1991) has indeed stressed the importance of continuousantecedent rainfall. Therefore, the preceding precipitation in the region caused by typhoon Unding and tropicalstorm Violeta are important contributors to the saturation of the slopes.referred to the ¥vorks by Matsuoka (2006) for' further in-formation. Brief description of the methodolo*"y forseepage analysis is given in the Appendix.The slope considered in the analysis is the one sho¥vn inFig. 7, where more than 100 people ¥vere killed by thesliding mass. The average slope lvas about 30 degrees, thedepth of the failed soil was about 4-7 m, and the surfacesoil ¥vas silty clay loam, as indicated in Fig. 15. TheFurthermore, numerous faults and fractures cutexposed surface of the failed slope was sandy layerthrough the major rock units in the south¥vestern sectionof Sierra Madre mountain. Because of these, the rocks incomposed of weathered residual soil. This indicates thatsliding occurred at the boundary between the surface soilthe areas are badly fractured or fragmented, makingand bedrock.them vulnerable to ¥veathering and erosion. During heavyrainfall preceding the mass movement, ¥vater seeped intoThe finite element model employed is sho¥vn in Fig. 17.The modeled slope is *"eometricaily a planar slope withsoil layers almost parallel to the *'round surface. Thethe fractures of the fractured rocks and promoted thedeterioration of the rock mass. If the amount of rainfallis of limited volume and time duration, the soil and rockmaterials could have easily absorbed the ¥vater. Ho¥vever,due to the excessive rainfall prior to the disaster, rock andsoil materials in the area became saturated ¥vith ¥vater,and this caused the failure of slopes.slope profile, as ¥vell as the location of the sliding surface,¥vas estimated based on the results of field investigation.The pre-failure profile ¥vas estimated from the aver'ageoutline outside the landslide body, ¥vhile the location ofbedrock ¥vas inferred based on typical depth of surfacesoil in the area.The unsaturated/saturat,ed soil parameters for theIt should be mentioned that the sediment disastersdescribed above ha¥'e similar characteristics as thosesurface layer were estimated from the grain-size distribu-experienced recently in other parts of the ¥vorld, such astion curve sho¥vn in Fig. 15. The model developed byin Kagoshima (1997) and Hiroshima (1999), Japan, andin Sumatra ('_OO1) and East Java (2002), Indonesia(ML,IT, '_004). What makes the Quez,on sedimentArya and Paris (1981) ¥vas used to approximate the soilwater characteristic curve ¥vhile the relation bet¥veencoefficient of permeability and matric suction ¥vas esti-disasters more remarkable, ho¥vever, is the fact that theymated based on the method proposed by Fredlund andoccurred in a mountainous area ¥vhich re*"ularly ex-4016mm annually). The series of tropical cyclonesRahardjo (1993). These unsaturated soil properties areshown in Fig. 18. Other required soll parameters ¥vereestimated using values obtained from parametric studiesdrenched the area ¥vith higher-than-average intensity(possibly a 100 year rainfall), causing catastrophicand from published data for soils ¥vith similar grain siz,edistribution cur¥'e, and these are indicated in Fig. 17. Indamage to life and property. The debris flo¥v in Infantathe table, k* and k, refer to the saturated permeabilityand Gen. Nakar covered an area of approximately 40coefficient in the lateral and vertical directions, respec-km2, ¥vith estimated ¥'olume of sediments up to '_O millionm3, almost er'asing these t¥vo to¥vns from the map.tively, V** is the criticai flo¥v velocity for seepage failure,periences very high rainfall intensity. (on the average,and S* is the specific storage coefficient.The slope surface ¥vas assumed as permeable boundarySIMULATION OF Sl,OPE FAILURF,In order to investi_ :ate the mechanism of the slope¥vhere rainfall ¥vas applied, ¥vhile all other boundaries,including the bedrock, were assumed to be impermeable.Since the initial distribution of pressure head ¥vithin thefailur'es and the role of rainfall intensity in inducing suchmodel is unkno ¥'n, preliminary analysis ¥vas conductedfailures, a finite element-based saturated/unsaturatedseepage analysis and limit equilibrium-based slope stabil-by subjecting the model to an antecedent rainfall consistin>' of a cycle of one day of rain and t¥vo days of no-rainity calculations ¥vere conducted on one of the failedfor a period of one year, such that the total intensity SEDlivIENT DISASTERS IN THE PHILIPPINES;60c:eoe'and through parametric studies, some input parameters¥vere modified to replicate the most critical settings forslope stability (Matsuoka, 2006). The results of seepageanalysis are sho¥vn in Fig. 19. At the initial condition,i.e., after the model **round ¥vas subjected to antecedentc:Oo40g(T'30e)20:rainfall for a period of I yr, it can be seen that most of the,:'E:: 10o o>1ioi OO1 OOoiooooSucton (kPa)2>!),e,10oq)e,:,o*oc:cDVEoc,70908Eo 06x 04;slope, except for the upper portion, is fully saturated.This condition is possible, considering that Infanta issubjected to a very high rainfall all throughout the yearand that Real to¥vn is very near the sea. As the model issubjected to successive hea¥'y rainfalls due to the passageof' the tropicai cyclones, the saturated zone propagatedtowards the upper portion. Such finding is corroboratedby the observation rnade by the residents in the area thatduring the hea¥'y rainf'all, rainlvater ¥vas cascadingdo¥ 'mvard f'rorn the slope tolvards the sea. Finally, at the02time of occurrence of the landslide on 29 No¥'ember, theoo1101 ooo1 OO1 oOoOSuction (kPaj¥vhole slope ¥vas already fully satur'ated.Using the results of seepage analysis, slope stabilityanalyses ¥vere conducted to simulate the sliding surfaceFig. 18. Unsaturated soil properties used in the anah.'sis: (a) soil waterand to back-calculate the str'ength parameters of thecl]aracteristic curve and (b) unsaturated permeabilit) funcrionsurface soil. For this purpose, slip circle method based onmodified Fellenius approach ¥vas employed to calculatethe factor of safety (F*) at the various time stages sho¥¥'nDegreeof S}*turaiionin Fig. 19. The location of thei009 .vater table ¥vithin theprofile at each stage ¥vas estimated based on the distribu-E ] 9O%] 809{.[] 709(.tion of degree of saturation lvithin the soil profile, as ¥vellI eo%b 05 November 2004as on the calculated distribution of total head at eachnodal point.Since obser¥'ation dictates that the slip surf'ace oc-curred at the boundary of the surface soil and thebedrock, Iarge strength parameters ¥vere assigned to thebedrock vhile those at the sm'face soil ¥vere assigned bytrial and er'ror. lvloreo¥'er, considerin*" that seepageanalyses revealed that portions of the slope along the(c) 21November 2004circular failure planes ¥vere either fully or' alrnost saturat-ed, shear strength parameters at full saturationvereused. Furthermore, the effect of seepage pressure on stability ¥vas neglected. The unit lveight of' the surface soillvas assumed as y= 17 kN/m3.To back-calculate the required stren_g:th parameters,the condition ¥vhen the actual slope failure occurred vassimulated, setting F* = I .O. Typical combinations of cohesion and friction an*'1e ¥vhich ¥vould result in slip surfacesimilar to that actually observed and at vhich F, = I .O aresho¥¥'n in Fig. 20. It can be observed that the relation isFig, 19. Results of saturated/unsaturatcd seepage anal)sis at differenlstages corresponding to the passage of various tropical c .'ctoneslinear, i.e. , the required cohesion incr'eases as the frictionang:le of' the soil decreases.Once the strength parameters were calculated, these¥vould sum up to 4016 rnm, which is equal to the averageannual rainfall in Infanta. In the calculation, a transientflux function, ¥vith values equal to the r'ainfall rate of the¥vere employed to calculate the factor of safety for otherstages prior to the collapse of the slope. Figure 21 showst.he relation bet¥veen F= and rainfall intensity. Note thatantecedent event lvas applied to the nodes along theground surface.Once the initial condition vas kno¥vn, the model wasnext subjected to the recorded daily rainfall intensitypr'ior to the series of typhoons (i.e., the initial condition),shown in Fig. 2. By considering the actual time history ofrainfall, the complexlty of the e¥'ent can be easily replicated. Different scenarios were considered in the analysis,typhoon Winnie (29 November), the ¥vhole slope becamefully saturated and F*= 1.0. Then, subsequent rainfallmade the slope unstable and slope failure occurred.the slope is stable ¥vith F*= i.53; ho¥vever, as typhoonsand tropical storms passed, the slope became saturatedand the f'actors of safety decrease. Finally, at the peak of ORENSE ET AL^710sho¥ved that at the time of failure of the slope in Real, the12For Fs=1 OO at critical slip surfaceE 10¥Z:o 8¥vhole slope is fully satur'ated and therefore unstableBased on these results, strength parameters of the surfacesoil ¥vere back-calculated using slip circle method.The slope failures and related mass movements re-s:O 6sulted in significant localized changes in the physicallandscape of the affected areas. Therefor'e, appropriatepreventive measures are necessary in order to minimiz,e4further landslides and to prevent similar disasters in the(,,s'Lo5 252030 35 40Friction angie, c (deg)future. The back-calcuiated strength parameters can beused to analyze the stability of other slopes in the area¥vith the same soil types, and to formulate a landslide riskFig. 20. Relation bet veen cohesion and friction angle correspoudingto F,= 1.0map due to rainfall for the region.ACKNOWLF.DGMENTSIn carrying out the ground reconnaissance ¥vork in the400landslide-affected areas, the overall cooperation ofMayor F. America of Infanta, Quezon, Prof. J. Foronda3eoof the University of the Philippines, Mr. N Resplandorof Infanta Congressman's Of ice, and Engr. E. Morales¥vas very helpful. The rainfail data for' the month of; 200G(o loOO1,1Q1 Q/30 1 1i ,9!4 1li/141 i24I /29 1 2!SFig. 21. Time histories of rainfall intensity and factor of safet)No¥'ember ¥vas provided by the PAGASA office, QuezonCity. The staff of DENR Mines and Cjeosciences Bureausupported the third author during the site inspectionimmediately after the disaster. Seepage analyses ¥vereconducted using the program developed by ChuoKaihatsu Corporation, ¥vhile slope stability analyses ¥vereThe results indicate that the simulation of saturated/unsaturated fio¥v ¥vithin the soil ¥vith a finite elementseepage analysis can provide useful insight for the reconstruction of pore water pressure variations, total headand degree of saturation (or volumetric water content) inresponse to variable intensity rainfall. The assessment ofpore pressure changes at all nodes, exemplified by theperformed ¥vith the program de¥'eloped by Kiso-jibanground ¥ ater smface, also allo¥vs an accurate slope1) Arya, L. *¥,1. and Paris, J. F. (1981): A physicoempirical model opredicl the soil moisture characteristic from panicle-size distribution and bulk densi y data, Soi! Sci. Soc. An7. J., 45, 1023l030stability analysis to be performed.Due to the absence of available data for the strengthparameters for this type of soil, the back-calculatedConsultants Co., Lrd. The authors ¥vish to express theirdeep gratitude for the assistance and camaraderie of thesepersons and organizations.RF.FF.RENC_ES2) Asian Disaster Reduction Center (2004): Phi!ippines, 11130rl*phoon, http: //1v¥vw,adrc.or.jp.parameters can be used in the analysis of stability of other3) Chuo Kaihatsu Corporation ('-003): UNSEEP (Simil!a!ion Progr 7mslopes in the area with similar soil condition for futurerainfall. Thus, the results presented herein can serve asbasis for more detailed analysis of slope stability in thearea as well as in mapping the landslide risk due to heavyfor Sleac!.v and Uns!eady Seepage A na!_vsis Prob!em) User 's IVlanua!4) Daido, A. (1971): On he occurrence of mud-debrisrainfall.5) Foronda. J. ('_005): Personai Commuoication.6) Frediund, D. G. and Rahardjo, H. (1993): Soi! Mecllanics forCONC_LUDING REMARKS7) Guadaqno, F.(in Japanese).o¥v, Bu!letinof Disaster Preve,1!ion Resear'ch hzstitute, K¥.'olo University. Part2, 21(87), 109-135.Unsattlra!ec! Soils= Johr ¥¥*{lev and Sons Inc.., Ne v York ,Heavy rainfall due to a series of typhoons and tropicalstorms in the Philippines caused extensive damage to lifeand property, particularly in the coastal towns of Quez,onProvince. Analysis of rainfall data recorded immediatelyprior to the disaster, as ¥vell as comparison ¥vith historicaldata a¥'ailable, showed that the November-December2004 rainfall event ¥vas extraordinary. Other factors,such as the steepness of the slopes and presence of faultsand geologic structur'es, ¥vere considered as contributingfactors.The results of saturated/unsaturated seepage analyses,1. (1991): Debris flo¥v in the Campanian ¥'ol-caniclastic soils, (ed), S!ope Stabi!it_v.' Engineering. Deve!op,1lentsand App!ications (ed, b_v Chandler. RJ.), Thomas Telford,125-130.8) Hutchinson, J_ N. (1988): General repor : ,Iorphologicai andgeotechnical parameters of landslides in relation o geology andhydrogeology, Proc5th InrS vmpLancl,s!ides. Lausanne, Slvitzer-land, l, 3-35_9) l¥,iatsuoka, K. (2006): In¥'estigation of the 7-004 sediment disastercaused by heavy rainfail in the Philippines using saturated-unsaturated seepage analysis, Bache!or Tllesis, Yamaguchi Uni¥'ersi y, 30({n Japanese).lO) ¥* ,iinistry of Land, Infrastructure and Transport (2004): Guicle!inesfor Deve!opnlent o.f I・ larni,1g (7nd Ev(lcu(7rion S!stenl agains!Sec!inlent Disas!ers in Developing Cotnltries, I 02. SFDli¥*IFNT DISAS'TERS IN THEl l) National Disaster Coordinaring Council ('_004): C'onlprehensiveReport on rhe After-Effects of Fbur Typhoorls. Damage Assess-nlent and Re!ieflRecover_,' Operations Undertaken by !¥rDC*CJVlember Agencies, h tp: //¥v¥vlv,ndcc.gov.ph.PHILIPPINESa is equal to either O (in unsaturated r'egion) or I (in satu-rated region).The initial condition is given by:h(x, z, O) = ho(x, z) (A3)12) National ¥Vater Inf'ormation Net¥vork (2002): Rainfa!! Synoptic,Annt/a!Vifax!mum J-da_v Rain_fa!l for Infanta. Que on,htt p : // v v¥v . n¥vi nn¥vrb . gov . p h .13) O ce for Coordination of Humanitarian Affairs ('-004): Phi!ippineF!oocls, OCHA Siruation Report N0. 4. http: //¥v¥v v.relief¥veb.int.14) PAGASA: http://¥v¥v¥v.pagasa.dost.gov.ph_15) Takahashi, T. (1991): Debris F!o,t', A.A. Balkema, 165.16) Tropical RainfaH lvlonitoring h,Iission (2004): Plli!ippines Inun-711¥vhile the boundary conditions are as follows:(1) at boundar'y lvhere lvater head is kno¥¥'n:h(x, z, t) =h (x, z, t) (A4a)(2) at boundary ¥vhere the seepage flux is known.dated b_v S2lccessive Stonns, ht p: //trmm.gsfc,nasa.gov l.= - vb(x, z, t) (A4b)APPENDIX¥vhere ll. is the initial water head, hb is the boundaryvaterThe unsaturated/saturated seepage analysis is based onthe follo¥vin : governing differential equation, based onthe total head, h:to the boundary and vb is the seepage velocity at thehead, v is the seepage veiocity at any time t perpendicularah atk allahaxfiax+ kaz azaa(A I )¥vhereh= +zft = c( ) + aS=(A2)In the above equations, is the ¥vater pressur'e head, z isthe elevation head, S* is the specific storage coefficient,c is the slope of the soil water characteristic curve (= ahla ), k is the coefficient of permeability and the pararneterboundary.The governing differential equation is sol¥'ed numerically using finite element formulation (CKC, '_003). Inaddition to conventional steady-state saturated fio¥vanalysis, its comprehensive formulation allo¥vs modelingof saturated/unsaturated probiems, making it possible toinvestigate seepage as a function of time. The transientfeature allo¥vs one to analyze such problems as themigration of a ¥vetting front and the dissipation of excesspore- vater pressure. Thus, various processes, such asrain¥vater infiltration, *"round¥vater lo¥vering, boiling andpiping, as ¥vell as investigating the eff ctiveness of' drainpipes, can be easily considered.
  • ログイン
  • タイトル
  • Evaluation of Run-out Distances of Slope Failures during 2004 Niigata-Ken Chuetsu Earthquake
  • 著者
  • Yoshimichi Tsukamoto・Kenji Ishihara・Yasuhide Kobari
  • 出版
  • soils and Foundations
  • ページ
  • 713〜725
  • 発行
  • 2006/12/15
  • 文書ID
  • 20957
  • 内容
  • SOILS AND FOUNDATIONSVol 46, NO- 6, 713-725, Dec 2006Japanese Geotechnlcal Soclet¥'EVALUATION OF RUN-OUT DISTANCES OF SLOPE FAILURES DURING2004 NIIGATA-KEN CHUETSU EARTHQUAKEYOSHIN,lICHI TSUKAMOTOi), KENJI ISHIHARAii) and YASUHIDE KoBARliii)ABSTRACTDuring) 2004 Niigata-ken Chuetsu Earthquake in Japan, a large number of landslides occurred on natural slopes,especially at the hillsides in the region of Yamakoshi. In many of the large slides, the debris has travelled through afairly long distance, aggravating the disaster caused by the landslides. In recognizing its importance, case studies ¥vereundertaken on the run-out distance of the landslides at several sites at Higashi-Takeza¥va, Mushigame and Naraki.Case studies ar'e also under'taken for' slope failur'es invol¥'Ing the man-made deposits behind retaining ¥valls surrounding the residential hill at Takamachi-Danchi in Nagaoka. In the first section of' the p 'esent study, a simple analyticalmethod is introduced based on the energy principle, in vhich the residual strength is taken up as a sole parameter todetermine t.he run-out distance combined ¥vith the geornetr'y of the landslides. The slope failure is herein assumed toconsist of two phases, sliding and spreading, and the sliding distance is defined as the length of a slope on which themass of soils slides down, and the run-out distance is determined as the one on a gentle slope or flat plane on ¥ 'hich thephase of spreading occurs. Soil samples were retrieved fr'om the sites of landslide, and laboratory triaxial tests areconducted on unsaturated soil samples with varying water contents. The residual shear strength thus obtained was usedas an input parameter in the simple analysis to forecast the run-out distance. The outcome of the present study ispresented in a form of simple charts in vhich the run-out distance is expressed as function of relevant geometricalparameters and the residual shear strength of soils invol¥'ed in the landslide.Key w"ords: landslide, residual strength, run-out distance (IGC:D6/E6)ed sources of landslides, the quality of landslide hazardmaps ¥¥'ould be improved. Therefore, from the vie¥vpointof landslide hazard mitigation, the estimate of run-outdistance would be among the greatest concerns. The rssueof the post-failure run-out distance has been the subjectINTRODUCTIONOn October 2,3, 2004, a devastating earthquake measuring a magnitude of M=6.8 struck the hilly regions insouth Niigata at a local time of 17:56, followed by t¥vohuge aftershocks measurmg M= 6.0 and 6.5 at 18: I i and18:34 on the same day. Due partly to the heavy rainfallof concern, and numerous attempts ha¥'e been made toclarif'y the mechanism of debris fiow based on case studiesand analyses. C*omprehensive overvie¥v and summary onthe outcomes of these studies are given by Hunter andthree days preceding the earthquake, the hilly region vasse¥'erely plagued during the earthquakes by the occur'rence of numerous landslides, especially in the mountainarea of' Yamakoshi, where tertiary deposits mainly pre-Fell (2003). In the present study, case studies are undertaken focusing on the run-out distance of the landslides,vail. Fi*aure I sho¥vs the epicentres of the main shock andwhich occurred during the 2004 Niigata-ken Chuetsuthe subsequent two aftershocks and also the locationsEarthquake and also during other past earthquakes.¥vhere large landslides occurred. The reverse fault planehidden locally in the direction of north east-south westare purported to have been responsible for triggering thesequence of these earthquakes, and the areas devastatedSITES OF LANDSLIDE INVESTIGATEDby numer'ous landslides vere located on the side ofslopes on both sides of the valley along Imo river. Thehanging- vall of the fault. The area of Yamakoshi formsa hilly region typically with 300 to 400 m high above thelocations of the landslides are sho¥vn in Fig. l. As illustrated schematically in Fi**. 2, the stratification of thesea level.sedimentary r'ocks composed of sandstone and mudstoneWhen it becomes possible to more accur'ately estimatethe run-out distance of collapsed soil debris from expect-is inclined about 15 to '-O degrees to¥vards the ¥vest forming the out-face dipping surface, ¥vhereas the western facei::iiiA number of landslides ¥vere triggered on naturalAssociate Professor, Departmem of Ci¥'il Elrgineering, 'Tokyo University of Science, Japan (ytsoilC・rs.noda, us.ac.jp).Professor, diuo.(,; raduate Student, dilloThe manuscript fbr this paper ¥vas received for revie¥v on l¥,・:ay 9, 2006; approved on September 26, 2006.¥Vritten discussions on this paper should be subrnitted before Jul_v I , 2007 to the Japanese Geotechnical Society, 4-38-2, Sengoku, Bunkyo-ku,Tokyo I 12-001 l. Japan. Upon request the closing date may be extended one momh_f 13 1TSUKAMOTO ET AL,_714fl l_andsiidesO EPicentreDD Sites aflands}ides:u iedTakttmachl-1rkmj LJL J_JJDanchieTochio¥Ta aOkaShin noRiverSumonee"amakoshi oe・iushi ameHi, eashi-Takez i YaeO j i } l:,,・ eeoe emo Riveree 'oejo '.Shi anoar lkii6e : e S:34eRivereecr AITHirokamiKawaguchiPhoto l. Bird*s e)e vielv of iaDdslide at Higashi-Takezawa (afterProfessor H. Marui)¥. ,i6 8o -e7:56. '3/0c ;2004_ ee+Horinouchi? ih'i6 o:* : ' ^+ * ;' /*** **"I'*""*+<[km]ll v ¥_f¥¥ ) _ Yamakoshif':;';i/ ; II;-;i -{'!{ SiiNa :aokaMushigamel'llr) r-)''//e/J* Naraki_1V- (.j e'e eoe.r¥,'e efrJss'*¥:/ec/ },¥ ¥Hi zashi-Takeza¥va¥J?oi -'sfl ie:1hnoeshPti Iti' ' //////eTldT ¥¥ /' {'1idin¥s";"' '"; ' :''* ;';-""_*; ' '*"' - "'s* "* *:*"" = *Photo 2. Ciose vielv of landslide at Hignshi-Takezawa (This landshdeoccurred on the gentle slope at the left bank of Imo river, whichforms an out*facing dip strncture)(1) O_ n the out-facing dipping slopes on the easternslopes, many patches of paddy fields and carp-raisin*' ponds had been cultivated, and the majority oflandslides occurred at depths of 5 to 10 metres alongthe interface bet¥veen the sand and clay layers,leaving clear-cut slickensides exposed on the slidin*'River-; "I'In'facing diP ovt-fa ; diP ;/ :it h ¥;(/' "'"I "" ';*Epicentres and iocations of sitcs of lands ide studiedrs - -'**'s' s' S ;'#""-"'o+-- oi eee;.rig. 1.)'¥,l/Imo Ri¥'erJ fSvnc line siruclure::- ; ;surface.(2) It is conceived that the slidin*' has occurred ¥vithinthe least cemented ¥vater-saturated sand depositsatop the mudstone. Because of the large depth and¥vider area of sliding, the volume of the landslides¥vas r'elatively large, invol¥'ing the soil mass of aboutFig. 2.Geo]ogical features associated witl] Iandsl des10,000 to 50,000 m3. On the contrary, the landslideson the in-face dipping surface on the ¥vest side wereshallo¥v in depth on the order of 3 to 5 m, and thesoil mass involved ¥vas about 1,000 to 10,000 m3.along the Imo river constitutes the in-face dipping slopes¥vhich are inclined steeper ¥vith the angle of 30 to 40de*'rees. As illustrated in the inset of Fig. 2, the sandstonelayer is less cemented because of the action of ¥veatherin_"*.induced in conjunction ¥vith readily seeping ¥vater ascompared to the portion of less permeable mudstone.Thus, the characteristic differences prevailed in the modeLands!ide at Higas/1i-TakezalvaThe Imo river' flo¥vs from the north to south directionin the area of Yamakoshi. The site of landslide atHigashi-Takeza¥va is located at the left bank of Imo river,as sholvn in Fig. I . The left bank of this river is kno¥vn toform geologically the dipping slope structur'es, ¥vhich areof slope failures, as follolvs, bet¥veen the landslides on theknown to be susceptible to slip failures. The stream ofeastern and ¥vestern slopes of the Imo r'i¥'er.this river was blocked at several locations by massive piles RUN-OU'T DISTANCE OF LANDSLIDE7154001. . A ,i/--_J__*/ // */ - ' *OO**_/* _200//*) *///11//// Ir I S*t,,- 160 . 'j60Jj-* ).-__l SO _* i +. *o- _?** *l 5A"(a) o'・: imo riYer200400 IiPhoto 3. Bird*s e .'e vielv of landshde at Milshigame (after C:. Asia AirSurve) )50 SpreadirzSlldin(m I Pre fallur : sLlrfacc AJA' Post-failure surf ce'r 1 t I :?:_---504O (b)Il ao 3 OO 400 5aOmI +#^;i ;!j:; ;i{!r' '/' '* S1"'!;f':;..;i:j;:i* .1 '***.{.!*** +t ' " 'SSS'i*_: !' ; :;* *:;"'"';'j (;f ;jplane was observed being exposed over the head¥vallportion of the landslide with an an*"le of 20 degrees. Thissmooth mudstone surface had an appearance resembling¥vhat is called the slickenside. The slidin_ : debris appearedto be composed predominantly of sand. It is most likelythat due to the heavy rainfall preceding the earthquake,the *・round¥vater had been accumulated above t.he lesspermeable surface atop the mudstone, and consequentlythe liquefaction-type failur'e might have occurred at someportions of the less cemented sand deposit, Ieading to theentire soil mass moving do¥vn. In the present study, theprocess of this kind of sliding is postulated to consist oft vo phases, that is, the sliding of soil mass on the slopeand the spreading of debris over the field do¥vnhill..**-;. = S.;ii"'+''''/ ; ' { ' #i- s' :? 'r"*"'*"+!" #;' s;'#'-"; #'fT ' i''r?";''"; slb-t *: Si '":"+ ''"- '! ""; "!'"**;ii '* ' #'- *t*";F;; {;;f #**i " * i' :;'*'!'*S ;*'r:: : i';;;;: Fii;;;::;s+' :!;/)1{ ':';7:#*'; ;'i$e S"1 ': ;^' I':?' '--'; :;'i'; :'' :' _ *s:-'+ ':*;+ *ii'' ^i' -+'- 1i7*;*_';'s^1'sT;* '-7S?--' # f;s;j + - ;-*s' : '*=:' ' -"'- ***' ;S " +: :i':"S'; SS *S*s ;,'*:'s*'ji" ;:i- :'_ ;"..? ' * ' #'; 'e ' '- ' :s*+ *" '*' * ;:" t *1 "i 'i'ri*'_ 'Si s!/'_'f's;{ '' ;';;:i;;;:'"/''1 -*;;1';1 Si '*""' -+ ' #reconnaissance traversing at the site, a clearly visible slip'i*f*'*i '**/*/f""* # +* i/:'i';;i: i;i** '/j'-S_ :i_- s;'-;j'#!'#*+- '" !'+*'s;!!' /+'* '*'{ ;*;* " ''*"*'*i* i /*+' * :**Higashi-Takezawa are shown in Frg 3 Dunng the_''-r""!*"'.;"-.,!' ""!!- "--s;'-:- ''i; ^:;;i:hy*""'tl)+ ;" ;ii;'-"#f:'{"'"'* /:i:*;# * '#i + {/"'"* ' ' ' ;_i ;$1 - ;{j# " 'i '* *Hi**ashi-Takeza¥va was one of them, formin_"*, the naturalreservoir behind the debr'is pile. Along the Imo river, thesoil deposit consists of veathered sandstone of tertiaryera, and this material collapsed and moved do¥vnwards tothe riverbed. Numerous landslides ¥vere generated in thisregion in such tertiary deposits of weathered sandstone.The bird's eye vielv and close view from the south of thelandslide at Higashi-Takeza¥va are sho¥vn in Photos I and?-, and the plan and side vie vs of' the site of landslide at' ' "*' *SSj = *_il: '1i:f:;* _*_ +_:_ i:;;d*'"S;j: ""'r!!'' ..." 'i*i="+#'!--s'--* "*--*;Fs';#t'-s" #";# T#i s'i's_'*l_- *'1s*'*i #s "s# S*-'***- ;#; -*' l'* '' "*'*s'*'".+x " 1:; t""';" ; ; /!: ; ;! ("i 't ' s' ;1l ; ' 1' ;*'"'of the debris induced by the landslides during earthquakes. The upstream riverbeds at such locations lvereflooded and the natural ¥vater reservoirs vere formed,¥vhich were found to be narro vly in danger of breachafter the earthquake. The investigated landslide at': . :; *;1'1'# ; l;;' ""'t !l#t":';; ;+;'(i'"___S ;s; ii;- '*: -ii'* s* ::'.' ' .** .i i s s't; **s-: (; #(j ;' '- t' '" ' '; ' i "-"' s" '{i i ;'OOFrg 3. Plan view and side view of landslide at Hignsl]i-Iakezawa1 ;i- i:::;""'* +:# "+S;;S 'S#i:'f ;f;: "!;^;;{si,:!!';j.; eji S 'ijiiii;;1;:'j" ,^ i,!%,;i':::!;ji;,i}.._.-{!;1 #"' '1, 'ii;f::;! : 1 ;{::'/j' "'<# I}:i ;;;:::i:';;::ii:;' }?i':!;_;i!i::j::;::::: 1"'lIox.:;;!!';i!:;;;:;r r ;:::://'1: ; { ':::;:::1I Possible slidiag surihcei;;i z "F . S:i; ;::;; ;;:::i ;:-' st __';rf; ;: _* ; ';i':; 'S S' * '!: #Photo 4. Ctosevie , of landshtle at Mus rignme (This landslideoccurred in the region where t h o lveatirered mudstone mainl)prevails)Landslic!e at Mushi*・anleThe location of the landslide at Mushigame is alsosho¥vn in Fig. 1. T'he tertiary deposits of ¥veatheredmudstone are distr'ibuted over the slope, on ¥vhich anumber of fish ponds and paddy fields ¥vere constructedin a form of multi-shelves. Figure 4 shows the plan andside vielvs of the site of landslide at Mushigame. Thedeep-seated slip failure vas found near the top part of thelandslide, follo¥ved by spreading of debris spillin*' overthe road downhill. From the inspection at the site, it isnot kno¥vn exactly ¥vhere it was originated from, thoughthere ¥vas a trace of a *'rey-coloured layer of non-cemented sandstone in the debris. Therefore, the heavy rainfallmight have seeped ¥vithin this relatively permeable layer.Then, the collapse might ha¥'e been tri**gered by theshaking during the earthquake. As sho¥vn later' in Fig. 12,the soil sample from lvlushigame can be disintegrated intoa*'gregates. Therefore, it is most likely that the fragmentation must have played an important role in the "spreading" phase. This landslide is also characterized by the t¥vophases of sliding and spreading. TSUKAh"IOTO ET AL.716S: TFees rJn do230)OOmS;n IYi h d: Areas coYered by de briSY de ris't; i! ; :/ {'!'-. ;'t'-1"'1' #!'/t*ma"40!2'i _:,_f;_;;_j i 1'* 1 {:' (if'!f " l:i:?srJ) * 6'";o -:o +400_:s : ¥!!'t' ¥J 'iootrees coYe,td* * .¥ +'r ROad;-"-;t)_-.; t, -J l "Lp: / r200i"r r: : ;/fT, ;.:[i,: Yf.ls// 's"--'/2 O'' ;___'i_t / J / !'E:Fll:f.t l'r ,'i" f*_s:'arl":1 ' l -: : '/r' ,.¥;* _':: _ej"'T';;/' 1.':_tL* _.1gl)'/'si'' :: 'oOad/*' ' fhr'+ :mo riYeFi.,**.**(a) ';'.-.' *--"' **S ---t /t-i'St: : $ir': - t' /;@" 20 ; !__i(a)600m400200ooo100mlOOi 50m(b) SPreei dingp'e-;l rsutme A1T *1 OO[ Pre-failure uri3ceIt40 I'ost-!post-r ] l :re 5urface;i re sL;t:att / -A* P0'"- i l*'50, ,.lidin ¥^s'*rl** .;sl Possibie slidirl surfacelOOoFig. 5.pJ(b)Plan vie v and sitle view of landslide at Mushi amelJImo riYerOO 100 200mFig. 4.lSpreading I Slid ri"200300400mPlan vielv antl side vie v of iandslidc at NarakiAs illustrated in Fig. '-, the right bank of this riverconsists of steeply dipping in-face slope structures. Surficial slips from the top of the cliff took place during theearthquake at several locations along the Imo river. Thelandslide at Naraki was one of such landslides. As sho¥vnin Fig. 5, the head¥vall of the landslide had a slope angleof over 35 degrees and spanned ¥videly, resulting in totalin the *'reat amount of debris cascading do¥¥'n the valley.F!ow Faih!res of Man-lnade Fi!!s at Residentia! Hi!l ofT(lka/ 1 1 acll i-DanchiPhoto 5. Bird's el.'eviel7 of landslide at Narnk(afterC. AEROASAHI)The resldential district of Takamachi-Danchi is locat,edon the elevated hilly terrace surrounded by the alluvialflat plain in Nagaoka City. Ther'e are several doz,ens ofhouses in this residential distr'ict. Some portions of thesteep slopes surrounding the terrace about 50 m high ¥verereinforced ¥vith retaining ¥¥'alls ¥vith backfills of siltysands. During the 2004 Nii*・ata-ken Chuetsu Earthquake,the retaining ¥valls at four portions of the hillsides hadcollapsed and ¥vere carried do¥vnhill together with slidingdebris. At the three portions, the retaining ¥valls ¥verebroken apart and ¥vere involved in the landslides. Theabove three sites*ere chosen for the reconnaissance anddetailed investigations ¥vere conducted in the presentstudy.Photo 6. Close vie v of landslide at Naraki (This surficial soil slipoccurred on the steep slope at the rigi]t-bank of Imo river, Ivhichforms an in-facing dip structure)Figure 6 shows t,he plan and side vie¥vs of the slopefailure observed at site No. 2. In the present study, theprocess of landslides is assumed to consist of t¥vo distinctive phases, sliding and spreadin*'. In the case of landslides on natural slopes, the first phase ¥vas characterizedLands!ide at Na/'akiThe site of landslide at Naraki is located at the rightbank of the Imo river, as sho vn in Fi_・*,. I . Figure 5 sho¥vsthe plan and side vie¥vs of the site of landslide at Naraki.by the slide-down of the collapsed soils on slopes,¥vhereas the second phase ¥vas characterized by the runout in ¥vhich the debris is in¥'olved in spreading overgently sloped fields. In the present case, the boundary 717RUN−OUT DISTANCE OF LANDSUD狂  ’    ’  ノ                    ノ                !!∫ご三婁三ここ\、、『Aノ    ’                             」   ’                             置   ’                            一ノ輔嘲韓    、                    !mm一一鄭 一一物を乏…ヨニキ=なミミい,、噛議 幽て1匁べグ1〆({、ミミ溶誌こ漣一40   、       、A    !砂べ鴛:ヨ三3黍キぐ』へ、・ノノ7    ■、、   、           、          ,40ノ    ’                            『 ’63一 1 、 ,!へ、’芥渓・さこ、  一一   〆     r                   、      、巴’   ナ  l  v\ll三蓑毒難こ   、、   ’  1   !・、 \、\も往薯一V之濃妥蒙     ’〆   ’ノ陶 \   、    岬   !’ ∫ \ 、、、__ 、      ノ!   !   ㌔、      魅鳳r20、☆1箋議雛無孝㌘謎態蓼薫濃茎i曇ξ20ノ/’、0 ∼  /             .、、一一撃一’ノ   ・\  、\0!…一冊昌冊謄一…こ・∫’許・、ん2ケ三さざミき一一一一一一、幅輪  、、こ∼こ議》グ∫老9 ∼一一く=て・、   B「ol踏ces ’A・ 、〆/一一一・、!/Br・kenpleces−齢一・一蒙髪一』\\\、\、、(a) o///一響1、、、写づ即一需『’§臭き\ぐll一’20(a)02040n1m40mPr配特faiiuぎε5口r魚ccAS疑摯e魏Re【衷滋ng、∼aU、里0Prc一魚輩ur20PQSト撫ilure SBr重込cεsur蝕cε            、R磁aming、職疑     騨OSト藍a薙u『e sur£ace、、 、.へ’0(b)02040m     SpreadingAq02040m.」Fig.6, PI盆nv産ewands蓬deviewofflowfaiiureofrecEaimeddeposiIsat   site No,20fTakamachipDanchiFig.7。 P屡該“view and s員de v曇ew off董ow fa董lure of recla亘med depos員s飢   si重e No.30f−r謎kamachi−Danchithe rear end experiences act玉ve v&lues,1eading to di貿erentvelocities of the movement between the front end andrear end.In the continuum model developed by Hungrbetween the two zones,sliding an(i sP「eadin9,is assume(i(1995),the multiply aligned lumped mass How mo(iel wasto be locate(i at the toe ofαretε葦ining wal1.F三gures7aud8used wl亡h the How resistance terms of various rheologicalshow simi三ar sets of重he p1&n and side views of the fai正uresobserved at sites No.3&nd4,respectively.functions.However,various rheologic&l properties arethe necessary parameters which are often d漁c撮t toqu&ntify.In the present study,t難e residual s勤ear strengthratio 玉s taken as & so至e parameter together with theSIMPLE ANALYSIS8ASED ON ENERGYPRINCIPLEgeometry of landsldies,controlling the run−out distancesduring laudslides,an〔i on this basis a simple analyticalGθ11ε1噸α1F1η1nεwoノ“たρブSiη7ρ1θ■nαリノ5i5method is introduced,as follows.  1t三s apparent重hat正andsl量des at a fa玉r正y rapid spee(i  The details of the proposed model are 呈1重us重rated inpose dangeτand devas亡ating effects to(iownslope areas。Flg.9.A mass of soil expected to collapse is representedThese are clted as debris fiows,earth flows,roc麹ava一byarec書angularblocklocaξedat theposition“A”sltting豆anches,and f董ows of liquefied sands.The process ofon the slope with an angleα.At thls posi芝ion,the blockdebris f董o∼v was studied and the ana豆yses of run−outhas a heightκ。alld lengthゐ。.When the block of soils isdistance were un(iertaken in the past includ圭ng the studiessubjected to perturbation during seismic shaking,failureby Hsu(1975),Hullgr(1995),Davies e重aL(1999),Daviesistriggeredalldthesoilblock&t“A”lsassume(Itomovedownwar(is to posit三〇n“B”until the right−hand si(ie ofand McSaveney(2002)aad Hun重er and Fell(2003).丁鼓emain topics focused by Davies et aL(1999)an(1Daviesthe bottom face of the block reaches the bouPdaryand McSaveney(2002)are亡he重ong run−out dista!1ces of between the regions of sh(ilng &nd spread量ng.Thispl’ocess is assumed to correspond to the sliding Phase.rock avalanches ln which the fragmen毛ation process wasconsidered to be a necessary mecha互1ism to explain suc紅Duriag thls sliding phase,重he potential energy of positionlollg run−outs,The且ndings from the study of Hungris rapidly lost,compensated wit熱the friction energy Ioss(1995)were the fac重ξhat the collapsed soil mass moves inin(玉uce〔i at the bottom face of the block.It is thell as−a retrogressive manner durillg rap圭d lan(isHdes in such asumed that the block ofsoils at“B”now si{裳ing on t難emanner t致at the front end of the collapsed soil massgentle slope w三th an angleβis subjecte〔i to spread量ngexPeriences the coe銀cient of passive earth pressure whilecleformationラ keeping the hyperbolic shape wh重le the lTSUKAMOTO ET AL^718length of the bottom face becomes greater. The block ofsoils eventually reaches position "C". At this position,the hyperbolically shaped block has a height of H*- and aAL¥ ¥¥ /¥' "¥¥tr'jj" ¥¥70 '/'f80lmlVl,iength of Xf, as shown in Fig. 9. This model involving thehyperbolically shaped soil block ¥ 'as frst developed by,V/ '60// //l' --- -"r :'Hungr (1995) to simulate the break of dams. The process/ / 651-J' " 1 /¥¥L¥¥i/t/';w".,y/i¥t1¥¥¥¥"/¥¥J ¥¥ 60 - l I I )_""'tlll ' !r/of the dam break ¥vas assumed to correspond to the/.ll If tt/spreadin*" phase described above. In the ori_ :inal model,Iffl!by Hungr (1995), the source of available energy to bespent in the base friction ¥vas assumed to come solely/l40L¥¥ ¥/from the change in the potential energy in height dur'in*'¥ -1 ! ll!//f:1_" i'-_ 55T;_' t " j¥ -/'Y'Broken pieces20of ¥vall.spreading. However, in the model proposed herein, theavailable energy emer_9:ing from the change in height in__" ;,/:7/-'- i/ )O" " llf Pre-failure surface1tSt " -!¥.'- 1','o"I//ii /'A'"" 46 '.___(a)//POst-fail sre surface /;{::: ;!E ;:;1 ;l ;F "It nsiidin'L 1"'':" ' _'/'/ _ possib{e_ surrsce- !!oI/40m,_o/l / /- 48' //r 1 ,/SPr dlng t S]idir] : /Xf} h- Lo¥.o//er--A; I ((,ii'ill11*OIn 1APre-fa ivre st;rf cel1 Retainin!;O++'IOPoiLi C T BHo__________-yH, f (X1i i :O (b)ovelll_---_ I -1post-failure surface-*rfi:i/ : Ah1;{IlISliding:A'I Snreadin ;c' I r lit-t2040xf________/ii :il2SOm60'LOH -1: f'_::1 :':li '-1- ___: -_liFig. 8. Plan view and side vielv of fio v failure of reclaimed deposits atT.From B to c _l,isite No. 4 of Takamael]i-DanchiFi('. 9.Table 1.Summar¥_ of parameters inferred from case studiesSliding distance L (m)SiteHi :ashi-TakezalvaRun-oudistance Lf (m)(m)L(m) ai (') fi (')16S17o )1157 3107_23o '14613l0633oa', 57l 47l¥,1ushi ame85NaFaki959,5H26107Takamachi-Danchi No. 2'Schematics illustrating tl]e assumptions in tl]e simple analysisl ?_.51917Takamachi-Danchi No. 310) 4 , 56132812Takamachi*Darlchi No. 4'16578, 31913Tsukidate40l026.550120 )Kanan32587,-8180 )Kotobukiyama-Danchi84941770150 )1004 Niigata-ken Chuetsu Earthquake (October 23, 2004)' 2003,1iyagiken-oki Earthquake (h・lay 26, 2003)' 2003 ,Iiyagiken-oki Ear hquake (July 26, 2003)' 197S h iyagiken-oki Farthquake (June 12, 1978)*1 The slope angles of lhe "spreading" region in these cases are assumed as fi=0. RUN-OUT DISTANCE OF LANDSLIDE150O71960O /Naraki Hi ashi- /No.4(2004 Niigata) TakezawaMushigame (2004 N 'iigata)(2004 N_ Ilgata)O /S 100Tsukidate o /)(2003 5 26 Miyag ) //Kotobuki-yama Danch!/ (1978 lvii.va".._i)Cl):'O //Kanano -n)vs::s//o///o/50e;C:No.2":l)N0.35 20Se ((2003 7 26 ¥_ ,1iya_ ;i)Al40o /olOO 150/of landslides studicd/////////Takam chi-DanchiFills behind relaininwalls(2004 ¥_ 'iigata-ken Chuetsu Earthquake)402060Sliding distance, Li (m)Slidin*' distance, Li (m)Fig. 10. Summan.' of "sliding" distance Lj and "run-out" distance Lr/////////Fig. 11. Summar, of "sliding" distanee Li and "run o lt*' distance Lof fiolv faiiure of reclaimed deposits at 'I'akamachi-Danchiboth the phases is assumed to be spent in the base frictionprocess from posltion "A" to "C" via "B", the energywith equally mobilized residual shear strength in thesliding and in the spreading phases. In the following,of position, E, a¥'ailable is given as follo¥vs,E = yH L I h , (1 )these two phases are characterized by the sliding (initial)distance, "Lj" and the run-out distance, "Lf", as indicated in Fig. 9.It is of interest to make comparison of these t¥vodistances ¥vith reference to the case studies shown inFigs. 3 to 8. The dirnensional parameters related to thewhere y is the unit ¥veight of the soil rnass and A/1 is thedifference in the height between the centres of' gravity ofthe on*'mal rectan*'1e "A" and of the hyperbolic shape"C" Durmg the first phase of "slidmg" ¥1'ith thescale of landslides as indicated in Fig. 9 are inferred fromthe cross sections of landslides, and these are summarizedmovement of the non-deformed block of soils from theposition "A" to "B", the ¥vork done by the residualshear strength mobilized along the bottom f'ace of thein Table l. Herein, based on the observations on theblock is expressed as follows;pre-failure and post-failure cross sections of the sites oflandslide, the boundary between the phases of sliding andW1 = rfL*x Lj (2)cos cspreading ¥vas determined at the toe of the pre-failuresurface. The average angle of a sliding surface a ¥vas thendetermined. The maximum depth of a run-out depositwas taken up as H., and the value of L. was determinedby assuming that the total volume of a run-out depositestimated from the cross section would be equal to thevolume of the rectan*'1e. H.XL . The value of L; ¥vasdet.ermined as the distance from the boundary betweensliding and spreading to the centre of gravity of thevhere Tr is the residual shear strength mobilized at thebottom face of the block. During the phase of spreadingwith the block of soils changing its shape from therectangle "B" to the hyperbolic shape at the position"C", the volume of the soil block is assumed to remainunchanged, and therefore, the following equation isproved to hold;2H L*= 3 HfLr (3)original r'ectangle. The values of t.he run-out distance, Lf,t.hus inferred are plotted against the sliding distance, Lj,in Fi*・s. 10 and I I . It can be seen that practically for allcases studied, the run-out distances are generally largerthan the sliding distances. The data obtained frorn theother case studies are also included in Fig. lO, concerningthe collapse of man-rnade fills at Kotobuki-yama inShiroishi of Miyagi Prefecture during 1978 Miyagikenoki Earthquake, and also rapid landslides involvingDuring the spreading phase, the work done by theresidual shear strength mobilized along the bottom of theblock is expressed with reference to the inset of Fig. 9, asfollo¥vs;Tf{(Lflcos ft)2 - L.2} . (4)_J' x'*/22!d!=W, = rL*/2at Tsukidate and Kanan in Miyagi Prefecture during theFor simplicity sake, the residual shear strength mobilized along the bottom of the block durin*" the t¥vo phasestwo earthquakes in 2003, Sam'iku-Minami Earthquakeon May 26 and Miyagiken-Hokubu Earthquake on Julyof soil mo¥'ements is assumed to be the same. Thus, byequating the energy of position available and the ¥vork26.done during the entire process, the following equation isobtained;100sely deposited fills on natural slopes, which occurredThe concept of the rnodel proposed herein based on theenergy principle is illustrated in Fig. 9. During the entire -TSUKA ¥,IOTO ET AL.720It is to be noticed here that the energy dissipated duringEva!uating Resic!ua/ Shear St/'ength RatioBased on the general frame vork of the simple model asexpressed by Eq. (5) combined ¥vith Eqs. (1) to (4), thedeformation of the entire volume of the soil mass in theratio of the residual shear strength to the initial overbur-spreading phase is not taken into account in thisproposed model.den stress can be expressed as follo¥vs;E= Wj T PV.,_ (5)Ho "' ( I ) L ) 2t fi_( )( ) 9"'=rf Ho Li If I'o)t* * ++' * 162 L 1L'I.- I ,l+iii=f (C ' fi 'oyHO Lo' I'o' I'( :)II,Herein, it is to be noted that the residual shear strengthr'atio, rf/yH., calculated from the above analysis is equalto the ratio of the residual shear stren*"th to the effectiveoverburden stress, rf/a(., as follo¥vs;fi(6)cos (x L. 4cos2fi ' 4Estiil7ating Run-out DistancesThe Eq. (6) can be interpreted as the relation bet¥veenthe residual strength ratio and the ratio L /1,. for a givenset of geometrical parameters pertaining to a given landslide. Thus, it is posslble to expr'ess Eq. (6) in terms of theTf rf(7)yHO (7 (,unknown variable I,r/L. as a function of the residualstrength ratio and the geometrical parameters as follo¥vs;' +[( '.){{ , ( )}( ' 5 (Z':1 ( +*+1'* .I ' *2 t* P L.*L4"* p yHyH 'os'os" IH . . ILli)+9'f:'fi(L . =- )J L (tan)(8)It is to be noted that the run-out distance ratio, I,f/1,.,assume that rapid landslides ¥vould take place underis no¥v expressed in a form of the third order algebraicequation. There are basically three solutions for Lr/Lconditions of little or practically no volume change. Insatisfying Eq. (8). Ho¥vever, the ratio of the distance, I,f/L*, needs to take a positive value and the run-out distanceunchanged during shearing, the overburden stress actinginitially on the soil mass needs to be reduced. In suchshould become greater as the residual shear strength ofcircumstances during deformation, a portion of thesoils becomes lo¥ver. It is easy to find one solution satis-initial overburden stress must be temporarily carried byother substances such as air-containing ¥vater or dust-fying such conditions, Ivhich is the largest among thethree. Therefore, the above Eq. (8) implies that given thegeometry of the problem, the run-out distance ratio, I,f/I,., can be estimated from the initial distance ratio, Lj/L*,based on the residual shear strength ratio, tf/yH..order for a potentially contractive soil to keep its volumecontainin*・ air existing in the voids of flowing soil mass.Without scrutinizing this aspect, it will be simply assumed in the present test scheme that the volume of partlysaturated soils be maintained almost unchanged duringSIMPLE LABORATORY TRIAXIAL TESTINC.shearing. To achieve this condition, the initial overburden stress was reduced in the triaxial tests during theapplication of shear strain.Genera! C017siderationsIn estimating the run-out distance of landslides, theTest Apparatus and Test ProceduresThe method of laboratory testin_g: adopted for theresidual shear strength ¥vould be of key importance,present triaxial tests is described belo¥v (Sa¥vada et al.2006). The lar_ :e triaxial test apparatus equipped with anMETHOD¥vhich is mobilized at a largely deformed state prevailingin the ¥'icinity of the slidin_ surface. Since the materialsinvolved in the process of slidin*' and spreading are notnecessarily saturated, it would be desirable to kno¥v theresidual shear strength in non-saturated states in general.With respect to rapid landslides such as those observedduring recent earthquakes, there would be little time forthe sliding soil mass to chan*'e their volume in the courseof rapid movement. Therefore, it would be reasonable toinner cell ¥vas used. This apparatus can accommodatecylindrical triaxial soil specimens of 12.0 cm in diameterand 24.0 cm in hei*・ht. The top of the inner cell has anopen mouth with a diameter slightly larger than that ofthe axial rod so that even a small change in the volume ofthe sample inside could be monitored by up-and-do¥vnmovement of the water level in the narro¥v annuai spacebetween the axial rod and the mouth of the inner cell. RUN-OUT DISTANCE OF LANDSLIDE721100l OOH i !za$h i*Takesaw:} sand80(crushsd to _ rBanules)- 60(complelely crushcd l;eJ)SS80/, iushigame soilt)Takalnac *D2Jlch iSi e No 2;,) 60eQTakarnac i-Dane iSite 'o 3))40l/i:20/ll/ps 40iunki20*//_ 1-oO OOG^OO I iParticie size (mTn),:oO.OO I O.OI O. I lParticle size (mm)lOrig. 12. Grain size distributions of soils Fetrieved from sites oflandsiideFig. 13. Grain size distributtons of soils retrieved from sites of fiowfailure of reclaimed deposits at 'Takamactri-DanchiThe constancy in volume ¥vas implemented in drainedloading conditions by leaving open the val¥'e to thepiezometer line. The non-saturated soil specimen ¥vasprepared ¥vith the method of lvet tamping, and an equalconfining stress of o' =98 kPa ¥vas applied to both theouter and inner cells to produce a state of consolidation,¥vhereby leaving air in the upper part of the cell so thatthe ¥vater in the outer cell stays at the same level as that ofthe inner cell. The soil specimen ¥vas then axially loadedat a constant strain r'ate. Since the inner cell is subjectedto an equal pressure from both inside and outside, there isno lateral defor'mation of the inner cell. Thus, a sli**htchange in the vater le¥'el in the mouth was considered toindicate the volume change of the specimen. The cellpressure ¥vas increased or decreased so that the water levelSaturation ratio, Sr ('/,)o 4O 20 40 60 80 100Hig:ashi-Takezalva sandDf30 - 400/0loC/:0.3(e=1.3 1 - 1.25)oalc=a3c=98kPanJ:b O.n*J1C-);phase of shearing, it ¥vas necessary to reduce the cellpressure to keep the volume of the specimen unchanged.The cell pressure lvas reduced until it became equal tozero. However, the axial loading vas continued until theaxial strain of 100/0 ¥vas achieved.Soi! Sanlp!esThe gram slze distnbutrons of the soil samplesretrieved from the sites of landslide at HigashiTakezawa, Mushigame and Naraki are sho¥vn in Fig. 12.The soil samples taken from the sites at Higashi-j!j{,:'1O.1'dj1e'¥oo204060Water content, Iv (o/o)inside and outside is maintained coincident during theconstant rate axial strain application. Since the specimengener'ally begins to increase its volume durin*' the earlylOFrg 14. Plots of residual strength ratto against water content(Hi**ashi-Takezawa)Test Resti!tsIt was found in the ear'lier study that amongst variousfactors including density and water content of soils, theresidual strength of non-saturated soils is predominantlyaffected by the ¥vater content, (Ishihara et al., 2005;Tsukamoto and Ishihar'a, 2005). Based on the outcome ofthis earlier st.udy, a series of the tests ¥vere conducted onthe soil samples by using the procedure as describedabove. With an aim to examine effects of water contentto be rather cohesive ¥vith the plasticity index of lp=26.on the residual strength of non-saturated soils, the tests¥vere carried out on specimens all consolidated to a confinmg stress of cr = 98 kPa.The residual shear strength of soils is defined as half ofThe sample as recovered from the site consists of manythe deviator str'ess, S * = qf/2, at a largely deformed state,granules of different sizes. This soil sample ¥vas mechanically crushed to granules of approximately equal sizes ofand the residual shear strength ratio or' normalizedabout D50=2 mm, and used in triaxial tests. The graint.he residual shear strength ratio thus determined aresize distribut.ions at the mechanically crushed state andthe completely crushed state are indicated in Fig. 12. Thegrain size distributions of the soil samples retrieved fromplotted against the ¥vater content in Figs. 14 to 18 for eachTakezalva and Naraki ¥vere of disintegrated sandstone.On the other hand, the soil sample from the site atMushigame ¥vas of ¥veathered mudstone, and was foundthe sites of Takamachi-Danchi No. 2 and No. 3 aresho¥vn in Fig. 13.residual shear strength is defined as S =1a . The values ofof the soils recovered f'rom the sites of landslide.In the case of the soil sample from Higashi-Takeza va,it can be seen in Fig. 14 that the residual shear str'engthratio takes a nearly constant value of 0.24 up to the watercontent of about 200/0, but it is reduced sharply to a value TSUKAi¥,iOTO ET AL.7?_2SatuFation ratio, Sr (o/o)O.4O.4' i lf {Cii II it)/Saturation ratio, Sr ("/,)6020oO 3c'o 2040 60 80 100bc; O.3iV { jt O) O.2e)e)*"lviushigarrle soil"Dfl 12 - 151"/.(e=1 70 - 1 52): O.lc):s O.1,")i *=a3 *=98 kPaO20o440O60Frg 15. Plots of(Mushigame)residual4020o¥Vater content, ¥v(o/o)60¥Vater content, ¥v(o/o)strena*th ratio against water coutentFia.i7. Plots ofresid}lalstrength ratioagainstwater content(Takamachi *o. 2)Saturation ratio, Sr ('/ )O.4o2040 60 80 1 OOSaturation ratio, Sr ("/.)O,4o 2040 60 80 100t)/ 'Takamachi-Danchi¥_ To.3t)O.3i*.Df58 - 970/00.3(e=0 99 - 1 20) j.(i c=(T3c=9 8 kPa [O 'b O.2)o: O.1S O.1,),C:::"*o::(:Oo204060OoWater content, Iv(o/o)206040Water content, ¥v(o/o)Fio. 16,Plots of residuastrength ratio against water content (Narnki)Fio.18. Plots of residual strength ratio against(Takamachi No. 3)vatercontentof 0.03 ¥1'ith increasin : ¥vater content. In the case of thesoil from Mushigame, it was difficult to prepare soilspecimens ¥vith a saturation ratio in excess of S*=600/0,and the data at a ¥vater content greater than w=400/0were not obtained in the present study. However, as seenfrom Fig. 15, the residual shear strength ratio may beinferred by extrapolation to take ¥'alues ran*"ing from0.32 to O.'_7 at the ¥vater content of 10 to '_Oo/o. In the caseof the soil from Naraki, the residual shear stren_._"th ratiois found to change from S */(Tg=0.19 to almost O, assho¥vn in Fi**. 16. The plots of the residual shear strengthratio a*'ainst ¥vater content for the soil samples takenfrom Takamachi-Danchi No.'- and No.3 are displayed inFig:s. 17 and 18.COMPARISON BETWF.F,N SIMPLF, ANALYSESAND TRIAXIAL TF.STSResidua! Shea/' StrengtllThe assumptions adopted in the simple analysis leadingto Eqs. (6) and (8) imply that the residual shear strengthratio rf/yH. can be determined by the parameters relatedto geometry of landslides such as I,j/L*, Lf/L*, H*/1,., Oiand fi. Therefore, by reading off the values of theseparameters from each of the cross sections as listed inTable 1, it is possible to infer the value of the residualshear strength ratio from back-analysis using the relationof Eq. (6). On the other hand, the soil sample ¥vas takenfrom each of the sites, and the range of the residual shearstren*・th may be determined from the data of the triaxialtests in the laboratory. Therefore, it ¥vould be of interestto make comparison of the values of the residual strengthratio thus obtained from the simple analysis and triaxialtests. Such comparison is displayed in Figs. 19 and ,_O. Itis to be noted, ho¥ve¥'er, that the residual shear strengthfrom laboratory tests is dependent on the ¥vater content,and therefore, the ¥vater content likely to exist at the siteof landslide must be estimated. It is a difficult task todetermine the in-situ ¥vater content precisely and there- RUN-OUT DISTANCE OF LANDSLIDE0.5We : heteti¥ ;eatheredsandssonemudstone /O.4tMushigam}/e''O.3l'TsukidateK nan j=j* 0.2x*/SSe/o SrOo/oFigs. 14 to 16 that it is alrnost in the range from S* = 60 toI Naraki4ro'if!shear strength ratio corresponding to S*=700/0 thusindicated are generally in fairly good agreement with thevalues inferred fr'om the simple analysis, except for theSi/800/0 that the residual shear strength tends to degrade,¥vhile one cannot see any tendency in the range from S* =O to 600/0. It is note¥vorthy that the values of the residuall# " f I Take:z.a¥vaO. lcontent. This choice is based simply on the fact that,which may be more intuitively estimated to grasp degreeof non-saturated state of various soils. It is seen in' S f I OOo/oa shi-degr'adation of the residual strength ¥vith increasing ¥vater¥vhile the ¥vater content is videly variable depending onthe grading of soils, the saturation ratio is a parameter! ' Sr=700/0,'S '} = I is Hi_t)H? :l//720iO O. I O.2 O.3 0.4 0.51 f/ c"o inferred from analvsisFig. 19. Comparison of the va]ues of residual strength ratios obtainedfrom the simple anah.'sis and laboratory triaxtal tests (sites oflandslide)case of Naraki. One of the rcasons for this differencemi**ht be the fact that this landslide occurred on the steepslope of over' 35 degrees, and the debris fio¥ved into theImo river and spread further up to the opposite bank of'the river. It is noted in Fi**. 19 that the residual shearstrength r'atio takes a value of 0.09 to 0.17 for soilsderived frorn sandstone, ¥ 'hile it takes a higher ¥'alue ofabout O.23 for soils derived from mudstone.o.5jf T'akamNaoc i- anchi j////o.4v,x/o SfOo/ov2//o SfIOOa/oo.3/Sp o.2llt)H O.1o/e Averag:e////// J?f/N0.4//?*tolNO 2 lNo.3O O. I O_2 O.3 0.4 O.5Tfla'o inferred frorn analysisFig. 20. Comparison of the values of residual strength ratios obtainedfrom the simple anah.'sis and laboratory triuxial tests (sites of fio vfailure of reclaimed deposits at 'Takamachi-Danchi)Similar comparisons f'or the case studies on the failuresof fill deposits behind retaining ¥valls are made in Fig. 20.It is seen that the values of the residual shear strengthratio are *'enerally t¥vice as rnuch as those obtained fromtriaxial tests. Herein, the soil sample from the site No. 4¥vas not available for triaxial testing. In the li*・ht of thef'act that the fill rnaterials ¥vere practically the samebetween the two nearby sites, the results of the simpleanalysis on the site No. 4 is compared in Fig. 20 with thedata obtained fr'om the triaxial tests on the No. 3 soilsample. Ther'e are significant difi rences bet veen thevalues back-calculated and laboratory-determined. It¥*ould appear likely that the presence of the retaining¥valls biocked free movement of' the landslide debris,thereby shortenin*' the distance of run-out, which ¥vouldha¥'e been larger if there were no such obstacles.Run-out Dis'tanceAnother way in vhich the estimate is compared ¥vouldbe to plot the run-out distances actually obser¥'ed againstthose ¥vhich are computed through the formula in Eq.fore the most likely ran*"e of water content or sat.urationratio vas estimated as indicated in each of the diagramsshown in Figs. 14 to 18.In Fig. 19, the values of the residual shear strengthratio for the case studies on natural slope failures areplotted. The data for the case studies on the rapidlandslides at Tsukidate and Kanan are also included. Thelaboratory triaxial test results on the soil sample f'romTsukidat.e were previously reported by Ishihara et al.(2005). The range of the residual shear stren*"th ratiochanging with the ¥vater content, w, or saturation r'atio,(8), based on the residual shear' strength ratio obtainedfrom the laboratory tests. The values of the run-out distance ratio, Lf/L*, obser¥'ed in the fields for each case ofthe landslide are plotted against those estimated frorn thesimple analysis in Fig. 21. The run-out distance ratios,Lr/L , are estimated from Eq. (8), by taking the geornetrical parameters of H /L , Li/L*, c and fi, and byassuming the laboratory-determined residual shearstrength ratio, if/yH., to be equal to that correspondingto S*=700/0. It is of interest to see that the run-outdistance ratios predicted for the cases of 2004 Niigata-kenS,, is indicated for each soil sample, and the value of theresidual shear strength ratio corr'esponding to S, = 700/0 isC*huetsu Earthquake are generally greater than thoseactually observed. For the cases of 2003 Miyagi-kenindicated as a dark round symbol, based on the testearthquakes, the predicted values give the run-out dis-results shown in Fi**s. 14 to 16. It is to be noted here thattances'hich are smaller than those actually observed.the saturation ratio, S*, is chosen rather than the ¥ 'aterFor all practical purposes, it is of use to plot the resultscontent, w, as a primary parameter controllin_ : theof the sirnple analysis derived from Eq. (8) in terms of the TSUKAMOTO ET AL.79 4f//2.5Hit:gashi- I o/'Takeza¥o 'a / Tsukidate; l.5Eoll/1J,oKananMushigame////)/ / / // / // //Tf/aro O 112 / O 14/ O,16/,13Naraki /(1 3s, 6 02) /7 .5/// ' // / / / 0,18// ////// // /' ' O.2 _ll / // ////// /+,:s777T7 l/o // // j infL'rred from T rcF'o - Iv Felations l/o // j correspondlng to Sr = 70e/oMuShlgameHo/L0=0.5Lf/Lo observed in the fields// / O,_"_5=01l .5 2l/// / / L L700/0/_// The residual stTength v lues areo.5//// Srs809/ // // / /1 .5-///// / /// // // O /""/// / // // /// /// / . ,/ -l./ ./// / / / / //// / / /// O,'4///// ////// / / // // /l /_' / /// / / / / / ////////// / oL=23 ,/o. )l .5l,Li/LoFig. 21. Comparison of the run-out distance ratio, Lr/1,*>, observed inthe fieids and estimated from simpke anal_vsis (sites of landslitie)3, .5/ / / / 70)/( / /i .5//'r/// / // / / /)' / / / / // /P//.i,/:///¥=1 .5/ /// / /// / /HigaShi-Takezalva///// // // //// //// / / /o / / Ho/L0=0. 1 5 5 !/1/// / (X'=17o , p=0 f0.5 /0.5/l .5i,Li/L,oFig. 22.P!ots of I.j/L,, agninst I.f/1.*, (Hignsho.5 /0.5///// / / /ll / / 6ll //1// // // // /ITf/a/ -O.)?_/ / / / / ////l/ / / //O 18/// / // ///'/ ..//// ///// !////////ll/// // / / // ///*=2.5O.16/ / / ////'/////// /' / 0.2/ /0'4//// / / /// // // // ///.O.'8 / / /O 16/// ///// / /// O.14S L//80)/(/// // //,3/////llo- o'r .// _/o /1'//.f/C/// // //// // ' / / /ZQPlots of L /1,*, against Lf/L*, (Mushig me)Fig. 23.///////////NarakiHo/L0=0 }?_3= ce=33' , =01.51,Li/L,oTal ezal a)run-out distance ratio, Lr/L , a*'ainst the initial distanceFiU. 24.Plots of Li/L,, against LF/I 'o (Naraki)degrees and fi=0. The dark round point corresponds toMethod ofEstil77atillg t/1e Run-out DistanceThe estimate of the run-out dlstance may be made viathe procedure as follo¥vs.(1) For a given slope, it is necessary to assume a poten-the values of Lf/1,. and I,j/L , Ivhich are observed in thetial sliding surface on which a landslide is postulatedcross section indicated in Fig. 3. The equal contour linesto take place. This may be made by tracing the zonelvith respect to the residual shear strength ratio, rflof ¥veak soil deposits ¥vithin the slope, based on theyH. rf/(Tg, are also indicated in this diagram, correlating the run-out distance ratio Lf/L. ¥vith the initial dis-geological and topographical setup of the slopetance ratio Lj/1,.. It is also seen in Fig. 2,- ho¥v sensitive(2) Dra¥ving upon the array of information as above,the run-out distance ¥vould be a*"ainst the residual sheargeometrical parameters such as the angles of genericportion of the landslide, ce, and of the spreadingratio. Lj/L . Figure '-2 sho¥vs such plots for the case ofHigashi-Takeza¥va, in ¥vhich I,./H*=0.155, c =17stren*'th of soils. In Fig. 22, by inferring the ¥'alues of theresidual shear strength ratio at S*=600/0, 700/0 and 800/0from the r /(Tg-S, relation sho¥vn in Fig. 14, the equalcontours for the respective values of S* are also shown.The same diagrams plotting the values of I,F/1,. a*'ainstthe values of I,j/1,. are sho¥vn in Fi**s. 23 and 24 for thecases of Mushigame and Naraki, respectively.under consideration.portion, fi, are kno¥vn. The distance of slidin_g:, I,i,defined in plan view through ¥vhich the mass maytravel do¥vn must also be assumed in advance.(3) The residual shear stren_gth ratio also has to bekno¥vn. This may be made by testing soil samplesreco¥'ered from the deposits representative of those RUN-OUT DISTANCE OF LANDSLIDElikely to be involved in the landslide.(4) Having kno¥vn all the par'ameters as above, theformula of Eq. (8) is used to deterrnine the ¥'alue of'Lf/L., and hence the run-out distance Lf'79 5of Science. The authors ¥vish to express their gratitude tothem. Thanks are also extended to Asia Air Sur¥'ey, AeroAsahi and Pr'ofessor H. Marui of Niigata University forpr'oviding them lvith the photographs presented in thepaper.CONCLUSIONSThe run-out distances of landslides on natural slopesand fill deposits ¥vere examined based on the case studiesobserved during 2004 Niigata-ken Chuetsu Earthquake.The simple analytical method ¥vas introduced based onthe energy principle, in which the residuai shear strengthratio is a sole parameter to determine the run-out distance. By employing this simple analysis, the residualshear str'ength ratio ¥vas back-calculated based on thegeometrical profile of each landslide. The laboratorytriaxial tests lvere conducted on unsatur'ated soil samplesretrieved from the sites of' Iandslide. By conducting aseries of tests¥'ith varying ¥vater contents, the residualshear strength ratio ¥vas infer'r'ed fr'om the results oflaborator'y triaxial tests. Based on the fact that the valuesof the residual shear strength r'atio back-calculated andlaboratory-determined are compared fairly ¥vell, thechart correlating the sliding distance and run-out distance¥vas drawn for each case of landslides, which is of practical use.ACKNOWLEDGEMENTSThe laboratory tests described herein ¥¥'ere conductedwith a help of M. Nakamura, S. Koh, R. Igarashi, formerstudents of the soil mechanics group at Tokyo UniversityREFERENCES) Davies, T. R., ilvlcSaveney, IM, ,J, and Hodgson. K. A. (1999): Afragmemation-spreading model for long-runom rock avalanches,Can. Geotecll. J., 36, 1096-1 1 lO2) Davies, T_ R. and McSaveney, .¥,1. ,J. (2002): Dynamic simulation ofthe motion of fragmentiug rock avalanches, Can. Geotech. J., 39,789-7983) Hsu. K. J. (1975) C atastrophic debris s reams (Sturzs roms) generat-ed by rockfalls, Geol. Soc. Am. Bu!! , 86, l'_9l40.4) Hungr, O. (1995): A model for the runout analysis of rapid flowslides, debris fiows, and avalanches, Can. Geotech. J , 32, 610-623.5) Hun er, G. and Fell, R, (2003): Travel dis ance alrgle ft)r rapidlandslides in constructed and natural soil slopes, Can. Geotech_ J.,40, I123ll41.6) Ishthara, K., Tsukamoto, Y_ and Nakamura, IM. (2005): Considerarions of lhe landslide in the 2003 N,iiyagiken-okl Earthquake, Int.S_vmp. E'arthquake E,,gineering Commernorating Ten!h A nnh*ersar_vof !he 1995 Kobe Earthquake (ISE'E Kobe 2005), January 13-16,Kobe and Alvaji. B-'_1 1B-226.7) Japanese Geotechnicai Sociery (2003): Report ofDamage Investiga[ion on Sanriku-Minanli E'arthquake anc! JVliyagiken-Hokubu Earthquake in 2003, 859-872.8) Sa¥vada, S., Tsukamoto, Y, and Ishihara. K. (2006): Residual deformation characteristics of par iaHy sa urated sandy soils subjected toseismic ex'citation, Soi! Dynainics and Et7rthquake Ejlgineering, 26,175-18,-.9) Tsukamoto, Y and Ishihara. K. (2005): Residual strength of soiiinvoh,ed in earthqtrake-induced landslides, Proc. Geotech. Earthquake En*"rg. Sate!!ite Co,if.. September 10, 2005, Osaka. Japan,TC4-ISSIMGE, I i7123.
  • ログイン
  • タイトル
  • Evaluation of Natural Slope Failures Induced by the 2004 Niigata-Ken Chuetsu Earthquake
  • 著者
  • Hirofumi Toyota・J.Wang・Kouichi Nakamura・Naoki Sakai
  • 出版
  • soils and Foundations
  • ページ
  • 727〜738
  • 発行
  • 2006/12/15
  • 文書ID
  • 20958
  • 内容
  • fSOILS AND FOUNDATIONSVol. 46,i¥'O727 738, Dec 20066._Tapanese Geotechnical SocielyEVALUATION OF NATURAL SLOPE FAILURES INDUCED BY THE 2004NIIGATA-KEN CHUETSU EARTHQUAKEHIROFUh,II ToYOTAi), JINxlNc}* WANGi;), KoUIC}{1 NAKA IURAiii) and NAO (1 SAKAii+)ABSTRACTSlope failures occur'r'ing near' earthquake centres have attracted much attention since the 2004 Niigata-ken ChuetsuEarthquake. The 1995 Hyogoken-nambu Earthquake affected large rnodern cities and urban infrastructure, but hillyand mountainous areas suffered hea¥'y damage from the Niigata-ken Chuetsu Earthquake. During that earthquake,numerous landslides occurred in Koshi of Nagaoka city (formerly Yarnakoshi ¥'illage). Social problems de¥'elopedwhen many to¥vns became isolated because landslides cut off traffic and public service lifelines. Soil from landslidesclosed river channels and formed natural dams along the Imo River, 'hich fio¥vs north to south in Koshi. The naturaldams subrner'ged some towns and emer_ :ency measures ¥vere prornptly undertaken to prevent debris flo¥vs caused bynatur'al dam breaks. In addition, many surface slides also occurred at steep slopes. The endogenous and exogenousfactors of those landslides must be clarified because of their fe¥v precedents. For that purpose, topography and geologyin this r'egion are arranged according to each failure type. In addition, triaxial tests of saturated and unsat.ur'ated soilsobtained from a slope failure site lvere conducted to examine the soils' strength properties. This research is intended topropose estimation items that can help to judge the risk of natural slope failure.Key words: earthquake damage, geology, site investigation, topo*'raphy, triaxial test, unsaturated soil (IGC: B1 1 /D6)A Ko hl 7INTRODUCTION<+i>j E ( ormer Yamakoshi viila ;e) /At 17:56 on 23 October 2004, the Niigata-ken ChuetsuEarthquake, ¥vhose rnain tremor ¥vas magnitude 6.8,380- tstruck central Niigata-ken (Chuetsu area) and seriouslydamaged infrastructures of hilly and mountainous areasincluding Ka vaguchi town, Ojiya city, Nagaoka city, and-' *l l:Ojj!a citKawa uchi town/!r'*¥Joetsu citdepth, with characteristic frequent strong aftershocksthat engendered increased damage.i(]Takada3 70-In the Chuetsu area, thick alluvium covers plains; hilly--' ;h+___'agaoka citytheir envir'ons (Fig. 1). The earthquake ¥vas an epicentralthrust-fault earthquake ¥vith a hypocentre of' about 10 kmL¥1tToch io cirh+lrishiQ a¥vH,pocentre ofmain shock"/_._/_j ) ]fiigataareas are composed mainly of soft mudstone of quaternary and tertiary deposits. Quaternary deposits are ne¥vstrata formed from two million years ago. Ter'tiarydeposits are geological structures formed between '_4million and 2 million years a*ao. The geomorphology1i Prer ;'c ure r{ i) ailchil 3 80 M)oko eit)l i'139'Niigata prefectvre/7'fl : "i '1] Ii Japanii ______T_formed by folding presents a prominent landslide area inthis region. The regron exhibrts catchment area dotted¥vith many ponds and r'ice terr'aces. In addition, rainfallFig. 1.Map of'iigata prefectureof more than 100 mm was recorded from typhoon No.ogical Agency, reached 115 mm. Under those circum-'_3, ¥ 'hich passed through the Chuetsu area t¥vo daysbefore the earthquake. Daily rainfall of 21 October 2004at Nagaoka city, based on data from the Japan Meteorol-stances, more than 3,000 Iandslides occurred in the hillyarea close to the seismic centre during the earthquake.i) Associate Professor, Depanment of C*ivil and Environmental Engineering. Nagaoka University of Techuology, Japan (toyota(jlvos ,, nagaok aut . ac . j p) .Graduate S uden , ditto,iii) postdoctoral Fello¥vship of JSPS, ,Japan.i ) Researcher, Public ¥Vorks Research Instituie. Japan=i)The manuscript for this paper lvas received f'or re¥'ie¥v on iMay 9, 2006; approved on Oc ober 2, 2006.¥Vritten discussions on this paper should be submitted before July I , 2007 to the Japanese C} eotechnical Society, 4-38-2, Sengoku, Bunkyo-ku,Tokyo I 12-001 l, Japan. Upou request the closing date may be ex.'tended one nlonth.727 TOYOTA ET AL.T28' _"'"- ' :''"f "' t! ' t _i]_:'i ''l't ' ' 't' ':*1Ic:sd'!"' : ; ' 'YOlltOcrlhlra'li's:_';''t*( '*c;lj*; 1:':::':::::*.::f::::b:::: a ':;:ii/::::i ;l - t t'1" ;' t i t f "-Jb J' *1 "' '/4 :_'J 'jil;t; ' ';t :j,._:: !.'.- ;d;:.; :_":'t'I ";*:i/i : !''t'Nlgorisa¥va'.I:;"'t '; ' ' -{ ! '=_I_' '7:!;.111"li;;i/_,'__'::'t i:,:::. ;,'tJ::;: 7:!: !: :fl o- L: 1¥'. '/"':: tst: l¥A JL't_"I " / ' :";J''J4-_ ' : " I" '. , } *+':;;;i!'_'=ai Vi "'t4*___'''"'''?''{';"'' '',,,'_'A.;'_i':'" :' w' :'.' __'i 'L'*l-t-"' ' .1F*__!t7" ' A"' ' '*t'-_ ¥::: :. -¥ : '"" l jr:/' I ':L;J ' ;L ;i: /'¥' fOhta4¥'Rrver!';' '' l':;:e' ' '! :iL+i:;= ' : !E'; ;;;' _:' t( / '-' _'{'::_I':L': :1_:Li? "_'_::'':' '';;*;i':'f{''1:v;f'+1_l':, _ :, ;_:'i;:'! :1i':'' 'j_': 'j' '4= ',':!t j!i':'-s!i t::' -':;_'.s't . "t'r':;: ';',}.:_'!'i_:'_'1}tl¥ v- iTanestthara ' ' l_: __e ;' _ t _>:? 1' t i- t --'' ._:' .__! ',' -s;L:;/'1' ;'")' j""';L =t' .")¥Iakano. ._ . */ '* .: *v/tr"';i ':i ::/: :; :t 'I.._*'_*.;¥_,:.'"a:;' " s'::' ' _ ;dt_ l.'_:¥I'f:/}'g _t I 11 , _-j;.i :_ --','}'' ' _1' :_Tel' ' ' l' ":' t:" anO' ' _ 'i{T..;t ;:'ir''":,,_,_ _; ..' 5.-'('*;;'!___;''''''' '! '"t' - -/'.' . rr::' .si': !,.ai:::.L:',_...:_j'_te/"=_o"L_'';1r//"/''" 's$-'_ "'e_ ''i;'ri.i '-.' _I:'.:_:',I,_.'.l_.;'/.,_ lss,*-a 'F:: '; s !1' 4ri ::: I '* '*' ;;{.:. t;' 's ';.;;:.t'l'tr rtr! 'P:7' '¥ tlt;::i !___:j_ ;; .1' ! -- i; ,- ::-: " : :] i :::!i_:i '-- ___; ;;.a; iL'*+** t'*i * ' '!* '* " *' ' ') * '!i :* !' lITI_ORlvel(! "l ' ft" ' ' ' . "', ' . ;¥. ;h' ' ' )';Jl;c1's",' II '. e'*'t-r ';' -} .t ' _""' ', ' I'i;r' i: ! ''ii¥ fi f 1""tt '{:t' :l:?' ,'e: ci": i.i._{...:s!:It: 1';1(!'/ e¥q f; _::itt ;::i;i;(-L L 'j f ;'s"F' :__rl i/l t;:t't; 'L t t ;'; i Jf:f';: :;'; t_!]r:LI "'iFr ' -.¥' /'**;:' :; i .i; l : t;;;:'tk:/t:;:I' i!:L ¥;:_ /'; "'fl /'el'l'^ ' / It -r- 4 ¥_¥t;;_;.5ei.;;";i;:t':, i ;"";:;:*::_f'.:;;r::,,,;: : *. 'S' "i* sj;?' :,,tl:';i'= 1i','.' _:. ;,s':1' *:.*: ',:'* ::,':t,' 'F, .':;!'1";i- 1:*'la_nds.. ;.t : i : ;'. _: ..'¥'¥. : ;::I /ai':¥::: "' "" 'i ; ' " -"- _a 1i ' '/PI 'ASahi Rlvel - ''t"*':' lr :'<0r"'"''='' ?'!i ''s 'Li''t ''s __ _ f'hIlde.' t ¥<¥_(i- i¥r: :L 1't *-tieJ'Qt' -- " -']'p*_I'1 ''Q'cj:I:cL?i:i;jl¥:j:!?''_ ;s ;':::,i:; !'/;;;T;ii:}'_ ' . _ . . . _G'¥/". _ _"; " .' ='_ _::..:. "_': l' _'_ :1-'! ':!__L r' j!Plr'?';; "( " : s ' ii " _ L* 'tl ' ''!:" $1': t'l'k 't i_ " !;- ' 'r '; j j I:1 _1";it . , ; g._t" :'. i"r'i ; ":: ; ltt: : t'l]:/f't'/J "" ;:_ l}'itamp e l 'i 'i:. ';i :'J ' : ' "' /'t ' { 'ii::ji:; :: i::: ; P:: 1 i { f(; ;; :'; ; l}'::# ' ! -rf,1:P{Asahlka¥¥ a//'p ' i' " " ' ' ' ')Jsrt:'-'1" ":1'L A" ' L s ' _ _ _ __ L _ ' " : " i'_' 'i'i i_ _ '_ '_' ¥! r: !""s #* L' : ' "' ' ' ! L'_$ ;_ ' ' "' ' __1*Lt '_ ' t_ ' ':' ';f;r'"'' 'tg ri:;"''-"l"''''!!':t'( j::ll)'S;: "fit;:::i:i :'";: :. ..' ,... ;'1' * v ;_ _ ; / : ;¥)":¥ 'rl. i(' ';_ . . ;dl "J' c 1i ii:1:r'i;;/i: :::1 :::j;;::;:jSi:::{::j:;{ ::{j';;i:? ; :i -_i', __'!1'I::./: NaranOkl'ti': _{-_*' ""i! 'Iele'! ; '-' ' i '-'f'h'tr!I:a:,i -i''r ''S _''!*_,.:A;s.:;';'*1 a'i"ee:/;ijji(:t!'i:jJ(i;i.;:_landSllde.,_:".e - 4!i : ";': j:; ''t:;;]l'L_;. ;"*" ;ti: "'j_ I ',;(,. I;_r''Nanatakl_' '"e' _tt';f;$_ _'f' -'ri_'__aRlvel'il:-i/!/:_;i+;__'_t'I':1r1ci__1f2 #;-- '-'e'-_ h .'-' _t_'_.!J:__.' ''-___ ;;J''''''+-"L:":i,. , . - .. ._ ;: +'! ' ' ;j;:!f' ;:: '( :6?: t_f:)'t'c?! i;?:" 1'" ' :j$': ; _e"':':; !;_.::r! l ;- ;Ir: t:;i: ;!' ij';::"f :;L(*:.i:"i'i ': : L *_t ' 'I :!';'.1p .- .., . ' -' ' :'-l.' .i ,:'c'''!'''tb_ " . ;, '.t i : ' __ i__ _ ;: ;'. ; '.' :;' "' -'' i ' ' '* e-*"_*_t'''1'' 'I' '_ :1 t' - . l"L_'_ _" tt =sfs,' 'Q :_ ;" ' '' '_:*- '_'-'-:'_? $'Its ?' _./ _ }e '1F';'!:' " -; " L'[-' ''I" f '' ;'_ L'f .j:--_F-';";'' ' ' L!Lfl-"'- '_;¥r '_:':__,_'"=}_ "- (': '; ''-"r C_ontourintel /al'' ' '' i'i ':'tes'--''_t'_'_"_':'1andsllde. '¥_ _'';'/- ;! ' - - iO:a:p'1 '- fL--'! t _"' !_'_t'-' --'9-___I -' ! i' 't';-"'__-t_-' -_' "'''-'; ,5c'IP')s':¥: ' "t't';,: :v':: +";; :: S : ffe'l'l{,tsj_$;:.::e'l':L'i;:' ,L'l ;''-*-;i':'" -_''Li '-'- : ;::::'iil ! j_ ' J 'tii!t'':::L:if-;i:.;! ;::";'''-L'"f"$'_l'jj{.flj::'f-' . :l;j'-;'; -::'*'_4 =';' 't'?r'' ; L ' bL. ':;(r'h)rx-1 Sirtpf Maesa¥¥a "'Rlveli ;' :'s'i':t_i::;s;:a0'f !;;i::; ,;'_ sl"E;: ".. /1- -' l'i'i-1 __' --_''-j;1 '' r''' : ::':Si:' :; ':__"* _'jl:' t t _it' L' _'/"t:i'tQf_-!;'_F P - Ie' *i*i- - '- '';'";":i' -'ij'=! ' ;i'_i'__ :1 ' :i _::'__ :;er' ----1!: i'f;' ' :;_::-t' ;sl:a;''i '_]'Il"-a''h'!iHlaaShi takeza¥¥ra i1 P"':e;:Lii ' S;s ' '*' qb' eF1:'I an d S I I d e):i : :?- :;: . -- L' * ' *;:v fl ;'*}1' _l 'i; ' ':I_";i ¥'T ' 1 t ' '- 'As7 _' If' ( ¥'_/:1'': '1_ "'t : uh ' ( "i'-;';11' Junl_dalla::i;; ] i--':'h 1' ' ' '_Scal,.Fig. 2.}. ./t '¥/-J ;(m)Disaster map around Kosivi of Nngaoka city (provided b,. Geographtcal Survey Institute)Their specific damage information ¥vas reported by theJapan Society of Civil Engineers (2006).The present paper describes topographical and geological characteristics of the landslide area caused by theearthquake. In addition, monotonic and cyclic triaxialas a representative designated landslide area. Figure 'sho¥vs a landslide map of those areas during the ChuetsuEarthquake provided by the CJeographical Survey Institute. River-clogging landslides along the Imo Ri¥'erduring the earthquake are "Terano" "Nampel"tests ¥¥'er'e conducted using undisturbed and reconstituted"Naranoki", "Higashi-takeza¥va" and "Junidaira"samples obtained from a slope failure point to elucidatemechanical properties of saturated and unsaturated soil.landslides from upstream (Fi . '_). Among those landslides. Terano and Higashi-takeza¥va ¥vere large-scalelandslides, ¥vhich required urgent measurements againstOLD l,Ar ITDSl,IDF,S IN THE KOSHI ARF,Ariver clogging."Tanesuhara", "i¥,Iushigame" and "Asahika¥va" inTable I sholvs t,he number of landslides that occurredbet¥veen 1949 and 7_OO _ around Koshi. Many landslidesKoshi (formerlv Yamakoshi village) have been mentionedoccurred in the Asahi River basin, extending from NATURAL SLOPE FAILURES 729Table 1. 'umber of landslides in Koshi (1949-2002): (a) classified b,basin and (b) classified by region(a) classified by basin (b) classified by regionNurnber of = Number oflandslides = IandslidesAsahi River 29 32Asahi RlverNanataki River 19 15TanesuharaOhtaRiver 19 7Imo River 17,1ushigameTable 2. Landslides in TanesulraraNo,Disaster Place ConditionsYearLandslide Disaster1 1824 Nakanohroughout theApril: Length l.4 km, village. 500/・ of rice fieldsthawin¥Vidth 700m could not be cultivated.The history of former Yarnakoshi ¥'iliage is ¥vell record-ed (1981). Its history frorn 1700 indicates Tanesuhara asthe site of the greatest recorded earth-flo¥v disaster.Table 2 sho¥vs records of landslides in Tanesuhara. Thelandslides are not recorded throughout the Koshi areabecause no such large ¥'illages existed except Tanesuhara.Although landslides in prehistoric tirnes are unknolvn,the first landslide recorded in Tanesuhara occurred in18・_4. An exogenous factor of landslides is sno¥v-melt¥vater in early spring. At that tirne, slopes ¥vere destroyedand t vo lar'ge clogging ponds lvere created in the ImoRiver. About 100 years after that e¥'ent, a landslideoccurred during the sno¥v melting season at almost thesame place, thereby forming a natural dam. During theChuetsu Earthquake,vhich occurred about 80 yearsafter the previous event, the river-clogging landslide vasbroken at Terano near Nakano. The decisi¥'e difference isseason Flood inducedby riverclo ging. Two ponds ofthat the precipitating exogenous factor was not sno¥v-lOOm len th and 10mriver clo*'ging caused by landslides is a commonphenomenon in the Imo River of Tanesuhara becausedepth ¥vere created._, 1926 Landslide Nakano Fields and mountains ofl¥,Iay Length l.1 km, more than 2km Ivere¥Vidlh 180 m damaged. Fioods inducedby river clogging.melt runoff, but an earthquake. As mentioned above,large landslides that clog the ri¥'er have occurred e¥'erycentury.Conceming unrecorded landslides of prehistoric times,the landslide history of Koshi ¥vas investigated using a3 1 929 Landslide Terano Prefec ural route (Tochio-topographical map that was compiled using inf'ormationbridge collapsed. Riverfrom an aerial photograph. Figure 3 shows landslidelocations durin*' the earthquake (provided by theAprii ¥¥'idth 50 m Ojiya)vas severed;clo :girrg is unclear.Landslide A Iarge landslide occur4 1932 BetweenN. akano red during constructionand of landslide measuresTerano De ails are unclear.Geo*'raphical Survey Institute) presented on the map of'old landslide topography (provided by National ResearchInstitute for Earth Science and Disaster Prevention(NIED) and the Japan Science and Technology Agency(JST)) in the Koshi area. Many landslide configurationsremain at the Imo River basin bet¥veen Terano andMushigame, and in the Ohta Rrver whrch hasYomogihira and Nigorisa¥va as designated landslideareas. Although 17 Iandslides occurred in the lrno Riverbasin, almost all occurred in the Tanesuhara or Nakanolandslide area located upstream. Landslides werereported only slightly in Nampei, Higashi-takezawa orJunidaira, Ivhere natural dams ¥vere created as a result ofthe earthquake. When classified by re*'ion (Table 1(b)),the number is large in this order: Asahikawa, Tanesuharaand Mushi**ame.It ¥vas repolted in the Japan Landslide Society"Record of landslide disasters in Niigata ('-003)" that alarge landslide took place in Tochio city (Fig. 1), Iocatedin the northern part of Koshi, during a large earthquakethat occurred at Takada of Joetsu city (Fig. 1) in 1751.Subsequently, this area carne to be called the "Higashi-nakanomata landslide area" and ¥vas desi*・nated as alandslide prevention area. Although it is interesting thatsuch a large landslide occurred at the distance of about80 km from the earthquake centre, its detailed cause isunknown. According to a geolo*"ical map (Takeuchi andKato, 1994), alternation of' mudstone and sandstone(mudstone dorninate) of the Quater'nary deposit formsthis place.-Nakano, although it is difficult to estirnate their ageBecause of their geological characteristlcs, Iandslidesr'eadily occur there. According to the village history(1981), the cause of the large landslides is not rainfaH orearthquake but snow-melt runoff from spring tha¥vs.Tanesuhara and Mushigame villages ¥¥'ere built on a sitethat was flattened after t.he landslides. Although theirvillages are covered by fertile landslide sediment and arepresently stable, mass movements are continuing aroundthe villages. Large-scale landslides such as those ofTerano or' Hi_9:ashi-takeza va rnight be re-slide typelandslides because they are located on old landslidetopography (Fig. 3). Those landforms jut out as ridgesand form r'avines in both sides of failures (Fig. 2). Seismicmotion tends to be arnplified in ridges.CHARACTERISTICS OF NATURAL SLOPEFAILUREh/lany river-clogging landslides occurred along the ImoRiver during the ear'thquake. F'igure 4 vas prepared byoverlaying the landslide locations during the ear'thquakeon a simplified **eological rnap pro¥'ided by Takeuchi etal. (2004). Landslide designated areas obtained from aconservation map of Niigata prefecture (1982) are also ;i'_' _..* i''TOYOTA ET AL730* ;s "I _ _* ' 'f= ' _#t;_ / *':' *_ _r;' ";;'/* "'''v-l" ' *l'_ i t':_i _'/*t;'-'=j+t:'s_*" '_'*" s'e7*,1;:'_i alulej*ulmg;i' /":''; '*''_"_***s#* #x"s#- t';:"+'r!;t*f'' ;'#i/ "'' i;_j',;i:;;;':j:' /}* ;' = __ : : ' ' L "'=#--'topo crliir :;i;';:i;;L' / i:i'aph:_' .' vt- 'i-Is;'1i)*+4'+'> E=,;i::t' (l ;:il:i'Old lan sliri;/ ' ; ithe earthqua' ; Is; 'iIff; : i c )1( ':'; //' *''ske; {' "' "':::'s:;: : )>t'dfst i:_ . i;+*/ixlTanesuhal a vl llacFe !p ?'s, lMushiaame# ;ri;i!:;;';;tt:'i (';;i;_***' : s;:i;*s'_ :'IS''1:';':_rl'';;ft;/:" :1::!i;;/ ;'#'s!'vlllage/'_;'^"i'!*'1;l;#";#}^#;*ji:*'"**_''j";'+*'; i'jit' '''!: i{* i"*i_ tll_* si:1rf:::i.,:ii:::..:'t;"''_i_ *_ /,,;#''.J;_ ";(t. s r/;"'s.,_ ".t#'-"'* '';_' 'rli; i_,;_ j ';i _.:';1;s'_s' ^,,/Jf:J//J;;. ;.""_p/ 1r_'*'''r'.,_ _ SS_ 'p_i:# 's{s_'1,s s'+*., ' /- '/ ^- #+#1'j _'_ ;;4:*#- =Ss:*i-ft.s't"'' j ' ':-__;***'*' ! j::/{"*:::;i:(;ij:I,{*;-! : ; ii'f;:::::i::' *._-_S', . j + __ '_ _ _s.__s ' :t^''""'I :i.-_* _'.;'1 !i""";fql?: " 1J;}!"' l11?,'; _' ' 'i -'_'_ '-IiS ii :4 ' '; 1lr!!:'' )/ ://'ii'i ';#;¥ V :;(iif:i :: "-''.lLi {{' :/'' ;": L' _! ;i:ff '" j ; i':;f'; '_ rf-;i;;/;- ;; i_'f/# "; ' _;'"s - _ _ '-/' ;;'ii*S;;;.'1 '#_ ' _ _(/'f2SS ;!"#{.'::; ;; ;i':';:: ;:;: f; /!L#(';;:;!*S;: ! ji; (f'!;;.#" ;'!'": # -'^;f '_sj. *. * 1';' { ; ? /!::ii)/r {i l)s;/ i:::::: ;i"'sjj {+* ;;{{; : *!;: i:}f ::::};.,ii;i'ii;x !i:i/'{j::a; i :: : ii;;; ;iilij{;' * F, +vi"!' _ '"f'*:i;i'{ '';'/' {_ '('';1;1 _';; '# ';_ i!{ !i +i=1 ;;j' Iandsllde2;;;f' I.ii- - _ ;:_ .*/ ' ;:; ';';1'::: s * ' i : $ .*.; ;'/'i; :; i;;; '" #'* i ;s _ '_' "!"'l;;i::!!";?i; ';'" si/ '!//";ii#": ' ' t"i i; i{_','!/' r ''- ? : ' f ( i;- -+"" *- : r * $-- _x; ' t*- ""4s; ' ^=' * - ' ';t "!{' ir'.'.!"!' ;; '::::;'f:;"sS;; ;;';:;';. ;. '; '' ss+iS* ; ;;;:i ('Yi.!'-''i';;'!' _"__. iI:+_^':i{;;* * cr;';'Lil^;s*'#*:'-i"'r' #?It' *x i'{S;1#_""f*x{j:-''_/_;__"/il;;:t"?'T;*'i!:l**s'r*+,,_;__'"-_-''st';''':'!P'i:/};;::i:::{_L s;_;' ;;_"'_s" /{; ;:;;i:.NamPei'; c-: -_';" "}:; '-i:_s''!;::ii; ;r _'$ -'_':''':r 'I'sit;;'*'ii:;;+;:'tS;;r'*'"_;::'"::* {"=s} :}t1!'_ '' ' .'" :'i" s" * +*i'*' **t_] -'*;; T4';s i ;;! 's;^' j;'i: #:;;ii :;;'Si t;'*i*i; {i;i{ I ;/:1::;;;;1'ii'i'-)'!X ;# '; SS'itP'; iFj t; ; !4 ;"':'fi''' :'' .*+' A s:; ';s_+ *j__:_ , : . ' . sfrsc;;';:/;i+ ij_;;/;/;1:''t :t'$*;; "# tv*' i'f- ^' '-'_ '-t ^- s; :s*- s'- -''" t '"';' :r'-#.; '_;; - i S 'ti'S; =:t' ' ;/ 'S ;i;1;i:;i::Siri'_e!_'_';!;'; :: / ;_ '# ;*;/"ilandfsl/;f /!;: j!:::{i: ;' ,f j;;';: i' :'" ;. _ '_= !:' SSi # f ; ';::(;::i:i'::{::i41:i;;i t's"";;j: 's #; Si! !_ ' ;*;1 ";"it:t ' L ;i ;#1}O' ',"'¥zASahl Rlvel;s:i ;i'! "_ ' ^l!/' ' - - ': - -;' ; s -+' '! 'f'x#-* ss * ' -r;i-;#T-" s' /''t'-; '# '' -- a: : j;* i! ;' ?:';;';l/ii: '':t:;)S; !#':landslldei' #i ;i:I ;i:::,#:S:' ''S; s#+t=; !*^-/_' /s')#" '!;x * i .ic/::{:;,;'s__ " ;**!^ *t .' *'*. ** " {_' ' " +*? s.* s9s:i' l'__ '/ !#'t 1 ' S i/;i -- '-'; 'r-rr!'/ r's'l r{t - 't ^" ' "* ;*i 1; "unlualras s/'-"ts'#s .' 'f!!''- -'""i:1' F#i '-'' x" '# ' ' #1":'^T1S'x;1.# :--(^*:{i/'#;;' ; '! ';'rl!#--;fs****s:1s'"il !'i'!j{ ii^ ;!;;; !; i ; :;tt{:i ' i^;;; ; :i r' !{'!:';s- - i' :"'1('*s ;''_s s^ i'#:ss-_-' - #/s;rli ?_* 'f 'f'r ,'_1'4'1andSlidef ' 4S: !:^x ;;*' ; ' '-- *"' 'L;s' { 'Fig. 3.*}Landslide map of Koshi (providetl b .' NIED and JST)indicated in the figure.The ¥vest side of the map, inciuding Mushigame, ¥¥'hichis classified in the Asahi River basin (Fig. 4), is geological-ly an Ar'aya deposit that is massive dark grey mudstone.Shifting tolvard the east side of the map, ¥vhich isclassified in the Imo River basin (Fig. 4), the depositchanges to alternation of sandstone and mudstone namedKawaguchi and Wanatsu deposits. Alternation of sandstone and mudstone is distributed mainly along theImo River, except in its upper course. As sho¥vn in Fi_ . 4,numerous landslides occurred during the earthquake inareas ¥vith alternated layers of sandstone and mudstonecompared with massive mudstone deposits. This findingen_ :enders the conclusion that sandy natural slopes aremore fragile than clayey. natural slopes during earthquakes. On the other hand, Iandslide-designated areasare distributed mainly in the massive mudstone deposits,indicating that the failure mechanisms of landslidesinduced by earthquakes differ from those induced bysno¥v-melt ¥vaters.Moreover, the notable *'eolo_ ical features of thisre_ :ion are syncline and anticline structures. They form acomplex topography in ¥vhich synclinal axes and anticlinal axes are arranged ¥vith a short interval (Fig. 4). Forthat reason a peculiarly cuesta topography appears in thisle*'ion. Fra*"ile and weak slopes are therefore easily NATURAL SLOPE FAILURES¥Y"A731ATY.>.! EVs・lrn RiverFauure durin"' Higashi)!ama anticline axiS;'_'*+,.<=*'lMassiVe mudstone;'pfAlternationofsandstone 'fe ' ;'/f:! I>*I A ' 'deposit# itiJ'i,:.' ' '/p'- ;i iii' #{{;'-<'Asahi River/"'S-;1:c1 -a r'l"; *-;","/I{;";:>'' ip¥j/'s/( :: ";:;:; / )'//**;'f*"' ;"To¥vcT_e anticlineaxiSticune axiS.i fy-> 'f';..iij:/::(:;r_':;::;:/';:: {;S --;"-';ai*tiL# SS' "{',hi"; i c?: '4" / " ' ' i '":-l _;j. ' '':"''=' ''""' 's "-" ' " ' '<' "' ';"/;;;iir:':,:"'= j' i '/!.,.¥"- , =_dKajlkanesyncllne-__'' S>-"/1"'"_'¥""' axls;;.- . _ _; '- ' " ,;i- ';';: H;・aas['1 ta[(eza¥ a _ jiI' s' ' lc,s's' :'' ' Ia ldsllde={1) '.r- =" =' Kon"ratsukura': s ;"' *. _, anticline axiS/:J'/'= i>._ s!lCross section cL' er'. r;I oooLandslide distribution and geological map of Koshiformed. In addition, the river scrapes a¥vay the groundsurface and the slope toe becomes unstable. Thosereasons explain the numerous landslides that occurredduring this earthquake.Figure 5 sho vs a _ *eological cross section of the areasouth of forrner Yamakoshi villa,_,_ae (Junidaira) publishedby Geological Survey of' Japan. The location of this sam-pled geological area is shown in Fig. 4. Sandstone (W)and sandy mudstone (S, Ku2) are distributed ¥videlyaround the Imo River. This is a representative foldedmountains area composed of syncline and anticline.Geological cross sections clarify that a dip slope appears-500scale (m )Fig. 4.lB O)in the left bank of the Imo River and a reverse-dip slopeon the ri ht bank of the river.Photo i sho vs landslides that caused clogging of theImo River. Terano is located southwest of Tanesuhara.The natural slope moved at the left bank of the Imo River(Photo 1(a)). That about 360-m-long slide ¥vas 200 m and290 m vide in its upper and lower parts, respectively. Theinclination before failure vas about 17'. The depth isabout 20 m in the deep part, although it is not clearlydiscer'nible. Soft sandy soil, ¥'hich ¥vas extracted forphysical tests, vas videly distributed in the main scarp.Temporary drainage measures against river clogging were _i_;' _.; =_;'::r: ,;;;:i; , jTOYOTA ET AL.732:*(a) Terano (Lefi bank)(c) Terano (Right bank)(b) Upper part of the slope* Ori**in l't .・_olY iE or nle 11 t;r _ '.'S(d) Nampal1' ;s. .1;'i' -'"' ;' 'i/-"' ;: '- < s -' i="'t .."*j'/i : fi;= ;':' ";: "il; F;*,i":;';';,:: ::i::::?< ,'-'. _.+*f'."'** ..:- ''fr +(e) ) !Taranoki(D Hi_O_ashi-takezawa(h) Temporary watercourse(j) Junidaira- _'_* :+ S; '- ;,f i;:; "' "'_;; ' ;i;/' /::.: .-'..i=}'"'s "':: Ii: i- :'i''; i*i_ :ti' ._;- }::: i:;;;!';. ;;' ;: ;:: _: '__f ';:;: ;i' :!;,:::';i -'; :f! '***-"' '_ ;!> L';"";: {r_i"':i *- '* ' '" i-';; il:" - ' ' - .:t"='"'-'-- - ;'' *_i '*""'' :1 t ;: T'+ : {: ;f:; _ - ' - ;_ -' "i*/**' ;"* :''. ** -- *- ;: ' 'r/' ";':{ ' "__;i't:* * . '" ; .i ;;.-.*.:s;[ *; ; i'/;;._"**" '***_; ';;{::{i' {:7}::;;*=;:{'{i'- ;Srrrooth slltstone(_)c'Photo 1.L Trdslides clog"*ed the Imo Rivermain shock and became weak. Other salient featuresKa.i jgane slcline BKomatsu'*.)ura an icline+;imO River, lZ:t: w-fIWj Ui; 24sNA28Rihr b nk ( dip siope){reverse-dip s ope)Eandstencen Y mudstenelass I e mvdbecause of surface failures.Nampei is about I km downstream from Terano. Thesteep slope is about 35' and a fe¥v hundred meters ¥vide;it ¥vas collapsed at both banks of the Imo River (PhotoKU2¥ Leri bankinclude several surface failures that occurred in the ri**htbank, ¥vhich is a reverse-dip slope (Phot,o 1(c)). However,the amount of sediment that flo ved into the river is smallonenndl mt dsione interbed edlv th nndst( ne1(d)). Urgent measures against river clogging ¥vereunnecessary because of surface failure of only a fe¥vmeters' depth.A slope of about 30', 4,_O m ¥vide and about 200 mvil]age pub ished b) National Insritute of Advanced IntiustriaiScience and Technoiogy (AIST)long had collapsed at the right bank of the Imo River inNaranoki (Photo 1(e)). Although the depth of this slide islarg'er than that in Nampai, it would be classified as asurface failure. The soil that appeared on the bedrock inconducted Immediatelv. to prevent failure of the naturaldam. The slip mass corresponds to an old landslide tracethe failure area ¥vas soft sandy soil. Sampling ¥vas carriedout for physical and shear tests. The properties of the soilsho¥vn in Fig. 3; small valleys are present on both sides ofsamples are discussed iater.the failure mass. Groundwater is apparently abundant inThe soil mass, which is 220 m and 340 m wide in upperand lo¥ver parts, respectively, and 340 m long, movedabout 70 m during the earthquake at Higashi-takeza¥va,Frg 5. Geological cross section of the south of former Yamakoshithis area: ponds are visible in the upper part of the slope(Photo 1(b)). One ¥vitness insisted that this landslide hadoccurred not during the main shock but rather during theaftershock. The .'round ¥vas probably disturbed by thewhich is at the confluence of the Imo and Maesa¥vaRivers. Photo 1(O sho¥vs the main scarp of the landslide. :::! -]i !:::J: 1rNA'TURAL SLOPE FAILURESThe main scarp sediment ¥vas poorly graded sand,-r)r]/ OOvhichlandslide trace in Fig. 3. Small valleys exist on both sides¥vas extracted for physical tests. At the failur'e point, theof' the failure rnass, just as in Terano's case. A temporaryphoto sho¥vs that *・round preparations, as a measuredrainage channel is constr'ucted ¥vith installation of adrainage pump to prevent the natural clam from seepagefailure or erosion caused by overflo¥v (Photo 1(h)).Slope failures occurred on both banks of the Imo Ri¥'eragainst river clogging, had already been conducted. Thefailure slope is the left bank of the lrno River; it has agentle 15' ridge, ¥vhich is about 21 ' inslde of the slidein the Junidaira area, but they vere not large. Photo l(i)sho vs a landslide at the left bank of the dip slope. Theslope ¥vas gentle: Iess than '_O'. It seems to be a deeperSrnooth mudstone, ¥vhose colour is subdued blue, appears in the lo¥ver part of the main scarp (Photo 1(g)).The smooth mudstone is inclined about 18'. Theslide than a mere surf'ace failure because a¥'alanche-mudstone sarnple ¥vas also acquired to obtain physicalproperties. The mass moveurent corresponds vith an oldmitigating facilities and trees are mo¥*ed just as they are.This slope is continuing from Higashi-takeza ¥'a;1 ootopographical inf'ormation suggests the possibillty that alarge landslide like Higashi-takeza¥va might occur.Peculiar slope failures occurred at the right and leftbanks of the Imo Rlver. Sul"face failures occurred f'requently at the right bank having reverse-dip slope strata,o-80e)*.*1)>Hi ashi*taker.elba S' ilts[ontf 60(.p*=7cm s L 406 ! 1)¥vhereas problematical large landslides occurred oc-;Q'casionaliy at the left bank, ¥vhich is a dip gentle slope.Figure 6 sho vs the grain size distribution arising fromthe failure area along the Imo River. All samples f'rom themain scarps are sand including less than 100/0 fines. This40-Q- Terano, p==2 6S3)-O- Nar ;nLI E p>=,); -O!tn *653 g!cnl*-A- H i ashi-lakczaLnmeans that sandy natural slopes are more fragile thanp.=2 681 g;cm*oO.OI O Io.oo lclayey natural slopes during earthquakes. The denslty ofioNaranoki soil particles is slightly less than those of other(i*rain siz.e (rum)locations. The exposed smooth mudstone at Hlgashitakeza va is cornposed mainly of silt; that is, siltstone.These results agree ¥vith information from the geologicalFrg 6. Grain size distribution obtained from fai ure areas along theImo Riveriivit' ',StlT: ;/i/;ji:f:, T:::+!1 ::r';! :rs;Fal:i:;::::r; cj r:::/::{:¥Ii;i"i, t' *.;,i;1l_¥ -""l)!': :i1//J::it:::;1 ' ='t;i ;: t' j!:1:( J//Lt:i/4/! ;:jS:/i i{ :; ';j;cL EEI ":-Iff':: ;1(: ( j:," J' *'ts';;i:::tgtt?:I. f .'Aota"1'__ _'_* It tE"' --: -';: -:'!i: ;L -'!_' i '1'j/ }k .t I'. ,__ ' 'j_- ; ): '' c;/ill"/17 /_ '!'-"'i;1!' _'"=;-- -i_* : -¥ -i<.!';'-"I+'ti:'l'F 1 )l ! _i"_" _!;__7''1 I e;ll='r =r;;f ' ; ;1!('!i'E i_: .::a'm]ondaira. t _ i tunnel '' )}:It. (':*t:*i"' tt; ;/=+ ¥:1j;]' 'i.,:Lt ", -* ' -t: t''¥: " (i'(:L'_ _eil:;) i?lc!(;;:::::;f/:/:li:;;f: :J"!__f 1'::I :Ii le; i ::)¥:¥1' Y1:::C¥. . ,::'. .'= 'ii ",,;!:1 ' - i '¥1:_j tlj; :::¥/;;.'ir .-i:i!;:::;i:!;:i:iii:::1 {i/!'! ;i'i: i :_/ ' TT :i¥ : h"Lk'_: :/ f ;:i'l!::-__ - Measu InO Polnt Ij;;e '.'___'_;<_j fS '- rl!' 't:I }v j :¥: ! : ¥¥ *" -*. /' ';: - :;;' i :;';'"J i;/¥* J・ _ -;i_-'i :i- '. ril'- ; , ;ii:._";;.r li7ii;! f'¥--::'i: _ :_ :j_iji..・:.s'l_:_ :_; iii-'(a'7'iJ'_"'=' :". .1,t! '11+ rl-'":l ' '--'::]i; : ':: 1*': l-"* ']I :; 'IlL'":";: !T /:4:',/ii:!:i:!)::: ii :i/'!:"Ili !:i7L ( r 'f ' (i- '(' : ' l ' /jr :'-t -::i'T:::; ;// iT:; ;'J : ' :'!':- :1;: ::; ' _]P'"='ii* !'v*-= ** # oh 'EII1'!Js=i-- -' 'tj 'i j'; J,'}'r : /!r* '}r :r_ }:'r' 's'JJ¥' -:O+:'-Ii::'i c:i 11_--- '** =1:'*¥L'* e r': -:r'-_ =1$' - ''Jl'Tr : " - '-=: .,.1' "i::1" :;:;::'¥I_ ""']_" :- ;: 1::jShiotani River]' * * jJ I * "-+ 'l ir F J;;;Is ) a -¥' '''t_'ts! *'i="=-IGovomi-'-- '+"I(/nl'l't''Rlver 't' !'i'/-tl:! 'l-F (_ '_ 'jcLi '( ';--' i"/ l jPl{ij'r'ts _i I ;:' * r' i"Jias]:' ' - :L :1' {!' '- ¥;; --1i' e :'i) !-;f !i:'ir1!;! _-" ¥ !/- -;fi- ¥ :: :. ;. i ' ':, :::;:'1:':;!':':: : i-__ i-L_ I'・;i¥' '-'1;i I "- Ltii,ii] :r=J!'/ --i ."'1 i ':_ . IJ. :.' ':,, ',j;'!!1::!* + * :t:,++::TI+__;'I; ::; /':::!":II *;f*};:-"' !e'jL 'ii!: * -"'_I "_- :'i''/(a) Miyauchi ofMyoko cityFig. 7.-'/ ¥)i"" ; ] -:;;; 1= : '-': ' ' :' 1 : :f i': ::: 1 $t'l ll;' ¥ e ;:"'easulln_'_a'_ polnt{ ; ! { _f"::1'_-;'s4 lilrjl Joshinetsu motcu Fva¥Ar:t'fl¥ 4 ' 'tt '¥'" : -_: 'IJ,i' : {, {. ]_ ?__¥ i¥T: : ; )i _'T'-'S - S : 1;'.:!' s){ A A7!:r : !::" _' a)!¥r ' _-: lHonj** +* **,.l!s p/1;;t...-_ f" :"':'1__ ¥. /il-__. !-;::?:: i '::i ;i':"i :'?' -'_1'::''!]'----:'.・i _;';;:::jij):: =::':::'Q::::: i;:::'- ii ' 1, . ';'{tt;¥t{;ilhi^ 'il:'f::;{j':r l t' :'_: 7 tl(i ::'1Lti:hl-ilr_'- "¥ A' : ¥j::/c¥ '-'¥. /-^ Y' 'j_' *' ' _' ' f1:e=Li } : :-=F:1i :':":':1'[iI'fr:,,!.t;'// ;f2l o _'l : " ' '' 'i fitl: :::1' ; i/;;ji-'-'-/_:'i' -..1',' if!'_' *:' * *; / * t[ ti ;':-:):;'1' j!':i/" *:(c-* ' _- *F ''//:T:"i;;' '1i7:'1-i _ I i j-:'i::(::/(((. ¥ h* "t$l'l:"' '<1'/:e;':"=_ /*'1*'i_1:' ;' lt:::"I ! I ¥i( :;' / !_ ' 'tL_)'t:_"'f -" ::'f?: :!:( !i'! /] {;_A _'¥/_';:.i':.. ,''** =:"*"'*r { ; ll_:::'-" !s¥ ':: ' : ''- - - ll';' ' = '' ; t;;<''/ '' -'-*-** ' /" Til' '; -l: ' r'--f - *_ i¥nyauchi' ' I 1-: -!:I:---=+;d* t::i* * ::* i;Q/;*'/ " * 'i-+; '+1:, :Ttf :i' (Li:L;I nsfll o:: {r:d ' ,1"i-'- . 38 t__'-_/I : ::'! !_ L_ ,_..___1 _ ' :;:1-'*((:" '" I:'' " i' "t1: "t:"+it J 't" - 'lirl:t'_/;*""+ " jl'1 " l: ';; r:T 4//_ ' i_ - ii :i' -' _ . II+.'1 !'fi/' ://*' "'1 '*: ¥'! *'*':;: ¥¥'vc ///;'1'h: v:'!;=!¥'j_ !^ ;;'I ':'r.+?,:::iC::_K:;__'- *t ! ' ' '*'=' {:- '*:*1';'f' J't":];' e;;i ::-i'1"sv¥ L e/ __¥¥'tb(b) Irishio_ a¥va ofTochio cityIopographical map around t le pore lvater pressure measuring poi lts (provided by Geographical Surve) Institute) =TOYOTA ET AL,.734)eptil Soli tlT)uSoil tvpeDtptllfeasurement¥1easVrenlento liofP¥Peasurementof P ¥1' Pof IYater lel'elSih' m¥¥ieasurementO mSils? nl1_rni, 9 111Soil Qbtained¥Yeathe[ed4 TllmudstoneSoi} oblinedlrom the siteStrainer4 Iilrom 1 e si e6 O m6 1116{SilYirll '*ral'els nS IlSand} r veliIAlterTla iOil ofincl udinl O'i3emoniteI O mI O Illocoom udsto le aild:cobb] essandstone;entonitc1? nll I O I11O nl-il lllSiltstoneSand¥Ynter pressureJi4 m14tml stit cer7 I1114 mlllSandTSand¥mudS One'o *" L(a) Miyauchi of Myoko cityFig. 8.(b) Irishioga¥va of Tochio cityDepth of sensors and soil profilesmap, ¥vhich sho¥vs alternation of sandstone and mud--100stone strata.5MEASURF.MENT OF WATER PRF.SSURF. IN-1 l¥. {lyauch . ¥. i} ko clt}-11^Ol;nder round 1( a)P(,r,: 1 uicr presSurc; ter ieYeiFigure 7 sho¥vs a topographical map around the measuring points at lvliyauchi of Myoko city (Fig. 7(a)) and-13:; *'*'* -i3 Oiignl -ke I Cl]uetstiS:(Inhquake (-i40lrishioga va of Tochio city (Fi**. 7(b)). Those locations inNiigata prefecture are sho¥vn in Fig. 1. The observationprotection ¥vorks such as draina_9:e system on the surfaceor a planting treatment ¥vere carried out. In lrishioga¥va,numerous infiltration wells ¥vere installed because it is anold landslide area (Toyota et al., 2006). Mass mo¥'ementsin both areas induced by the earthquake are not ob¥'iousbecause the slope deformation had not been measured;deformation was not confirmed by ¥'isual observation ofthose slopes.Figures 8(a) and 8(b) r'especti¥'ely sho¥v the depth ofground-installed ¥vater pressur'e sensors and their geolog-ic column of Miyauchi and lrishiogalva. A borehole ¥vith' O h ? I st-* tain S!loek)1,nd 23rd zslh2 th26th6- 3 7Ti ne. Octobe ' 1004 (day)2507i Takada (37Q 6 3' ¥. '. 13S' 14 S'points are inferred to be the upper parts of the landslides.The measuring point of Miyauchi is located next to theJoshinetsu motor¥vay. When the motor¥vay tunnel vasconstructed, the natural slope ¥vas cut and the slopes,-134-120Water level and pore ¥vater pressure have been measured by our laboratory since 1995 in old landslide areas.,*1 3>oPREVIOUS LANDSLIDE AREAS-13j¥'oE)(b)200Effect of Tvpheon N+0. 2315l OAccunl ulated rain F ! lHourlrainfal OOoN. ii ; ta-ken Chuclsu)E :rthc]uzikc (¥. i50in shOck)o:-o'Oth 2 1 sl"nd23rd ,sth? tll?6tho<Time. October 2004 (da}')Fig. 9. Records of water pressure and l]ourh.' rainfall during the earthquake at Mi,.'auchi of M .'oko cit .' NATURAL SLOPF FAILURES*6 )-ll.6 ;::Vnd r2round I altr :cYels;- l4 S>) -O3,Pors IYuter prcSSure/-lO O(a)OOcl*s o-9 cI ¥ IY'¥'*! ,*:-11 6 *,:t;: - en c!1ueisJ-6 f)rPorc IY ter prcssurc-ll 7)',,,j-llS:h-ll S*1 1Jr!-shlo_r'_'al a' i ochlo crt)- IUnder"rt)wld IYatcr c:(i-ll5(a)I-7 o>30-i I Oj f[rI・.; ti"'" -11T""I'^'t" C:I }:L1._('...'].,. "TOc}11C)T 35' -9 ()'i r'h LILlks f ain s c : } [-lOl lh'(Jlhll<t 'Ind '3rd 15lh2 Lil-12J)-jl?a-1 lth 14th ISth i6th 17th I lhTime. jul) 2004 (day)12thTime. October 2004 (day)100lEfcl of T piroo!1 ;o. 2iS() E(b)Accu nuiatei r 'n li]160rrd}floO;400e'40}lol:rlv r inth>)Hourlv rainfalr T¥. l]:1-ke 1 Ciluit uEur hqu ko9600Ochl04 ,,150rl l¥//,!,:-,ll'Olll80(b){o9thOthI200 L5020sln sh )c : tl・,Tt'2nd23 rd 2Time. Oclober(Lh 25th 2 6thoVo<<004 ( ay)Fig. 10. Records of water presswe and hourly rainfall during theearthquake at lrishiogawa of 'T'ocl]io cit)O'tl 13th i4th i lh 16thTime. July 200417th ISth19th 20thO(da ')Fig. 11. Records of vater pressure and hourh.・ rainfall during thehcav) rainfall disaster in Niigata on 13 Juiy 2004 at lrishiogawa of'I'ochio cit)a strainervas used for groundwater' Ievel rneasurements.For pore ¥vater pressure measurement, ¥'ater pressuretransducers in a borehole vere buried in sand. Theirupper and lo¥ver sides ¥ver'e sealed ¥vith bentonite, asshown in Fig. 8. The boreholes for pore ¥vater pressur'epressure dropped on 7-0 Oct. '_004 ¥vith low atmosphericpressure; it then increased gradually. Although the pore¥vater pressure reached a peak at the time of Niigata-kenChuetsu Earthquake, the pore ¥vater pressure increase ofabout '_O cm in the ¥vater head is too small to affect sloperneasur'ement ar'e bur'ied completely using soils obtainedfailure. The pore ¥vater pressure sensors might not befrorn the site.installed near the slip surface ¥vhere pore ¥vater pressureFigures 9 and 10 respecti¥'ely sholv the r'ecords ofvater pressur'e and hourly rainfall at Miyauchi andlrishioga¥va. The values of ¥vater pressure have beenwill change most during a certain event. It seems thatacquired using a data logger on the hour. This means thatthe ¥vater pressures ¥vere measured after 4 minutes fromthe main shock because the earthquake occurred at 17:56.The rainfall data of the nearest areas from our measuringpoints ¥vere selected f'rom among data provided by theJapan Meteorological Agency. The measuring points ofMiyauchi and lrishioga¥va are about 6 km and 4 km apartfrom the rainfall-recorded points, respectively. Rainfallt.here vas considerable rainf'all on the night before theearthquake at N,Iiyauchi (Fig. 9(b)).As additional inforrnation for comparison, the recordsof lrishioga va during the heavy rainfall disaster inNiigata on 13 July 2004 (Toyota et al., ,_006) are presented in Fig. I l. Unfortunately, data of Miyauchi durin_"*,this event vere unobtainable. The total amount of rain-fall reached 400 rnm and the groundwater level rose noless than 10m. Never'theless, the pore water pr'essureaccumulated rainfall reached 100 mm in those areasincreased only dozens of centimetr'es in the ¥vater head.Those results underscore the possibility that pore vaterpressures ¥vere measured at strata vhere excess pore ¥vater(Figs. 9(b) and 10(b)). The ground¥vater level rose quickiypressure is not generated sensiti¥'ely.occurred by typhoon No. 23 of 20 Oct. 2004. Thewith the rainfall: about 2 m in Miyauchi and 4 m inlrishiogabva. The differ'ence in the rise of groundwaterbet¥veen Miyauchi and lrishioga¥va is considered to resultfrom their geological and topographical features. Theseareas are inferred to be catchment areas because elevatedunder*"round water levels descend gradually over a fewSOIL PROPERTIES OBTAINED FROM A NATURALSLOPE FAILURE SITESoil properties obtained from the slope failure atNaranoki (Photo 1(e)) Ivere examined in detail. Blockdays. The rise in groundwater table induced by thetyphoon had dropped by about I m in Miyauchi and 3 msamples lver'e obtained from the site of the failure (Photoin lrishioga¥va by the time of the Nii**ata-ken ChuetsuEarthquake. Similar behaviour ¥vas observed concerningthe pore water pressure at both areas: the pore waterdirect (box) shear tests in the laborator'y (Photo '_(b)).2(a)) and trimmed to make specimens for triaxial andThe specimens ¥vere extracted laterally from the soilblock. Therefore, the direction of compression in triaxial TOYOTA ET AL.736600(a) str'ess-strainl¥Taranoki sand ( Undisturbed)l ria500ial co np,essionDen eCDlest (p '*const )1- - -400r,'=200Pie ivm dense. ,r,=0 672F'=200 kPa- t' =C 83*.*oop'* { OO200P . , =0.67 1L 100 kPa. e -0.863* - -/1 oo-------p s(, };,; G7-----1L¥Ji i kz :iloO96C< = 2/3 ( j2i5) (9・b)-12¥.r+ -9cr)les_.,_¥ ';. I ;;aranoki sand (Undis turbed)rll,p'=const )- - - - hledlum den^ -c-6p )OU kP . (!==0.672, ,* .''*¥ p'=50 kPz . L'L=0.7]?,?) *p'50 kP*・ **' O 838_ /7i;]: -- --r. "I _-e;+*TlhDIL:ndiE:m dL;'n^ -ep'=1 OO kPs_ C *{).67 l:*. .tb) L: - s;;Trieixial compressionrp**s o・r・ ・'[';>I OV kPs. e*=O S63p 100 kPa. *'=={, S3i*3oq 1263Pl]oto 2. Soil samp]ing at the N*aranoki failure area: (a) sampling15g ) (""). = 2/3 (point and (b) trimming for triaxial tests50O(c} Failure line'aranoki sand (LJndisturbed)tests becomes a horizontal plane in the ground. The soil is400very soft and can be easily broken and shaped by hand.The grain size distribution of the soil is sho¥vn in Fig. 6.Specimens lvere d= 50 mm in diameter and /7 = 100 mm inheight in triaxial tests, and d=60 mm and /7 =20 mm indirect (box) shear tests. The specimens ¥vere saturatedusing the vacuum saturation procedure. Drained triaxialcompression tests ¥vere performed under constant p' con-Tri ¥iul compres'sion3 * sin(・ :*. 300' 200bO n)en<e{c)*=48l OO6cosLtd c*otests ¥¥*ere conducted under constant pressure ¥vith shearrates of 0.02 mm/min. There ¥vere t¥vo types of density inapplication because it contains about 100/0 fines.Figures 17_ and 13 respectively sho¥v the shear behaviour of undisturbed specimens in triaxial and direct sheartests. Direct shear tests ¥vere performed only in mediumdense specimens. In the case of dense sand, peaks areclearly apparent in stress-strain relationships (Fig. 12(a))and dilative behaviours occur, as sho¥vn in Fi_ . 12(b). Inc llkPa,!ed Um densetc'.=") 9c O 7 kPa)3 * sinc,do50 1 OOl _p'= ((Tj + 2(nearby similar layers: one ¥vas dense (e=0.69) and theminimum void ratios based on Japanese Industrial Standards (JIS A 1'_24) ¥vere 1.207 and O.676, respecti¥'ely.It is noteworthy that the soil is beyond the standard'siI,Itereeptditions ¥vith axial strain rates of 0.020/0 /min; direct shearother ¥vas medium dense (e=0.85). The maximum and6sin(,CDtest (p'*const. )O'OO 250oo・" 3 (kPa)Fia. 12. Shear behaviour of samples obtained from Naranoki intriaxial compression testsdense sands agree roughly ¥vith those of medium denseones. This indicates that dense and medium dense sandspossess mutually similar physical properties. With regardto results of direct shear tests, behaviour resembles thatof medium dense sand of triaxial tests (Figs. 13(a) and1 3 (b)) .Figures 12(c) and 13(c) sho¥v failure lines obtainedcontrast, peaks of medium dense sand are indistinctrespectively from triaxial and direct shear tests. The dense(Fig. 12(a)) and contractive beha¥'iours appear at thesand has a much larger angle of shear resistance, cd, andsmall shear-strain re*"ion of large confining-stress (p' =cohesion, c', than the medium dense sand (Fig. 12(c)).Regarding the medium dense sand, although the angle ofshear resistance, cd, that is obtained from the triaxial200 kPa) case (Fig. 12(b)). The void ratios and shearbehaviours indicate that the dense Naranoki sand isextremely dense. At the ultimate state after the peak inthe stress-strain relationships, the ultimate stren*'ths oftests is greater than that from direct shear tests, cohesion,c', from the triaxial tests is less than that from direct NATURAI SLOPE FAILURES200¥-7O[l OOaranoki saJr}d (¥ edium dense)(a }Direct : "hear tes t (Constani pressure)=-_ 150* 80! Dr) ing (dell} dratioo)Undis urbed. c .=2eO kPa, e ={) 8 9*00L; ldis tlrbcQl. c;:* DO kP60t' O S97pUndisturbcd. el*=50 kPa. i.=0 S9450,, 40)fo1_o4 6o,loolO18Matric suction. s (kPa}Shear displacement. 5 (mm)Fia. 14. Soil-water characteristic curve of sample obtained f rom-o 6Na!anoki sand (lvlcdium dcnsc)Naranoki(b)Direct shear test (Constant pressure)-O.4500N.. -0.2Und]stur ed cs: jOkPa, t.=0,S9;400e;= O S45 - O.SS_1 . p=constantSv- O Oaranoki sand (Reconstituted)Tri a ¥ ial compressionl300¥=50 kPa:;:u6sin(t*3 - sine'*1Drained (Constant suction)Undis Vrbed. c:. = OO kPa. d =0 897O.2Undismrb d cF)> 044 6oShear displacement, ()b100 kPa.,:=0.S49800N_ Iaranoki s nd (iv edium dense)39 O t. c'*(] kPa)O kPaSa ura ed{ ooo(c)50p'.Direct shear test (Consiant Pressure)_ 120i OO 1 50 20050300p***1 = ((TI + 2( ;)/J) l! (kPa)c!J-!'5 IQF'io. 15.:2* 90/'60F'ailure lines of unsaturatedmedium dense Naranoki sandof 100/0, was placed in a 5-cm-diameter mould andtamped carefully at I cm layers using a 2-cm-diameterrod to control the specimen density. The drying portion;e,(/) 30of the soil- vater characteristic cur¥'e obtained from thecd=6i: kPao=i *(mm)ol 50{o50 1 50l OOVertical stress, c:200(kPa)Fig. 13. Shear behaviour of samples obtained from Naranoki in directshear testspressure plate method is sho¥vn in Fig. 14. The experiment was carried out under net rnean principal stress,p *t= 100 kPa, using triaxial apparatus. This r'esult suggests that the air entry value of the soil is approximately5-lOkPa. The degree of saturation dramaticallydecreases at the air entry value because of the sandy soilspecimen. Therefore, when the soil is under stable condi-shear tests. Those differences pertain because the strengthparameters, cd and c' , vary subtly ¥vith even small scatter-tions, the soil is in either fully saturated or in a low degreeing data. The cohesions are very small in both tests.Therefore the cementation and ageing effects of mediumShear properties under an unsaturated state were alsoexamined, where suction was controlled by the pressureplate method using a ceramic disc. Drained (constantsuction, s) tr'iaxial compression tests ¥vere performeddense Naranoki sand layer are small. Those results implythat failure vould occur in the medium dense sand layerbecause the strength of the dense Nar'anoki sand is muchgreater than that of medium dense sand.The medium dense layer is considered to be moreimportant for failure than the dense one. Reconstitutedspecimens were prepared as medium dense (e=0.85)using the moist tamping (MT) method to elucidate theunsaturated soil shear behaviour of medium dense specimens. By MT method, the soil, ¥vhich has a ¥vater contentl+-of saturation.under constant p' conditions, vith axial strain rates of'0.010/0/min using reconstituted medium dense Naranokisand. Figur'e 15 sho¥vs f'ailure lines of' the unsaturatedspecimens. Under saturated state, strength pararneters ofreconstituted specimens, cd = 39.0' and c' = O kPa shownin Fig. 15, are almost the same as those of undisturbedspecimens, cd= 39.9' and c' =0.7 kPa in Fig. 12. Reconstituted specimens having similar strength ¥vith that of' lTOYOTA ET AL73sundisturbed ones could be reproduced by usin_ : MTMiyauchi and 4 m in lrishioga¥va, ¥vhich are oldmethod to make the specimens the same void ratio ¥vithundisturbed ones. Failure lines of different suctions arelandslide area, by typhoon No. '-3. But the rise inapproximately parallel to that of saturated soil. Cohesionis greater ¥vhen the matric suction is larger. Although theMiyauchi and 3 m in lrishioga¥va by the time of thecohesion changes to some degree bet¥veen saturat,ed ands= iO kPa, small differences exist between s= 10 kPa ands= 50 kPa because the tested conditions (s= 10 kPa and5. Very dense and medium dense sand layers weres=50kPa) are a part of the soil-water characteristiccurve ¥vhere little difference exists in the degree of satura-tion vith suction. The result sho¥vs that the stren9:thground¥vater table had dropped by about I m inNiigata-ken Chuetsu Earthquake.involved in the Naranoki slope failure. The verydense sand had much larger strength than that of themedium dense sand, although the ultimate strengthsof both sands ¥vere similar. Therefore it is supposedthat the failure ¥vas triggered in the medium denseincrease induced by matric suction cannot be expectedenou*'h. Concerning the external force, the slope at6. The air entry value of the medium dense NaranokiNaranoki is steep, which is about 30', and is located nearthe epicentre. It is considered that large shear stress,cohesion of sand, induced by dryin a, js no mor'e than¥vhich consists of static shear stress by the slope anddynamic shear stress induced by the inertia force, ¥vasacting in the slope during the earthquake. This slope issupposed to be fragile and unstable during the earthquake because the soil possesses only small cohesion inspite of the large external force induced by the slope andthe seismic motions. For example, for this estimation,Tsujioka (2006) performed stability analysis based on aslice method ¥vhere moisture and seismic force effects onsoils were considered.c_ONCl,USIOr 'SThe Niigata-ken Chuetsu Earthquake caused heavysand layer'.sand is approximately 5-10kPa. The apparentabout 10kPa; furthermore, the cementation andageing effects of Naranoki sand layer' are small. Fromthese facts, the Naranoki slope is inferred to befragile during an ear'thquake.ACKNOWLF,DGEMF,NTSWe investi_ :ated the stricken area ¥vith many research-ers and obtained useful information. Soil tests ¥vereconducted ¥vith the assistance of students from theCjeotechnical Engineering L,abor'atory, the Departmentof Civil and Environmental Engineering, NagaokaUniversity of Technolo_g:y. The authors express sincereappreciation for their contributions to this study.damage to infrastructure in hilly and mountainous areas.Soil masses of landslides clog*・ed river channels andproduced natural dams. Effects of geo*・raphical andgeological features, such as foldin*・, ¥vere apparent. Thematter' of naturai dams de¥'eloped into a soclal problembecause to¥vns ¥vere submerg:ed and all residents offormer Yamakoshi village fled their homes. Moreover,residents of do¥vnstream areas ¥vere also exposed todanger of debris flo¥v caused by t,he collapse of naturaldams. A detailed investigation was carried out on landslides along the Imo River to clarify the failure mechanisms.Main findings from the study are summarised as:i . L,ar_ e-scale landslides during the Niigata-kenRF,FERF,NCF,Sl) Editorial Committee of Histor¥' of Yamakoshi village (1981):Hisrory of Yanlakoshi Vi!lage (in Japanese).2) Geographical Survey Institute (2004): Disas!er' * ifap of!¥rii*"a!a-kenClu!etsu Earrhquake (http://zgale gsi go^jp/niigatajishin/indexhtm).3) Japan Society of Civil Engineers (2006): Reporr on !he 2004 Nii**"ataCllue!su Earthquake, JSCE, CD Rom (in Japanese).4) National Researchnstitute for Earth Science and Disas er Pre¥*en-tion (NIED) and Japan Scienc,e and Technology Agency (JST)(2004): Lands!icle il[ap (http://kslvebl.ess-bosai go.jp/jisuberi/jisuberi_mini/jisuberi_top,htm}).5) N{igata Branch of the Japan Landslide Society ()-003): Recorc! ofLands!ide Disa,sfers in _'¥rji*"afa, C_D Rom (in Japanese).Chuetsu Earthquake are re*・arded as re-slide typelandslides because they occurred on old landslide6) Niigata prefecture (1982): Con.servarion Itrfap (in Japanese)7) Takeuchi, K^ and Kato. S. (1994): 1,50.000 Geo!ogiccl! ! lap ofllletraces.8) Takeuchi, K , Yanagisa¥va, Y., iiyazaki, J and Ozaki, h'l ('_004):I:50.000 Di*"ita! Geo!ogica! I fap of the Uon,lma Region, .,¥rjigata2. The greater part of slope failures during the earth-quake occurred at alternating sandstone and mudstone strata, but landslides, ¥vhich moved graduallyduring sno¥v melting season, have usually occurred atmassive mudstone of quaternary and tertiarydeposits.Takada-Tobu Region, Geologicai Survey of Japan,Prefecrure (Ver. !), GSJ Ope -file Report, l¥To 412 (http://¥v vw.sj. jp/GDB/openfile/files/n0041)_/index.html).9) Toyota. H.. Nakamura, K. and Sakai. N. (2006): Evaluation ofdike and natural siope failure induced by heavy rainfall in Niiga aon 13 .Iuly 2004. Soi/s and Foln7c!ari0,Is, 46(1), 83-98, 2006,lO) Tsnjioka, T. (2006): Stren_ h e¥*aluation of unsaturated slopes3. During earthquakes, surface failures occurred easilyduring heavy rainfails and earthquakes. Master T/1esis. Nagaokaat steep reverse-dip slopes, ¥vhereas lar'ge landslidesoccurred at gentle dip slopes.ll) Yanagisa¥va, Y., Kobayashi, l., Takeuchi. K., Tateishi, h,l ,4. The ground¥vater table ¥vas raised by about 2 m inUniversity of Technolog.v (in Japanese).C_hihara. K and Kato. S (1986): J,50.000 Geolo*"ica! Map of !heOjh,a Re*"ion= CJeological Sur¥'ey of Japan
  • ログイン
  • タイトル
  • Damage to Earth Structures for National Highways by the 2004 Niigata-Ken Chuetsu Earthquake
  • 著者
  • Junichi Koseki・Tetsuya Sasaki・Nichiro Wada・Junichi Hida・Masaki Endo・Yukika Tsutsumi
  • 出版
  • soils and Foundations
  • ページ
  • 739〜750
  • 発行
  • 2006/12/15
  • 文書ID
  • 20959
  • 内容
  • fSOILS 4LND FOUNDATIONS¥rol46,No6.739-750, Dec. 2006Japanese Geotechnical SocieiyDAMAGE TO EARTH STRUCTURES FOR NATIONAL HIGHWAYSBY THE 2004 NIIGATA-KEN CHUETSU EARTHQUAKEJUNIC 'HI KosEKli , TETSUvA SASAKiii), NICHIRO WADAiii), JUNICHI HIDAi+),MASAKI ENDO ) and YUKIKA TSUTsUhll+i)ABSTRACTDamage to earth structures for national high vays by the 2004 Niigata-ken Chuetsu earthquake is reported. Afterovervie¥ving damage to roads, se¥'eral case histories on se¥'ere damage and rehabilitation ¥vork for national high¥¥'aysare documented. Immediately after the mainshock, national highways and ps'efectural roads ¥vere closed at 101 sites.Among them, main national hi**h¥vay routes Nos. 8, 17 and il6 were closed at 17 sites, ¥vhile all of them could bere-opened within ten days after the mainshock. On the other hand, national high¥vay route No. 291 that runs throughmountaneous areas sufi:ered from extensive dama*"e at t vo sites, ¥vhere alternative routes had to be constructed. Thenational highhvay route No. '-91 suffered also from damage to embankments constructed on valley and/or' Iandslideareas and damage to cantilever and gra¥'ity type retaining ¥¥'alls constructed in *'or*'e area. The damage to the nationalhighway route No. 17 included collapse of a gravity type retainin*' wall constructed on uncemented debris or veatheredterrace deposits and collapse of a segmental retaining ¥vall supportin*' a cut slope. The national high vay route No. 8suffered from embankment failures due possibly to liquefaction of a loose sand subsoil layer and/or' concentration oflongitudinal or transverse ground ¥vater flo ¥'.Ke .' words: earth structure, earthquake damage, liquefaction, national highlvay, rehabilitationNiigata-ken Chuetsu earthquake (IGC: E8)¥vork ,the '-004In this report, after overvie ving the darnage to roadscaused by this earthquake, several case histories on severeINTRODUCTIONAt 17:56 on Oct. 23, 2004, the mainshock of thedamage to earth str'uctures for national high¥vays areNiigata-ken Chuetsu earthquake with a magnitude specified by the Japan Meteorological Agency, MJ lA, of 6.8documented, focusing on the details of the damage,ground conditions and rehabilitation works.hit the central part of Niigata prefecture, Japan. It wasfollowed by three major aftershocks ¥vith MJ A equal toor larger than 6.0 t.hat occurred ¥vithin one hour after themainshock. In addition, several other subsequent after-OVERVIEW OF DAMAGE TO ROADSTable I surnmarizes the nurnber of sites ¥vhere thetraffic was suspended completely along national high vaysshocks continued for about two lveeks (JMA, 2004).By this earthquake, many types of infrastructuresfailed completely or ¥vere dama*'ed severely, in particularand prefectural roads (MLIT, 2005). Immediately afterthe mainshock, these roads were closed at 101 sites.natural slopes and earth structures. It ¥vas because theAmong them, national high¥vay routes Nos. 8, 17 and 1 16strong earthquake occurred in mountainous areas and astrong rainstorm by Typhoon No. 23 hit the affectedthat had been operated directly by Ministry of Land,Infrastructure and Tranport (MLIT) ¥vere closed at 17sites as sho¥vn in Fig. 1, while all of them could bere-opened after competing temporary rehabilitationregion immediately before the earthquake (Tatsuokaet al., 2006). The daily precipitation due to this typhoonrecorded on Oct. 20 (i,e., three days prior to themainshock) at Na*'aoka City that is located in the affected¥vorks within ten days after the mainshock.In the affected region, there also existed Kan-etsu andregion ¥vas as heavy as 100 mm (JMA, 2004).Hokuriku Express'ays operated by Japan High¥vayProf ssor. Insti ute of lnduslrial Science, University of 'Tokyo. Japan (koseki@,iis.u-tokyo.ac jp)ii} Senior Researcher, Ground Vibration Team, Pt blic Vorks Research lnstitute, Japan.iii} project hieasures Official, Toyama River and National Highlvay Office (formerly Director of Restoration Measures Division. Nagaoka,)Na ional High vay Ofiice), IMinistry of Land,nfrastructure and Transpor , ,Japan.*'} Deputy Director. Hokuriku Regionai Developmeu Bureau, IMlnistry of Land, Infraslructure and Transport, Japan.+) Director, Niigata National High vay Offlce (formerly Nagaoka National High¥vay Office), *¥,Iinistry of Land. Infrastructure and Transpon,Japan.' 'Technicai Staff. 11lstitute of Indus rial Science, University of Tokyo. Japan.The manuscrlpt for this paper ¥vas received for revie¥v on May lO, ._006; appro¥'ed on September 29, 2006.¥Vritten discussions on this paper should be submitled before July I , 2007 to the Japanese Geotechnicai Socie y, 4-38-2. Sengoku, Bunkyo-ku,,,Tokyo 112-0011, Japan. Upon request the closing date ma"v be extended one month.739- 調KOSEKI ET AL。7婆0τ段ble1、NumberofsiIes・轍htr面cs糠spensionalongna醸o鶏田勤lghwaysandprefeαu田lroads(MUT,2005)Ocτ、23,2004(immedl飢eiyafter由e BQ)NoV、3,2004Dec,27,2004July 15,2005Oct.28,2005(in10days)(i鷺2mo蹴t蝕s)(魚9mo鷺由s) (in玉year)0007NH No.891NH No.17NHNo.王160000000000由er NHs2421121010PrefeCtura亘6089483735nO604745TQta1101NH:natlo照1厩gぬway→畢・壷』Sitesclosed afterthe rnain shockNational highway      ∼・,   へ_、!、_、,       1 /一 、   Kan−etsu..乙Expresswayke㎜ Railwayノノメ   \   、 カR8、3『’一 City limitR8−2ノ_!hio        域、ド㊥.R8−4        ∼ 寿 R116−1 ・、閉「〆/       ’        ,、ノ競轡 )、・イ.、!.、』R&5/HokurlkuExp.ソ !1ン        \』ぐ二ダ油翼一  /     戸\山ノKashiwazaki一1詞舜、煮卜曇Ω壌レぐ。、ご /R窪7−8R8−6ズR8−7、距’識・△〔」二Epicenter・f’ノ、、∼シ渡…照論糊gcピ、i、             ド                                R17−7     ・;議曽….黛、唖彫u。numa        、  』} 一     、 、一 自P自呼   バ        閉〆 R17−6晒’ゴ〉』・、.・1自毒黛9「me「yHi「oka師  ぼ ,鷺ド’う1一脾P^ ㌔∼、ゴ騨/ゴ\Fi9,LLoc頗on of siIes where I田餓c w我s suspended immedi訓ely af窒er the mginshock段long臓郎ional hlghway rouIes NTos、8ラ17a臓d1藁6Corporation(currently East NipPon Expressway Compa−ny Limited).Althoug熱part oギthese express∼vays wasbefore the earthquake,its damage was so extenslve that aclosed for ins夢ection immediately after the malnshock,of10km on behalf of t}1e original operator.part ofthis routewas rehabilitated by MLIT for alengththey cou玉d be re−opened to emergency traf温c suc封 as&mbulances,負re trucks and police cars、vithin 19hoursafter the mainshock(JSCE,2006).NATIONAI、HIGHWAy ROUTE NO.291 Hereln,case histories on severe damage to earth struc− Location ofsites wit}1major damage to national high−tures along nationa歪hlgbway routes Nos。291,17an(i8way route No.291in Yam&koshi district(currently partthat were rehabilitated by MLIT are reported。1t shouldofOjiya City)is indicated in Fig.2.Refer毛o Fig.1for itsbe noted that,a王though the national highway route No。general location within the a狂ecte(i region.T}1is route291had been operated by the of駐ce of Niigata Prefectllreruns through mountalnous areas that were damaged se一 NATIONAL HIG H¥¥*AYS 2004 CHUETSU EARTHQUAKE741National highway route 291 rehabilitated by MLIT for length of 10 kmL _ '. i constructedA[ternat[ve route li Existing route reconstructed l/' ="'1lj::.:::.:::::',1//:::,___ _=_' "'"'"'1//::: , ::{ _ ; .,::::::'1,:;_ ; '< ::_To oijya i'i:.t yJ-J; i i_ _! . - LJ-ir:.:l i '"_ ;" : T"ii';r'":l!lQRW-1 -:.: =l ::-BR,1 2=,j;):,i";f+-ASahi nver .'...'.i'.'1"-" ::'= j" l; 'i.,_.iTi":j: Ta.kezawa! _.Hagurotunnel' Minami- ' ' =L, i-';+.* ¥,,!+,cee:+, , , RW-: ./ J _:::: ___imo riveKam'sawa riverPartly rehabilitatedby otfice of Niigata].I=- *・・・-"_._* i--・-・-==--*...-"_.__._*1ipref .l; l_ National highwayeprefecturai roadsiEIESE-Damage to sloperetaining walisiBR Damage to bridgesFig. 2.Location of sites/ Higashi'r takezawa*SE-1OSE-1 io i 'x'rL" '+!'To Uonurna:,, - : ¥ <SE-6and embankmentDamage toy'*-Jee .,*"SE-5R ver]Kaijgarle iSE-4i -S E-9*= ,city- Maesawa r'ver'! >'; fSE-8vith major damage to national ilighway route No. 291 in Yamakoshi districtverely by the earthquake.As sho¥vn in Photo 1, several slopes and road embankments facing Kamisawa River at Takeza¥va (site SE-1 inFig. ?-) failed extensi¥'ely. For rehabilitation ¥ 'ork, itvasdecided not to reconstruct this part of road for a length ofabout I km, but to excavate a ne¥v tunnel as analternative route in the mountains located on the otherside of the river. Such decisions ¥vere made based ontopographic maps and digital cross-sections that ¥verepr'epared in t¥vo weeks after the earthquake by usin__aaerial photo and laser sur'veys combined ¥vith the globalpositioning system.Since the above tunnel excavation requires rnor'e thanone year (tentatively scheduled until autumn of 2006), alocal road ¥vas employed as a temporary detour. Still, ittook about eight months to make the through tr'afficpossible, because part of' the local road had to bereinforced and reconstructed. It should be also notedlrrthat, due to heavy snolv fall, execution of such rehabilita-tion ¥vorks had to be paused during the ¥vinter seasonfrom Januar'y to April.In addition, at Higashi-takezawa (near the site BR-3 inFig. 2), a huge slope failure cr'eated a natural darn alongImo River (Tatsuoka et al., 2006), ¥vhich caused submer-gence of Shin-ugaji bridge at site BR-3 Iocated in thelh_Photo l.b:*o.Fa ureplanesSlope failures at 'Takezalva a]ong national tligtlWa) route291 (site SE-1) KOSEKI ET AL,,742Photo 2. Embankment failure at Kajignne aiong national higlnva .'route N. To. 291 (site SE-3): a) a:erial vie v and b) coilapsed soilsPhoto 3. Embankment failure at Kajigane along nationahigh,vayroute No. 291 (refer to Photo 2(a) for the direction to lvhich eachphoto lvas taken)'*.""' ="*='* =**=* '*'"=*'"'***" '*.***,**"*'*'* = .'-' 11[ i '""=''*/** *1+*--* * /l'' =1 ='+i J*=¥'=/_'' ' :])t '=- / lTilfl__:Illi! -+ ''= . );7l**= ' *= 'L¥ _l¥**'=**::';::::::::i:::;::::':::;:;iS 1:j;;;;i=j::i::<:i;;;j;i::i'-'i;i-;; ;;i:;-:;;;;' =;i; --'- "-'-- :-iLt ( _"_"'--''),.- = *'* ***=='***'*=*==++'erev,Td svce beFor2*"='=erouna Ye:{2tien s. 0=1¥^ :-; :"-i " }**i '. r"s-' $" =/ rlls' W __ ,: SS ;''i"eelegfcclsssi s tien SvmbStreF; ly Yfeaered mut,*'es' :=efBd rT7t;ePhoto 4. Aeria;s** rTCe:! pSed selY8!= ,,iC re nd stJrfas ? s *eri rll^ ' Bter",Sd f2! vr s_J !viewoflandsl desatKajignnealongnatiomahiginl'ayroute No. 291 (site SE-2)eSn ck **",rQc'(2b)ly e th retirllVdfe<:k *!**Stemporary reservoir. The slope failure ¥vas so extensivethat re-routing of the national hi**h¥vay at this site had toB iet*/AE O3-tefi yF.* Or fecensts** .! l:.・;*Is __ --D -; i:J :('----1' . /.- Qe dit(;'r.,/-t_r-i: / _,- - *'r** '*J/// 1/ :,..' ::_* // _----f "/_"I_i_--*' Q ---:r:L- l.*__t;___ t' r2'!,r3s h ef w :ef ow- /;d___ 'e c B i r- c)e li :l:L* !a*be implemented for its rehabilitation ¥vork.In the follo¥vin*・s, other case histories on severe damageto embankments and retaining ¥valls are reported. All ofthem ¥vere rehabilitated by reconstructing the existingroute.L*nit : r lmFlg 3. Original and rehabiiitated cross-sections of failed embankments anti estimated ground profilo at Kajigane along nationalDalllage to En7bankillents Consn'ucted on Val!ey andlo,'highlva)' route No. 291: a) cross*section at site SE-3, b) cross-Lanc!s!ide A reasAt Kajigane in Yamakoshi distr'ict (site SE-3 in Fig. '-),sec