Typhoon Rainfall and Landsliding in the Pacific Ocean Side of Japan

Ranjan Kumar Dahal1,2, Shuichi Hasegawa1, Minoru Yamanaka1, Astuko Nonomura1 1Dept. of Safety System Construction Engineering,

Faculty of Engineering, Kagawa University, Takamatsu, Kagawa, Japan

2Department of Geology, Tri-Chandra Multiple Campus, Tribhuvan University, Ghantaghar, Kathmandu, Nepal

ABSTRACT Japan is an island chain located off the north-western rim of the Pacific Ocean. Generally, in Japan, precipitation occurs mostly during typhoon seasons on the Pacific Ocean side, and in winter (heavy snow) on the Japan Sea side. This paper deals with synoptic descriptions of failures that occurred in various areas, along with rainfall and failure relationships during the typhoons of 2004 in Shikoku. In this paper, the effective amount of rainfall that triggered landslides in the Shikoku Island during these typhoons was identified along with information of occurrence time and landslide-triggering thresholds for Shikoku Island were interpreted. KEY WORDS: Typhoon; Shikoku, rainfall induced landslides; debris slides; debris flows; triggering thresholds. INTRODUCTION Experience shows that landslide occurrences on hill slopes have very close relationship with availability of water. As a result, many types of landslides occur after heavy rainfall in tropical and temperate climatic zones (Jakob and Weatherly, 2003). Landslides triggered by rainfall occur in most mountainous landscape of the world. They pose a significant natural hazard and they have a high damage potential. Many statistically meaningful analyses have been published to demonstrate threshold values of rainfall and landslide triggering (e.g., Caine, 1980; Wilson and Wieczorek, 1995; Crozier, 1999; Aleotti, 2004; Guzzetti et al., 2007). Within the last few years, many studies focusing on rainfall thresholds for triggering landslides with hydroclimatic condition, antecedent rainfall (Crozier, 1999; Glade et al., 2000), and hydraulic conductivity (Terlien, 1998) were conducted. In this context, this paper describes some scenarios of rainfall-triggered landslides that occurred in Shikoku Island, Japan (north western Pacific Ocean side) during various typhoon events in 2004. It is 225 km long and 50-150 km wide, with more than 80% of land consisting of steep mountain slopes. It is a heavily forested mountainous region of Japan. It has a few plain areas along the coastal lines and elevated peaks in the central part. There are only small villages on the mountains but the mountain bases are considerably populated. The mean annual precipitation of Shikoku ranges from 3,500 to 1,000 mm. Shikoku Island can be roughly divided into three geological zones: Ryoke, Sambagawa-Chichibu, and Shimanto belts from north to south (Fig. 1).

The three zones are bounded by two northerly dipping major faults, the Median Tectonic Line (MTL) and the Butsuzo Tectonic Line (BTL) from north to south, respectively. The Ryoke Belt consists of late Cretaceous granitic rocks, late Cretaceous sedimentary rocks (Izumi Group), and Miocene volcanic rocks (Sanuki Group). Cretaceous granite is widely distributed in the north of Seto Inland Sea. The MTL topographically marks a distinct sharp boundary for Shikoku Range. This range, with a maximum altitude of nearly 2,000 m, is occupied by the Sambagawa-Chichibu Belt. The Sambagawa Belt is composed of low-grade metamorphic rocks. The southern Chichibu Belt is mainly composed of carboniferous to Jurassic sedimentary rocks and low-grade metamorphic rocks. The Shimanto Belt consists of Cretaceous and Palaeogene sedimentary rocks and this belt occupies the two southern peninsulas protruding into the Pacific Ocean. The middle Miocene granitic and partially gabbroic rocks are sporadically distributed along the axes of Muroto and Ashizuri peninsulas. Owing to the geological and morphological settings, landslides and floods caused by typhoon rainfall are frequent in Shikoku. Also, in 2004, Shikoku experienced extreme events of typhoon rainfall and faced huge losses of life and property. In this paper, rainfall events due to typhoons of 2004 are critically evaluated and landslide events during the rains are also described. The main objectives of this paper are (a) to document the pattern of landslides in response to the typhoon rainfall events of 2004, (b) to derive the intensity-duration relation of rainfall that triggered the landslides during the typhoons of 2004, and (c) to determine the effect of typhoon rainfall on the different geological and physiographical terrain of Shikoku Island. TYPHOONS IN JAPAN AND SHIKOKU Typhoons are the most significant meteorological events for natural disasters on earth, which are caused by strong winds, heavy rains, river floods, storm surges, and high waves. It is noticed that the horizontal scale of a typhoon ranges from several 100 km to a few 1,000 km, while that of the cumulonimbus clouds is in the order of 10 km (Tsuboki, 2005). The heavy rain is usually localized in the eye wall and spiral rainbands. From 1951 to 2005, there were 1,468 typhoon events in the northern part of the Pacific Ocean, 163 of which hit the Japanese archipelago. In 2004, Japan was hit by 10 typhoons which are the maximum annual typhoon events within the last 55 years (Fig. 2) and the lowest number of typhoons occurs in February, whereas the highest number is in August (Fig. 3).

Proceedings of the Eighteenth (2008) International Offshore and Polar Engineering ConferenceVancouver, BC, Canada, July 6-11, 2008 Copyright © 2008 by The International Society of Offshore and Polar Engineers (ISOPE)ISBN 978-1-880653-70-8 (Set); ISBN 1-880653-68-0 (Set)

795

Fig. 1, Geological map of Shikoku Region (modified after Hasegawa and Saito, 1991), location of Shikoku in Japan is shown in inset.

01

2

3

45

6

78

9

10

1951

1955

1959

1963

1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

Years

No.

of T

ypho

on

Fig. 2, Number of typhoon events in Japan during last 55 years (source: Digital Typhoon, 2006 and JMA, 2005)

0

50

100

150

200

250

300

350

1 2 3 4 5 6 7 8 9 10 11 12

Months

Tota

l Eve

nts

Pacific Japan Shikoku

Fig. 3, Total typhoon events in North Pacific Ocean, Japan and Shikoku during last 55 years (source: Digital Typhoon, 2006 and JMA, 2005)

But, in Japan, the months of January, February, March, and December do not have any typhoon records and it is most abundant in months of August and September. Similarly, there were some records of typhoon in June and July to October. As seen in the isohyetal map of Shikoku, the island usually gets more extensive typhoon rainfall in its southern part than in its northern part (Fig. 4). The last 30-year rainfall pattern (Fig. 4) shows that mostly Kochi and Tokushima as well as the southern part of Ehime receive extreme rainfall. High-intensity rainfall is concentrated over most parts of Kochi and Tokushima prefectures. The elevated peaks in central Shikoku Island are the main cause of rainfall concentration in these two prefectures. Rainfall-related slope failure phenomena are very common in central and southern Shikoku. However, north Shikoku (Kagawa) has also been suffering from some events of failure. Fig. 5 illustrates the condition of rainfall-related disasters in four prefectures of Shikoku. In 2004, 10 typhoons hit the Japanese archipelago and nine of these affected Shikoku Island (Table 1). Storm and flood damages in 2004 resulted in a total of 227 persons killed and missing in the whole of Japan, which is the highest number since 1984 (JMA, 2005). Ehime, Kochi, and Kagawa prefectures were mostly affected by typhoons 23 and 21 whereas typhoons 4, 6, 10, 11, 15, 16, and 18 resulted in extensive damage and loss of lives in Kochi, Tokushima, and Ehime. The number of typhoon events and the damage outlines are given in Fig. 6. In 2004, rainfall intensity was abnormally greater than those of previous typhoon events. DATA COLLECTION In this paper, presenting data of all landslide events is almost impossible; however significant events were documented using sources such as field visits, government records, scientific publications, and technical notes. One interesting feature of the landslide data is good information about rainfall and timing of landslides for nearly 20 landslide events.

796

Fig. 4, Isohyetal map of Shikoku prepared from rainfall data of 1971 to 2000 (modified after MLIT, 2004)

Kochi

Ehime

Tokushima

Kagawa Landslides

Debris flow on streams

Steep slope failures

0500

10001500200025003000350040004500

Num

ber o

f Pla

ces

Fig. 5, Occurrences of rainfall related disasters in Shikoku (source: ISA, 2006)

564 461

1535

214 42

326

0 200 400 600 800

1000 1200 1400 1600 1800 2000

Debris Flow Deep Failures Shallow Failures

Japan Shikoku

Fig. 6, Number of landslide events in 2004 (source: ISA, 2006)

Therefore, such data were used to do a representative analysis of rainfall and failure relationship. Aerial photographs taken after the event, maps, and personal and public documentation were also considered to enable landslide inventory maps of some representative areas to illustrate the problem. In 2004, Shikoku Island faced one of the worst disaster events in the history of Japan. A scenario of occurrences of landslides in Shikoku Island after the 2004 typhoons is shown in Fig.7. Although nine typhoons affected the island, the more effective typhoons were only six (Typhoons No. 10, 11, 15, 16, 21, and 23).

Kisawa

Oogawa

Uwajima

Touon

Toyohama Niihama

Shikoku Chuo

Takamatsu

Moriyuki

Higashi KagawaMonnyu

Takamibo

132o 133o 134o

132o 133o 134o

34o

33o

34o

33o

Kagawa

Ehime

Kochi

Tokushima

Fig. 7, Landslides in four prefectures during typhoons of 2004 and damages are mainly concentrated in north-east Shikoku. But all six typhoons did not bring torrential rainfall in the whole island. Some typhoons 10, 11, 15, 16, and 21 were more aggressive in Tokushima, Kochi, and Ehime, whereas typhoons 15, 21, and 23 were most troublesome for Kagawa Prefecture. A greater concentration of landslides was noticed mainly in Kagawa Prefecture, that year’s third most disaster-affected prefecture in the whole Japan. Based on available information, the different typhoons hit various places and significant damage was observed in areas of maximum rainfall concentration. Example of such events within each prefecture is described in the following sections. Landslides in Tokushima From late July to early August, typhoon 10 brought heavy rainfall (more than 2000 mm) to the southern part of Shikoku. This typhoon rainfall created four huge landsides around the Kisawa village of Nakagawa, occurring mainly in the Ooyochi, Kashu, Azue, and Shiraishi area (Fig 8).

Table 1, Typhoon in Shikoku Island in 2004, rainfall values are taken from AMeDAS, JMA, 2005. 4 6 10 11 15 16 18 21 23

6/11 6/19-6/21 7/31-8/2 8/3-5 8/17-18 8/29-30 9/7-8 9/29-30 10/19-21Conson Dianmu Namtheun Malou Megi Chaba Songda Meari Tokage

960 915 935 996 970 910 925 940 940Cumulative (mm) 138 403 1234 188 386 507 536 168 542

mm h-1 32 45 55 52 51 56 46 49 54

Rainguage Station Saga Asahimaru Sawadani Asahimaru Ikegawa Jojusha Jojusha Kamihayashi Tomisato

Prefectrue Kochi Tokushima Tokushima Tokushima Kochi Ehime Ehime Ehime Ehime

Name of Prefecture

Typhoon No.DateName of TyphoonMin. Pressure (hPa)

Maximum Values

797

Fig 8, The Azue landslide In Kaminaka, data recorded by Shikoku Electric Power Co. Inc. showed hourly rainfall reaching more than 120 mm; the highest daily precipitation (1,317 mm) was on 1 August 2004 (Fig. 9). It is believed that this value surpassed the previous record held by Kito (1,114 mm). This village is 16 km southwest of Kisawa village and was visited by Typhoon Fran on 11 September 1976 (Wang et al., 2005). Similarly, typhoons 6, 11, 16, 18, and 23 also hit Tokushima Prefecture and small- to medium-scale landslides were reported along national and express highways and also in the Naka River catchments. The failure areas of Kisawa and Kaminaka area were more affected by rainfall from typhoons 11, 16, 18, and 23. Some 550 mm of rain fell in Uekatsu, Tokushima during typhoon 16. Likewise, a daily rainfall of more than 300 mm was recorded in the southwestern part of Tokushima.

2043

0

20

40

60

80

100

120

140

20 24 4 8 12 16 20 24 4 8 12 16 20 24 4 8 12 16 20 2430-Jul 31-Jul 1- Aug 2-Aug

Rai

nfal

l (m

m)

0

500

1000

1500

2000

2500

Cum

ulat

ive

Rain

fall

(mm

)

1317 mm

Fig. 9, Rainfall pattern in Kaigawa area during Typhoon 10 (31-Jul to 2 Aug, 2004), this hyetograph represent one day maximum precipitation ever recorded in Japan (source: Shikoku Electric Power Co. Inc.) The Ooyochi landslide occurred at 20:00-23:00 h of 1 August and Azue was advanced (2300 h, 2 August) after a total relaxation of typhoon rainfall around 16:00 h of 2 August (Fig. 10). The area mainly consists of Palaeozoic greenstone, Palaeozoic and Mesozoic pelite and greywacke, and serpentinite of the Mesozoic Kurosegawa terrain, as well as scarce limestone and chert. On the landslide area, cedars are the main species in the forest. The river

valley is steep and most of the settlements are located on gentle slopes formed by past landslide materials and on narrow riverside terraces.

Fig. 10, Rainfall pattern and time of landslide in Kisawa area during typhoon 10 Landslides in Ehime Ehime prefecture also suffered tremendous losses due to debris flows and flooding induced by the typhoon rainfall of 2004. Typhoons 15, 16, 18, 21, and 23 affected Ehime in 2004. In these typhoon events, hourly rainfall exceeding 50 mm and total rainfall of more than 400 mm were recorded in many stations. Rainfall due to typhoons 15 and 21 triggered many landslides at different locations in northeast Ehime. The hardest hit area was Niihama City (Fig. 11). The debris flows near the hill base of this city took the lives of some 25 people and destroyed property worth 40 billion yen (Bhandary and Yatabe, 2005). Geologically, Niihama consists of green schist of the Sambagawa Belt in the south and sandstone and shale of the Izumi Group in the north. The area that suffered extensive failure during the disaster is part of the Izumi Group. This geological belt is primarily made of sedimentary deposits of sandstone with frequent intercalation of shale. Most of the failed slopes consist of decomposed sandstone, either stiff clay resulting from weathering of shale or less disintegrated rock mass. For many landslide researchers, the Izumi Group used to be little known in relation to debris flow or landslide disaster. But, in 2004, during typhoons 15 and 21, the intensity of rainfall was so high that hundreds of landslides and debris flows occurred. Typhoon 23 also triggered many landslides in Ehime. Slope failure events in Ibukimachi (Uwajima City), Kawanouchi (Touon City), and Shingumachi (Shikoku Chuo City) are some other examples (see Fig. 7).

Fig. 11, Landslides in Takamatsu-Matsuyama Express Highway at Niihama, Typhoon 21, September 29, 2004 (source: MLIT, 2004)

798

Landslides in Kagawa Kagawa Prefecture was hit mainly by four typhoons (15, 16, 21, and 23) in 2004 and suffered loss of lives and property because of many landslides triggered by typhoon rainfall. Although mean annual rainfall of Kagawa is less than 1000 mm, it was the most affected prefecture in 2004. It was observed that hourly rainfall exceeding 50 mm and total 24-h rainfalls of more than 200 mm were the main causes of slope failure at different locations in Kagawa. The hardest hit areas were Moriyuki and Monnyu of Kagawa Prefecture. There was localized damage in the highway and forest area also around Takamatsu City (Fig. 12). Central Kagawa is also affected by many landslides occurred in the forest. In Toyohama and Onohara, west Kagawa, a heavy rain in the afternoon of 17 August due to typhoon 15 killed five persons despite a few landslides and debris flow events. The rainfall that started only from 11:00 h of that day was enough to create failure in both towns during the typhoon. In the case of typhoon 21, extensive damage was observed in the whole area. During typhoon 15, a total of 246 mm of rainfall was recorded in 24 h, with maximum hourly rainfall of 54 mm. Similarly, during typhoon 21, 249 mm of rainfall occurred with a maximum 1-h rainfall of 65 mm. Almost all streams coming from the mountain slopes brought a huge pile of debris. During typhoon 23, in east Kagawa, hourly rainfall reached 125 mm in Maeyama. Likewise, Kusaka Pass received a maximum of 116 mm h-1 and Yodayama received 107 mm h-

1. This extreme rainfall created many landslides and debris flow in the Shiratori, Moriyuki, Minamigawa, and Monnyu area (Dahal et al., 2008).

302

279

243247

0

100

200

300

400

1 3 5 7 9 11 13 15 17 19 21 23 1 3 5 7 9 11 13 15 17 19 21 23

Oct-10 Oct-20

Cum

ulat

ive

Rai

nfal

l (m

m)

Zoda Myojin

Kotohira Takamatsu

Small soil slips was started in Kotonami

Failure at Tatsuyama, Kotonami

Hiyama LandslidesTakamatsu

Failure at Shiroyama,

Kotonami

Mineyma LandslidesTakamatsu

Fig. 12, Rainfall pattern and landslides timing in central Kagawa during typhoon 23 Landslides in Kochi In 2004, typhoons 4, 6, 10, 11, 15, 18, and 23 hit Kochi Prefecture. Rainfall caused by typhoon 15 (Megi) induced many landslides in the Yoshino River basin of Shikoku on 17-18 August. Reihoku District in Kochi Prefecture was greatly affected by typhoon 15. Okawa and Uwezugawa villages were severely damaged. Many landslides occurred along the roadside slopes too. As a result, Okawa was isolated from the other parts of the prefecture (Yatabe and Hasegawa, 2004). Rainfall data recorded in the Komatsu rain gauge station showed hourly precipitation exceeding 100 mm at 1600-1800 h of 17 August and total 2-h rainfall was 205 mm. Cumulative rainfall from 01:00 h to 18:00 h reached 518 mm with maximum rainfall intensity of 104 mm h-1. Total precipitation reached 1,000 mm in 3 days. Many landslides and debris flow occurred during 17:00-18:00 h of 17 August (Fig. 13).

The Okawa area mainly consists of crystalline green schist of the Sambagawa Belt, which includes pelitic schist, psammitic schist, and siliceous schist. There are thin to thick deposits of quaternary colluvium on the mountain slope of the Yoshino River Valley. The crystalline schist is usually well known for landslide hazard and the sliding mass mainly consists of weathered and jointed schist (Hong et al., 2005).

131

1055

1086

0

20

40

60

80

100

120

140

160

180

200

1 5 9 13 17 21 1 5 9 13 17 21 1 5 9 13 17 21

17-Aug 18-Aug 19-Aug

Rai

nfal

l (m

m)

0

200

400

600

800

1000

1200

Cum

ulat

ive

Rai

nfal

l (m

m)

Komatsu

Ashatani

Komatsu

Ashatani

Many landslides

Fig. 13, Rainfall pattern recorded in Asatani and Komatsu during typhoon 15 CHARACTERISTICS OF DEBRIS SLIDES AND FLOWS In the 2004 typhoon events, the question of how many landslides happened is still not well adequately answered. Because of the thick forest in unpopulated hills, many landslides were only recognized long after the event Researchers are still observing unreported small- to medium-sized landslides triggered by the typhoon rainfall of 2004 in the forest of central Shikoku Island. Nevertheless, the International Sabo Association of Japan stated that nearly 600 various types of slope failures have occurred in Shikoku during the 2004 typhoon events. This number includes slope failures that need mitigation measures, such as sabo dam. But field and air photo studies revealed that a total of 659 slope failures occurred only in the Toyohama (west Kagawa) area during the 2004 typhoons. Field observations of more than 250 slides in all prefectures indicated that both sliding and flowing are critical to the failure process. Various research studies established the nomenclature and classification of landslides in Shikoku, coming up with different terms used for such failures (Curden and Varnes, 1996), like debris flows, debris avalanche, soil slips, debris slides, flow-like landslides, slide-flows, so forth. In a general sense, landslides in Shikoku Island after the 2004 typhoon events can be divided into translational slides, rotational slides, and a combination of both on the basis of the shape of the failure surface (Fig. 14). The translational slides were found to be the most predominant failure mode of debris slide. The rotational slides occurred rarely but in a large scale. The main characteristics of landslides are as follows. • The volume of failure generally ranged from tens of cubic meters to

hundreds of cubic meters. • In many cases, majority of the debris slides and flows were shallow,

with a depth of failure of <2 m. • In central and eastern Kagawa, landslides were composed of

intensely decomposed plutonic and volcanic rocks, whereas those in Ehime, Kochi, and Tokushima had thick colluviums and intensely fractured green pelitic schist.

• Some of landslides occurred along the colluvium and intact bedrock contact also.

• It is noticed that many of the landslides (e.g., Moriyuki, Monnyu,

799

Toyohama, Okawa, and Kisawa) were translational debris slides occurring first on steep zero- order valley or concave slope and debris materials were run down through first-order stream channel. Commonly, the slides are several to tens of meters wide and tens to hundreds of meters long. The slip surface, 0.3-2 m deep, was near the irregular contact between the colluvium and decomposed or fresh bedrock. All of these initial soil slides were mobilized completely to form debris flows and the liquefied debris also scraped off the saturated soil of the first-order gully and the whole mass of the channel began to flow, like a liquid following the liquefaction flow rule (Darve, 1996). The flow continued to erode its route and either piled up huge debris on the mouth of the stream or continued traveling through the second-order stream to a considerable distance on the sloping terrain. Therefore, during a field visit, the first-order streams were noticed as totally cleaned up (Fig. 15).

• In some cases, the down-to-top failure was also noticed and it was exactly like the headward erosion of the first-order stream onto the zero-order stream by extremely high rate of erosion. The Kashu landslide of Kisawa area is a suitable example of such failures.

• In addition, field observations pointed out that landslides were occurred both in natural and man-made slopes having thick residual or colluvial deposits.

Translational Failure Rotational Failure Compound Failure

a. Longitudinal Section

b. Transverse Section

Rectangular Shape U-shape V-Shape Fig 14, Schematic representation of slides observed in Shikoku after 2004 typhoon events

Fig. 15, Clean-up of colluviums and weathered rock from first order stream after debris flow, Kotonami area, central Kagawa RAINFALL THRESHOLD OF THE EVENTS For the prediction of rainfall-triggered landslides, the concept of hydrological landslide-triggering thresholds has already been developed, to some extent. Similarly, correlations between rainfall intensity, rainfall duration, and landslide events are also already established. Experience shows that shallow failures are usually

triggered by comparatively short, intense storms (Pasuto and Silvano 1998, Wieczorek et al., 2000; Jakob and Weatherly, 2003; Aleotti, 2004), whereas most of the deep-seated landslides are affected by long-term variation of annual rainfall and daily rainfall (Hong et al., 2005). The rainfall intensity threshold for triggering landslides is one of the debatable topics in the last 25 years. Caine first published a paper in 1980 about the threshold rainfall intensity for landslide triggering. After that, many attempts were done to establish the threshold rainfall intensity in the global and in the regional context. In the case of Shikoku, data of rainfall amount and exact landslide timing were collected from 18 major and minor landslide events. Although there were hundreds of landslides, only 18 landslide events were standardized with the exact time of occurrence. In terms of classification, all these 18 rainfall-triggered landslides were either debris slides or debris flows (Curden and Varnes, 1996) or flow-like landslides (Hungr et al., 2001; Hungr, 2003; Jakob and Weatherly, 2003). Fig. 16 illustrates the temporal pattern of rainfall intensity vs. duration conditions during the various typhoons of 2004. The continuous lines represent rainfall intensity vs. duration, whereas time of occurrence of landslides is shown as circular symbols. The upper graph (i) corresponds to rainfall recorded at the different rain gauge stations during typhoon 10 and the time of landslides are shown using circle symbols. Likewise, graphs (ii), (iii), and (iv) represent rainfall and times of occurrence of landslide during typhoons 15, 21, and 23, respectively, in different places of Shikoku. The data of effective temporal rainfall intensity for producing landslide vs. rainfall duration are also plotted in log-log graph to analyze the empirical rainfall thresholds (Fig. 17). It was noted that the intensity of rainfall to produce landslides around the Shikoku mountain range was higher than that in northeast Shikoku (area highly affected by typhoon events of 2004). Similarly, all failure events during various typhoons began only after a minimum of 10 h of continuous rainfall. However, during typhoon 23, the many failures began only after 35 h of continuous rainfall. The threshold line of rainfall intensity proposed by various researchers (Caine, 1980; Cancelli and Nova, 1985; Wieczorek, 1987; Cannon and Ellen, 1985; Larsen and Simon, 1993; Ceriani et al.,; 1994; Crosta and Frattini, 2001; Zezere et al., 2005; Guzzeti et al., 2007; Dahal and Hasegawa, 2008) area also plotted in Fig. 17. A comparison of the curves showed that threshold rainfall to produce landslides for Shikoku (as per data on rainfall intensity of 2004 events) follows the threshold line for storm-triggering landslides proposed by Dahal and Hasegawa, (2008) for Himalaya (tropical monsoon region) and Larsen and Simon (1993) for Puerto Rico (humid-tropical region). The curve prepared by Dahal and Hasegawa (2008) using the 55-year data of the Nepal Himalaya showed relatively better congruence for the Shikoku data than the curve proposed by others for different part of world. The curve proposed by Caine (1980), Cancelli and Nova (1985), Ceriani et al. (1994), Crosta and Frattini (2001), and Aleotti (2004) for the Alps could not cover up the average rainfall intensity of Shikoku during failures due to the extreme rainfall events of 2004. The curves proposed by Wieczorek (1987) and Cannon and Ellen (1985) for the Rockies also do not have any similarity with the Shikoku data. CONCLUSIONS Based on the above observations, analyses, and discussion, the following conclusions are drawn. • A study of the last 55-year of typhoon data shows that Japan is very

susceptible to typhoon-brought rainfall disasters. Shikoku Island (northeast Pacific Ocean side) also has a severe problem of typhoon-brought, rainfall-triggered landslides.

800

Niihama T-21

Toyohama T-21

0

1

2

3

4

5

6

7

8

9

0 5 10 15 20 25 30 35 40 45 50 55 60

Duration (hours)

Rai

nfal

l Int

enst

iy (m

m h

-1)

Niihama T-23

Kusaka Pass (Moriyuki)

Monnyu

Toyohama T-23 Zoda

Takamatsu

0

2

4

6

8

10

12

14

16

18

0 5 10 15 20 25 30 35 40 45 50 55

Duration (hours)

Rai

nfal

l Int

enst

iy (m

m h

-1)

Ashatani Ty-15

Niihama Ty-15Toyohama Ty-15

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

Duration (hours)

Rai

nfal

l Int

enst

iy (m

m h

-1)

Kominono Ty-10

Nakanose Ty-10

0

5

10

15

20

25

30

35

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

Duration (hours)

Rai

nfal

l Int

enst

iy (m

m h

-1)

Graph (i)

Graph (ii)

Graph (iii)

Graph (iv)

Fig. 16, Temporal pattern of rainfall intensity vs. duration condition for the 2004 rainfall events. The left of the graphs represent initiation of rainfall. Symbols indicate the occurrence of landslides in Shikoku during various typhoon events of 2004. • The number of debris slides and debris flows of 2004 in Shikoku

Island results from severe rainstorms brought by various typhoons,

along with the presence of unstable colluviums, moderately to highly weathered rocks, and steep mountain terrains.

• Most of the slope failures are normally shallow and consist of translational slides, rotational slides, and combination of both. The flow of materials after failure along the first-order stream on steep mountain slopes (30o–45o) is a common scenario in many events.

Fig. 17, Rainfall intensity and durations associated with slopes failures during extreme rainfall events of 2004 in Shikoku, Japan and curve proposed by various researchers for similar case. From the data of 2004 landslides events, threshold rainfall intensity for Shikoku nearly follows the threshold line for storms triggering landslides proposed by Larsen and Simon (1993) and Dahal and Hasegawa (2008) for Puerto Rico (Humid-Tropical Region) and the Nepal Himalaya, respectively. For northeast Shikoku, curve proposed by Larsen and Simon (1993) as well as Dahal and Hasegawa (2008) is relatively better than line proposed by Caine (1980) from worldwide data. Similarly, for Central Shikoku, the curve proposed by Zezere et al. (2005) is the most representative one. • The rainfall pattern and failure process in central Shikoku and

northeast Shikoku are entirely different because of bedrock geology, thickness of colluvium, permeability contrast between colluvium and bedrock, slope gradient, and average rainfall intensity.

• A study of average rainfall intensity during failure helps to gain an understanding, to some extent, of threshold rainfall intensity behind slope failure in Shikoku Island. When various threshold curves for rainfall-triggering landslides as proposed by different researchers are compared with Shikoku data, the curve proposed by Larsen and Simon (1993) for Puerto Rico (humid-tropical region) and Dahal and Hasegawa (2008) for the Nepal Himalaya (tropical monsoon region) more or less reflect the data of average rainfall intensity during failure in Shikoku due to the 2004 extreme rainfall events. However, historical average rainfall intensity and failure data of Shikoku need to be analyzed in detail to understand the threshold curve for Shikoku. The 2004 data certainly give some helpful insights. Shikoku has very similarly in geological settings with Himalaya (Kikkawa, et al., 2003) and rainfall threshold also shows such similarity in some extent.

• Further studies to investigate hydrological responses of mountain slopes and mechanical behaviour of soils from major to minor catchments scale are essential to understand the local variation in rainfall intensity thresholds. The influence of antecedent rainfall, if there is any, should also be incorporated in such a study.

ACKNOWLEDGMENTS We thank Mr. Toshiaki Nishimura, Ms. Aika Minno, and Mr. Eitaro Masuda for their help in the field and laboratory data collection. We

801

also acknowledge the indispensable field data provided by the Shikoku Regional Development Bureau, MLIT, Japan; the rainfall data from the Japan Meteorological Agency; the information about typhoons from Digital Typhoon, National Institute of Informatics, Niigata University, Japan); and the rainfall data from the Shikoku Electric Power Company. Dr. Netra Prakash Bhandary is sincerely acknowledged for his comments in manuscript. We would like to thank Mr. Anjan Kumar Dahal and Ms. Seiko Tsuruta for their technical support during the preparation of this paper. The study has been partly funded by the Sasagawa Fund for Scientific Research, The Japan Science Society. REFERENCES Aleotti, P (2004). “A warning system of rainfall-induced shallow

failure,” Engineering Geology, Vol 73, pp 247−265. Bhandary, NP and Yatabe, R (2005). “Influence of precipitation pattern

on triggering surface-layer failures of slopes during 2004 Niihama disasters,” Proc. One-day Symposium on Geo-disasters and Geo-environment, Ehime University, Japan, pp 63−70.

Caine, N (1980). “The rainfall intensity-duration control of shallow landslide and debris flow,” Geogr. Ann., Vol 62A, No 1−2, pp 23−27.

Cancelli, A and Nova R (1985). “Landslides in soil debris cover triggered by rainstorms in Valtellina (Central Alps -Italy),” Proc. IV International Conference and Field Workshop on Landslides, Tokyo, August 1985, pp. 267−272.

Cannon, SH and Ellen, SD (1985). “Rainfall conditions for abundant debris avalanches, San Francisco Bay region, California,” California Geology, Vol 38, No 12, pp 267–272.

Ceriani, M, Lauzi S and Padovan N (1994). “Rainfall thresholds triggering debris flows, in the alpine area of Lombardia Region Central Alps–Italy,” Man and Mountain '94, First International Congress for the Protection and Development of Mountain Environment. Ponte di Legno (BS, Italy), June 1994, pp 123–139.

Crosta, G, and Frattini, P (2001). “Rainfall thresholds for triggering soil slips and debris flow,” Proc. of EGS 2nd Plinius Conference 2000, Mediterranean Storms, Siena, pp 463–488.

Crozier, MJ, (1999). “Prediction of rainfall-triggered landslides: a test of the antecedent water status model,” Earth Surface Processes and Landforms, Vol 24, pp 825–833.

Curden, DM and Varnes, DJ (1996). “Landslide types and processes,” In: Turner AK and Schuster RL (Eds.), Landslides: Investigation and Mitigation. Sp. Rep. 247, Transportation Research Board, Washington D.C., pp 36–75.

Dahal, RK, Hasegawa, S, Nonomura A, Yamanaka, M, Masuda, T, Nishino K (2008). “GIS-based weights-of-evidence modelling of rainfall-induced landslides in small catchments for landslide susceptibility mapping,” Environmental Geology, Vol 54, No 2, pp 314–324,

Dahal RK and Hasegawa S (2008). “Representative rainfall threshold for landslides in the Nepal Himalaya,” Geomorphology, 15p (in press) DOI:10.1016/j.geomorph.2008.01.014

Darve, F (1996). “Liquefaction phenomenon of granular materials and constitutive stability,” Engineering Computations, Vol 13, No 7, pp 5–28.

Digital Typhoon (2006). “Data of typhoons published in http://agora.ex.nii.ac.jp/digital-typhoon/index.html,” National Institute of Informatics, Niigata University, Japan.

Glade, T, Crozier, M, and Smith, P (2000). “Applying probability determination to refine landslide-triggering rainfall thresholds using an empirical Antecedent Daily Rainfall Model,” Pure and Applied Geophysics, Vol 157, pp 1059–1079.

Guzzetti F, Cardinali, M, Reichenbach, P, Cipolla F, Sebastiani C, Galli M, and Salvati P (2004). “Landslides triggered by the 23 November 2000 rainfall event in the Imperia Province, Western Liguria, Italy,”

Engineering Geology, Vol 73, pp 229–245. Guzzetti, F, Peruccacci, S, Rossi, M, and Stark, CP (2007). “Rainfall

thresholds for the initiation of landslides in central and southern Europe,” Meteorol. Atmos. Phys., Vol 98, No 3-4, 239−267.

Hasegawa, S and Saito, M (1991). “Natural Environment, Topography and Geology of Shikoku,” Tsushi-to-Kiso, Japanese Geotechnical Society, Vol. 39-9, No 404, pp 19−24 (In Japanese).

Hong, Y, Hiura, H, Shino K, Sassa K, Suemine A, Fukuoka H, and Wang G (2005). “The influence of intense rainfall on the activity of large-scale crystalline schist landslides in Shikoku Island, Japan,” Landslides, Vol 2, pp 97–105.

Hungr, O (2003). “Flow slides and flows in granular soils,” Proc. of International Workshop on Occurrence and Mechanisms of Flows In Natural Slopes and Earth fills, website http://www.unina2.it/flows2003/flows2003/articoli/Hungr-Flows.pdf.

Hungr, O, Evans, SG, Bovis, MJ, and Hutchinson, JN (2001). “A review of the classification of landslides of the flow type,” Environmental Engineering Geoscience, Vol 7, 231–238.

ISA, (2006). “International Sabo Association,” various data published in official website of International Sabo Network http://www.sabo-int.org/

Jakob, M, and Weatherly, H (2003). “A hydroclimatic threshold for landslide initiation on the North Shore Mountains of Vancouver, British Columbia,” Geomorphology, Vol 54, pp 137−156.

JMA (2005). “Rainfall Data of Japan as observed by the AMeDAS at 1300 points in Japan,” Japan Meteorological Society, 4 Compact Disc.

Kikkawa, K, Ohno, H, Inagaki H and Hirata N (2003). “A comparison between Nepal and Shikoku by omniscape geology,” Jour. Japan Soc. Eng. Geol. Vol 44, pp 14-24.

Larsen, MC and Simon A (1993). “A rainfall intensity-duration threshold for landslides in a humid-tropical environment, Puerto Rico,” Geografiska Annaler, Vol 75, No 1–2, 13–23.

MLIT, (2004). “Information about typhoon disasters of 2004, Shikoku Regional Development Bureau, Ministry of Land, Infrastructure and Transport,” web page http://www.skr.mlit.go.jp/sabo/dosha/dosha-f.html

Pasuto, A and Silvano, S (1998). “Rainfall as a trigger of shallow mass movements, A case study in the dolomites, Italy,” Environmental Geology, Vol 35, No 2−3, pp 184-189.

Terlien, MTJ (1998). “The determination of statistical and deterministic hydrological landslide- triggering thresholds,” Environmental Geology Vol 35, No 2-3, pp 124−130.

Tsuboki, K, (2005). “High-resolution simulation experiments of typhoons using the cloud-resolving model on the earth simulator”, extended abstract, 7th International workshop on Next Generation Climate Models for Advanced High Performance Computing Facilities, Kona, Hawaii 24−26, February 2005.

Wang, G, Suemine, A, Furuya, G, Kaiboris, M, and Sassa, K (2005). “Landslides in Kisawa area, Tokushima Prefecture during the 2004 Typhoon Namtheun,” Annuals of Disas. Prey. Res. Inst., Kyoto Univ., No. 48 B, pp 1−8.

Wieczorek, GF (1987). “Effect of rainfall intensity and duration on debris flows in central Santa Cruz Mountains, California,” Geol. Soc. of America Reviews in Engineering Geology, Vol 7, pp 93-104.

Wilson, RC, and Wieczorek, GF, (1995). “Rainfall threshold for the initiation of debris flow at La Honda, California,” Environmental and Engineering Geoscience, Vol 1, No 1, pp 11−27.

Yatabe, R and Hasegawa, S, (2004). “Sediment disasters in Shikoku caused by Typhoons 10, 15, and 21 in the year 2004,” Journal of the Japan Landslide Society, Vol. 41, No. 4 (416), pp 94−96 (in Japanese).

Zezere JL, Trigo RM, Trigo IF (2005). “Shallow and deep landslides induced by rainfall in the Lisbon region (Portugal): assessment of relationships with the North Atlantic Oscillation,” Nat. Hazards Earth Sys. Sci. Vol 5, pp 331–344.

802