Distribution and characteristics of landslide in Loess ...

Engineering Geology 236 (2018) 89–96

Contents lists available at ScienceDirect

Engineering Geology journal homepage: www.elsevier.com/locate/enggeo

Distribution and characteristics of landslide in Loess Plateau: A case study in Shaanxi province Jianqi Zhuang a,b, Jianbing Peng a,⁎, Gonghui Wang c, Iqbal Javed d, Ying Wang a, Wei Li a a

College of Geological Engineering and Surveying of Chang'an University, Key Laboratory of Western China Mineral Resources and Geological Engineering, Xi'an 710054, China State Key Laboratory of Geohazard Prevention and Geoenvironment Protection, Collaborative Innovation Center of Geological Prevention, Chengdu 610041, China Research Centre on Landslides, Disaster Prevention Research Institute, Kyoto University, Gokasho, Uji, Kyoto 611-0011, Japan d Department of Earth Sciences, Abbottabad University of Science and Technology, Abbottabad, Pakistan b c

a r t i c l e

i n f o

Article history: Received 4 July 2016 Received in revised form 25 February 2017 Accepted 5 March 2017 Available online 6 March 2017 Keywords: Loess landslide Distribution Characteristics Landslide mechanics Loess Plateau

a b s t r a c t Every year about one third of the geohazards in China occur in the Loess Plateau causing human loss, damaging gas and oil pipelines, destroying highways, railways and degrading farmland. Field investigation and monitoring, in-situ tests and laboratory experiments were performed to improve our understanding of the factors effecting the distribution, characteristics and causes of loess landslides. First, we find that 79% of the landslides are shallower than10m, 85% have a volume of less than 100,000 m3. Second, landslides on the Loess Plateau occur primarily on concave slope profiles that have slope angles of 20–35° and that face south-east. Third, the equivalent coefficient of friction of loess landslides is very low resulting in long run-out with a low angle sliding surface. Loess landslides generally transform into mud-flows resulting in an increase in volume in transit and forming a geohazard chain. Antecedent rainfall plays an important role in triggering loess landslides. Finally, clusters of landslides in the Loess Plateau occur because the loess easily disintegrates under high pressure due to its loose and highly porous structure. There is a sharp decrease in cohesive strength with increase in deformation and water content and thus landslides tend to undergo static liquefaction during sliding. © 2017 Elsevier B.V. All rights reserved.

1. Introduction Loess is sediment formed by the accumulation of wind-blown fine sand and clay, and is generally found in arid and semi-arid regions. In China, loess covers a total area of approximately 631,000 km2, about 4.4% of China's land area (Liu, 1985). The Loess Plateau, which has an area of approximately 430,000 km2, constitutes the majority of the deposit area, and sits within the central part of the Yellow River basin. The thickness of the loess deposits in this area varies from a few meters to more than 300 m (Derbyshire et al., 2000; Li et al., 2013). Typical characteristics of loess include macro-pores, vertical joints, loose texture and high sensitivity to suction stress, which makes it prone to landsliding (Derbyshire et al., 2000; Dijkstra et al., 1995; Xu et al., 2007; Zhang et al., 2009; Zhang and Liu, 2010). One third of the geohazards in China occur in the Loess Plateau resulting in human loss, damage of gas and oil routes, destruction of roads and railways, as well as reduction in farmlands (Li et al., 2007; Xu et al., 2009; Wang et al., 2014a,b). The frequency of landslides in the Loess Plateau is increasing with time due to human activities and extreme weather conditions. Field investigations revealed that around 15,000 geohazards occurred in the Loess Plateau of Shaanxi province with an average ⁎ Corresponding author. E-mail address: [email protected] (J. Peng).

http://dx.doi.org/10.1016/j.enggeo.2017.03.001 0013-7952/© 2017 Elsevier B.V. All rights reserved.

density of over six per square kilometer (Lei 2001). Among these geohazards 85% were landslide, which is one of the most disastrous geohazards in the area because of its high velocity, large volume, long run-out distance, and strong impact force (Jakob et al., 2005; Zhang and Yin, 2013). The initiation and development of landslides in the Loess Plateau is often a response to three necessary factors, (i) pore water pressure, (ii) steep topography, (iii) and thickness of the loess. Human activities, such as cutting slopes, placing an overburden load on slopes, and irrigation usually play a major role in changing the stresses within a slope, as well as decreasing the soil strength and making the slope steeper. Such changes can lead to a landslide, resulting in a great catastrophe (Zhuang and Peng, 2014). Generally, loess has high cohesive strength at low moisture contents, but this cohesive strength reduces significantly with increasing moisture content (Derbyshire et al., 1994). Due to its high silt content, loess can liquefy under the continuous shaking induced by earthquakes or human activities. The loess then becomes unstable and can travel long distances (Zhang and Wang, 2007; Wang et al., 2014a,b). Catastrophic loess landslides in China on 6th October 2006 in Hua County, 10th March 2010 in Zizhou county, 17th September 1983 in Xi'an, and 7th March 1983 in Dongxiang county resulted in 12, 27, 32 and 237 casualties, respectively (Xing et al. 2013; Derbyshire et al., 2000; Evans and DeGraff, 2002; Zhuang and Peng, 2014). Furthermore, the loess in Heifangtai and the south edge of the Jing River has been very active since the 1980s,

90

J. Zhuang et al. / Engineering Geology 236 (2018) 89–96

witharound120 landslide events along the edge of Heifangtai killing 70 people (Xu et al., 2014). In order to reduce the risk of landslides, a wide variety of studies have been conducted on loess landslides. Some of these studies have established relationships between the environmental parameters of slopes and landslide occurrence on these slopes, examining the influence of landforms, precipitation, geological features, slope profiles, and vegetation cover to identify landslide susceptible zones (Gao, 1988; Zhang and Liu, 2010; Zhuang et al., 2015). Other studies have used ring shear and triaxial tests to examine the mechanism and conditions that lead to soil failures. Such tests indicate a decay in soil cohesiveness and change in angle of internal friction with: changes in water content, capillary forces and suction as well as transfer of salts and clays (Skempton, 1985; Sasitharan et al., 1993; Crozier, 1999; Crosta and Prisco, 1999; Zhang et al., 2009; Zhang et al., 2013). Several models for loess landslide triggering have been proposed, such as: runoff erosion, increased ground water levels, cut slope engineering, or irrigation; progressive failure due to external loading and shallow loess landslides due to precipitation (Li et al., 2007). However, the mechanic of loess landsliding is still not clearly known, and therefore, cannot be used for landslide mitigation (Peng et al., 2014). Many research groups have focused on mapping landslide hazard zones and on risk assessment using statistical and physically based models (Zhuang et al., 2015). Although the mechanism of loess landslide initiation and the factors that trigger loess landslides have been analyzed by many researchers, there remains considerable debate on two key issues. First, why are there so many landslides in the Loess Plateau and what makes loess so susceptible to failure? Second, where do landslides on the Loess Plateau occur and what controls their spatial distribution? To address these questions we conducted a series of field investigations and laboratory experiments with in a large study area in the Loess Plateau of China. 2. Study area The Loess Plateau is the largest and most extensive loess deposit in the world. Its within the central part of the Yellow River basin and is

bordered by the Tengeri Desert, Wuqiao Ridge and Riyue Mountain in the west; Taihang Mountain in the east, Qinling Mountain in the south and Yinshan Mountain in the north (Fig. 1a). Almost 50% of the Loess Plateau sits within Shaanxi province. So, in this study the portion of the Loess Plateau in the north of Shaanxi province has been considered as a case study. Recently, the Loess Plateau area in Shaanxi province has become an important economic zone due to the availability of large amounts of natural resources, such as oil, gas, coal. Engineering activities, such as slope cutting and irrigation have greatly changed the geomorphology and hydrology of the region, increasing the area's susceptibility to geo-hazards. Loess mainly forms three types of characteristic geomorphic structures: platforms, ridges, and domes. All three structures have steep slopes on their sides that can easily slide under favorable conditions, such as infiltration of rainfall, irrigation water and slope cutting. Slope angles are 23.5° on average and decrease with elevation. Due to Quaternary climate variability, the sediments are nonuniform in color, composition and structure in the section, forming alternating beds of brownish-yellow loess and brownish-red paleosol. The stratigraphic units in Shaanxi province primarily comprise Quaternary strata (Zhang and Liu, 2010). Quaternary loess covers the entire area of the Late Pleistocene Malan deposits (less than 10 m depth and now largely removed by erosion); the entire area of the Middle Pleistocene Lishi deposits (about 80 m, which cover nearly the entire area of the plateau); and a few Earlier Pleistocene Wucheng deposits that are located near the lower elevations of the Lishi loess (Liu, 1985; Fig. 1b). Rainfall is a common triggering factor for landslides in most loess areas (Peng et al., 2015; Zhuang et al., 2015, 2016). The Loess Plateau has an annual precipitation between 450 and 720 mm. High rainfall has been recorded between June and September and is often associated with storms. Loess itself has a very low permeability; instead, vertical joints on the top of loess slopes are the main pathway for the vertical movement of rainwater. Erosion along the joints forms sink holes and opens fissures to permit a large amount of rain water to infiltrate. As the water encounters a relatively impermeable bed, such as paleosol,

Fig. 1. The geological map and typical stratigraphy of the loess in Luochuan (Liu, 1985).


tertiary red clay or bed rock, it accumulates on the bed to soften the soil and induce landslides. 3. Data preparation and methods Data on landslides were recorded via field investigations, and geohazard reports published by the local geological environment monitoring station in the Shaanxi Province. The dataset included information on landslide locations, depths and volumes as well as their consequences in terms of human and economic costs. Field investigations and aerial photographs were used to map the locations of the landslides at 1:50,000 scale, and to analyze the features of the landslide areas. The depth of the landsides was measured by

91

Portable Laser Range Measuring Instrument compared with the intact terrain. Landslide volume was calculated by assuming a rectangular failure geometry (area multiplied by depth). A Digital Elevation Model (DEM) with are solution of 25 m was obtained from State Bureau of Surveying and Mapping. The cohesion of loess with different water content was tested using direct shear tests following soil test standard GBT50123-1999 (2007). Daily rainfall data were obtained from National Meteorological Observation Stations with a precision of 0.1 mm/day. Ring shear tests were used to study the sliding surface after landslide initiation at large shear displacements using undrained ring shear apparatus developed by Kyoto University (Sassa et al., 2004). The undrained ring-shear apparatus (DPRI-Ver.5) used in this study has an inner

Fig. 2. The loess landslides in North of Shaanxi province associated with landform.

92


Fig. 3. Distribution of landslide depth (a) and volume (b) in the Loess Plateau of Shaanxi Province.

diameter of 12 cm, an outer diameter of 18 cm and a maximum sample height of 10.9 cm. A remodeled soil sample was used in the ring shear test, with consolidation and saturation processes similar to those used in consolidated undrained triaxial tests. The normal stress, effective normal stress, pore pressure, shear resistance and volumetric strain were automatically recorded (Sassa et al., 2004). The specimens were saturated with pure water assisted by CO2 saturation, until the pore pressure coefficient exceeded 95%, considered the point of full saturation. The undrained shearing was carried out with a constant shear velocity of 0.1 mm/s and a normal stress of 150 kpa. Additional details on the design and construction of this ring shear apparatus, as well as the operating method, can be found in Sassa et al. (2003, 2004).

4. Distribution Identification of landslide occurrence distribution is very helpful in landslide susceptibility assessment (Dai et al., 2001, 2002; Guzzetti et al., 2006). The landslide database compiled for this analysis includes 4122 landslides that occurred in Shaanxi province (Fig. 2), and provides information on landslide location, volume and depth. The landslides in the Loess Plateau of Shaanxi province are mostly shallow landslides with depth less than 2 m. Fig. 3 shows that 64.5% of

landslides were less 2 m deep and 20.1% of landslides had volumes less than 10 × 104 m3. Fig. 4 shows the variation in landslide frequency with slope angle, aspect and profile curvature. The highest density of landslides (82.5%) occurred at slopes of 20–35°. The landslide frequency ratios (the ratio of landslide area to total area for each category) indicate that the probability of landslide occurrence increases with slope angle, especially for slopes greater than 25° (Fig. 4a). 46.3% of landslides are located on slopes oriented between 90 and 225° (i.e. East to South West). Frequency ratio analysis shows that slopes with southeast to southwest aspects are associated with a high probability of landslide occurrence (Fig. 4b), and that most of the landslides occurred in the south-east direction. We calculate slope curvature using the profile curvature tool in ArcGIS 10.2 (Evans, 1980). Positive values indicate convex–upward surfaces and negative values indicate convex–downward surfaces, while a value of zero indicates that the slope is planar (Wilson and Gallant, 2000). Therefore, the values of curvature were classified into three types of surfaces: concave, planar and convex. Landslides occurred in the study area on concave (57.91%), planar (10.29%) and convex slopes (31.79%). Concave surfaces were particularly susceptible to shallow landslide occurrences in the Loess Plateau. Meanwhile, frequency ratio values greater than 1 were distributed across locations with concave surfaces (Fig. 4c) since these surfaces

Fig. 4. Landslide distribution and frequency ratio for varying: slope (a), profile curvature (b), and aspect (c) in the study area.


93

move longer distances than other landslides of the same volume. According to Wang's report (Wang, 2000), a landslide behaves as a fluid when the equivalent coefficient of friction is below 0.17. It was found that 23% of the loess landslide exhibited an equivalent coefficient of friction below 0.17.This percentage includes loess landslide triggered by earthquake and precipitation. Loess landslides triggered by earthquakes have particularly low equivalent coefficient of friction indicating that loess landslides can transform into flows resulting in a longer travel distance. Fig. 5. The aforementioned landslide parameters.

have steep upper slopes with high potential energy relative to the gentle lower slopes and thus fails easily. 5. Characteristics of loess landsides 5.1. Long run-out distance with lower equivalent coefficient of friction The equivalent coefficient of friction demonstrates the mobility of landslides, and is defined as H/L; where H is the difference in height between the top of the landslide scar and the toe of the landslide deposit and L is the plan form landslide travel distance from top to toe (Scheidegger, 1973; Hungr, 1995; Hunter and Fell, 2003; Zhang and Yin, 2013). The aforementioned landslide parameters, including height (H) and travel distance (L), are shown in Fig. 5. Thirty-one rock-slides triggered by the 2008 Wenchuan Earthquake with long travel distances were studied, including eleven rock-slides from the literature and 20 from field investigations immediately after the earthquake. The datasets include information on landslide height, travel distance and volume. Fifty-seven loess landslides triggered by the 1920 Haiyuan earthquake were also investigated. Although some evidence of these landslides has disappeared, characteristics of the larger landslides are still identifiable and quantifiable. These 57 landslides were identified using remote sensing and field investigations. An additional dataset of 46 rock landslides triggered by precipitation was collected from the literature (Zhang and Yin, 2013). Twenty-five loess landslides triggered at the edge of the loess tableland by irrigation on the Heifangtai Range and on the left bank of the Jing River were investigated. These 25 landslides were either identified from the literature or from field investigation but were verified in the field in every case. The relationship between landslide volume and the equivalent coefficient of friction reveals an interesting pattern (Fig. 6). The equivalent coefficient of friction of loess landslides is lower than rock-slides for the same landslide volume. This suggests that loess landslides can

5.2. The antecedent rainfall plays an important role in triggering landslides Antecedent rainfall, rainfall intensity and rainfall duration are also important factors that affect landslides. Current literature on rainfall thresholds for predicting landslide hazards suggests that there are two viewpoints regarding the selection of appropriate precipitation parameters. One focuses on rainfall intensity (Guzzetti et al., 2007) and the other on antecedent effective rainfall (Cui et al., 2008). The influence of rainfall on landslides differs substantially depending upon the landslide type and materials involved (Guzzetti et al., 2008). A study on tropical soils in Hong Kong revealed that antecedent rainfall of any duration is not significant in triggering landslides (Brand et al., 1984). However, a similar study in Wellington, New Zealand, revealed that rainfall in antecedent periods of up to 10 days can influence the soil moisture balance and trigger landslides (Crozier, 1999). Thus, the effect of antecedent moisture may be related to regional climate and soil permeability properties. In order to determine the general relationship between antecedent rainfall and loess landslides, we used daily rainfall in 2009–2011 and July 2013 in Yan'an for study. In 2010 and 2011, no landslide was reported in this area, even with daily rainfall of 64.8 mm (June 18, 2010), 70.6 mm (June 21, 2010) and 54.4 mm (July 29, 2011). However, for these dates, the antecedent rainfall was relatively small, with cumulative antecedent rainfall values (calculated for a 10-day period before the three dates) of 42.4 mm, 89.2 mm, and 26.4 mm, respectively. As indicated in Fig. 7, the precipitation that triggered landslides in the study area was characterized by high antecedent rainfall and moderate intraday rainfall on August 25, 2009 and July 12, 2013.The cumulative antecedent rainfall values in the 10-day period before these two rain dates were 130.7 and 190 mm, respectively, with the intraday rainfall of 64.3 and 87.7 mm, respectively. In order to validate the influence of the antecedent rainfall on loess landslides, the data were also collected from other monitoring stations in the Bailu tableland (also in the Loess Plateau). We used the days with the largest daily rainfall: 71.8 mm (July 19, 2007), 75.5 mm (August 09, 2007) and 53.9 mm (September 17, 2011) as examples. For July 19, 2003 and August 9, 2003, the antecedent rainfalls were relatively small, with only 19.6 mm and 29.6 mm of cumulative rainfall respectively during the preceding10 days. No landslides occurred on either date, although there was sufficient intraday rainfall (Fig. 8). As indicated in Fig. 8, the precipitation that triggered a landslide was characterized by high antecedent rainfall and relatively low daily rainfall (46.7 mm). Therefore, it can be concluded that antecedent rainfall plays an important role in triggering landslides in loess areas which is also mentioned by Du (2010), but differs from other regions, e.g.: Hongkong, post-fire areas in the USA and southwest mountains in China (Brand, 1992; Cui et al., 2008; Cannon et al., 2008; Zhuang et al., 2015). 5.3. Geohazards chain

Fig. 6. The relationships between landslide volume and equivalent coefficient of friction for loess and rock landslides triggered by earthquake and pore pressure increase.

Loess landslides most commonly change into mud-flows due to continuous decrease in material strength under shearing. The velocity and volume of the flow increase with distance along the route during landslide transition to mud-flow, which can be highly destructive. Continuous and heavy precipitation triggered a large loess landslide at the Dagou Village of Mapaoquan Town, Tianshui City, Gansu Province

94


Fig. 7. Daily rainfall in April 2009, April 2010, July 2011 and July 2013.

of China on July 21, 2013 (Peng et al., 2015). The loess landslide rapidly turned into a mud-flow with a volume of 1.9 × 105 m3 (the landslide volume was about 1.6 × 105 m3). Based on field investigations, the peak flow velocity and peak discharge of the mud-flow were estimated to be approximately 7.2 m/s and 730 m3/s, respectively. Both the velocity and discharge of mud-flows in loess tend to increase from the initiation of a mud-flow (5.1 m/s and 482 m3/s). The loess landslide-mudflow travel distance was 1100 m and the equivalent coefficient of friction was 0.169. The loess landslide that occurred on September 21, 2013 in Heifangtai, Gansu province, transformed into mud-flow, moved with a high velocity (peak velocity of 9.0 m/s) and traveled a long distance (Fig. 9). It blocked the river and formed a landslide lake. A similar phenomenon also occurred in Yan'an, in the central region of the Loess Plateau, where a group of loess landslides moved as mud-flow triggered by prolonged and intense precipitation on July 13, 2013, resulting in severe damage (Wang et al., 2015; Zhuang et al., 2016).

6. Discussion 6.1. Static liquefaction Evidence of static liquefaction of loess due to shearing or increased water content has been observed in the field after many loess landslides (Zhang et al., 2009; Xu et al., 2012; Peng et al., 2015). We sampled the loess from Jingyang, typical loess in the Loess Plateau, to demonstrate experimentally that loess can experience static liquefaction during sliding. Samples were saturated with BD = 0.99 then consolidated under normal stress of 150 kPa, representative of conditions at a failure plane 5–10 m below the ground surface. The samples were sheared to a shear distance of 120 mm at which point shear resistance and pore water pressure became constant. The test results for a saturated loess sample with initial void ratio of 1.55 are shown in Fig. 10, where normal stress, pore-water pressure and shear resistance are plotted against shear displacement (Fig. 10a), and the effective-stress path is shown

Fig. 8. The precipitation process during 10-day period before July 19, August 09, 2007 and September 17, 2011.


95

Fig. 9. A loess landslide in Hifangtai on September 21, 2013.

Fig. 10. Test results of ring shear test (a: pore water pressure generation, shear resistance and normal stress against shear displacement; b: effective stress path).

in Fig. 10b. Pore water pressure increased sharply immediately after shearing began, to a peak of 130 kPa, which is nearly equal to the normal stress (about 92% of the applied normal stress). Shear resistance increased immediately after the shearing began, then reached a maximum, and finally decreased to about 57% of the maximum value when

the shear displacement reached around 50 mm, after which the resistance leveled off. The effective stress path and failure line shows that shear resistance increased immediately after initial shearing, and the effective stress path extended toward the final point but did not cross the failure line. This suggests that mass liquefaction in the loess sample under ring shearing. From Fig. 11, it can be concluded that the saturated loess samples are highly prone to liquefy under undrained shearing, and the loess landslide may liquefy during sliding. 6.2. Loess strength decreases sharply with increasing water content

Fig. 11. Changes in cohesion versus water content.

Unsaturated loess with a low moisture content as a large cohesive strength, which decreases sharply with increase in water content (Derbyshire et al., 1994) because the macrostructure is destroyed due to suffosion. We tested strength of loess samples with a range of moisture contents, using direct shear tests and compared these results with those for other soil samples. Graphs (Fig. 11) of cohesion against water content illustrate that the strength a range of soil types, including colluvium soil, silk clay soil, red soil, soft soil and expansive soil decreases with increased moisture content. However, loess is more sensitive to changes in the water content than other soils, illustrated by its larger decrease in strength compared to the other soils. These observations suggest that once water infiltrates into the loess, the strength of the loess decreases dramatically, thereby triggering loess landslides.

96


7. Conclusions The following conclusions were drawn from the current study: 1) Landslides in the Loess Plateau were mostly shallow landslides with small volumes. 79% of the landslides have depth less than 10 m and 85% of the landslides have the volume less than 100 × 104 m3. 2) Landslides in the Loess Plateau are mainly distributed on the concave slope profiles with slope angles of 20–35° and slope direction in South-east. 3) The main characteristics of the landslides in the Loess Plateau are; (i) equivalent coefficient of friction is very low resulting in long run-out, (ii) the sliding surfaces are flat, (iii) they move into debris and form mud-flow resulting in an increase in volume in transit and forming a geohazard chain, (iv) maximum infiltration depth is less than 2.0 m, and water infiltration occurs through loess cracks, (v) and antecedent rainfall plays an important roles in triggering landslides in loess areas. 4) Saturated loess is highly prone to liquefaction under undrained shearing, suggesting that loess landslides may liquefy during sliding. Loess strength is more sensitive to changes in the water content than other soils, increasing moisture content results in a dramatic decrease in strength of loess, and can triggers loess landslides. Acknowledgements The authors are very grateful to the anonymous reviewers and editors for their thoughtful review comments and suggestions which have significantly improved this revised version of our manuscript. We would also like to express our gratitude to Dr. David G. Milledge from Durham University for polishing and revising the manuscript. This study was financially supported by the National Natural Science Foundation of China (Grant No. 41572272 and 41661134015), the National Basic Research Program of China (No. 2014CB744703), the State Key Laboratory Program of SKLGP (Grant No. SKLGP2016K002) and the Central University Funding of the Chang'an University (310826163503). References Brand, E.W., 1992. Slope instability in tropical areas. Proc. 6th Int. Symp. on Landslides. Vol. 3. CRC Press, Balkema, Rotterdam, pp. 2031–2051 Christchurch, New Zealand. Brand, E.W., Premchitt, J., Phillipson, H.B., 1984. Relationship between rainfall and landslides in Hong Kong. Proc. 4th Int. Symp. on Landslides. vol. 1. BiTech Publishers, Vancouver, BC, Canada, pp. 377–384 Toronto,Canada. Cannon, S.H., Gartner, J.E., Wilson, R.C., Bowers, J.C., Labe, J.L., 2008. Storm rainfall conditions for floods and debris flows from recently burned areas in southwestern Colorado and southern California. Geomorphology 96, 250–269. Crosta, G., Prisco, C., 1999. On slope instability induced by seepage erosion. Can. Geotech. J. 36, 1056–1073. Crozier, M.J., 1999. Prediction of rainfall-triggered landslides: a test of the antecedent water status model. Earth Surf. Process. Landf. 24, 825–833. Cui, P., Zhu, Y.Y., Chen, J., 2008. Relationships between Antecedent Rainfall and Debris Flows in Jiangjia Ravine, China. In: Chen, C.L., Rickenmann, D. (Eds.), Proceedings of the Fourth International Conference on Debris Flow. Springer Press, Wellington, pp. 1–9. Dai, F.C., Lee, C.F., Ngai, Y.Y., 2002. Landslide risk assessment and management: an overview. Eng. Geol. 64 (1), 65–87. Dai, F.C., Lee, C.F., Xu, Z.W., 2001. Assessment of landslide susceptibility on the natural terrain of Lantau island, Hong Kong. Environ. Geol. 40, 381–391. Derbyshire, E., Dijkstra, T.A., Smalley, I.J., Li, Y., 1994. Failure mechanisms in loess and the effects of moisture content changes on remolded strength. Quat. Int. 24, 5–15. Derbyshire, E., Meng, X.M., Dijkstra, T.A., 2000. Landslides in the Thick Loess Terrain of North-West China. John Wiley & Sons Ltd, London. Dijkstra, T.A., Rogers, C.D.F., van Asch, T.W.J., 1995. Cut slope and terrace edge failures in Malan loess, Lanzhou, PR China. Proceedings of the XI ECSMFE Conference. 61–67. DGF-Bulletin; Danish Geotechnical Society, Copenhagen, Denmark. Du, J.W., 2010. Forecast and Warning of the Geo-Hazards Triggered by Rainfall-take the Qinba Mountain and Loess Plateau as Study. Science Press, Beijing (in Chinese). Evans, E.A., 1980. Analysis of adhesion of large vesicles to surfaces. Biophys. J. 31 (3), 425–431. Evans, S.G., DeGraff, J.V., 2002. Catastrophic landslides: effects, occurrences and mechanisms. In: Evans, S.G., DeGraff, J.V. (Eds.), Reviews in Engineering Geology. vol. 15. Geological Society of America, p. 411.

Gao, G.R., 1988. Formation and development of the structure of collapsing loess in China. Eng. Geol. 25, 235–245. Guzzetti, F., Peruccacc, S., Ross, M., Stark, C.P., 2008. The rainfall intensity–duration control of shallow landslides and debris flows: an update. Landslides 5, 3–17. Guzzetti, F., Peruccacci, S., Rossi, M., Stark, C.P., 2007. Rainfall thresholds for the initiation of landslides in central and southern Europe. Meteorog. Atmos. Phys. 98, 239–267. Guzzetti, F., Reichenbach, P., Ardizzone, F., Cardinali, M., Galli, M., 2006. Estimating the quality of landslide susceptibility models. Geomorphology 81, 166–184. Hungr, O., 1995. A model for the runout analysis of rapid flow slides, debris flows, and avalanches. Can. Geotech. J. 32, 610–623. Hunter, G., Fell, R., 2003. Travel distance angle for “rapid” landslides in constructed and natural soil slopes. Can. Geotech. J. 40, 1123–1141. Jakob, M., Hungr, O., Jakob, D.M., 2005. Debris-Flow Hazards and Related Phenomena. Springer, Berlin, pp. 136–157. Lei, X.Y., 2001. Geo-hazards in loess pletuea and human activity. Science Press, Beijing, pp. 1–56 (in Chinese). Li, T.L., Long, J.H., Li, X.S., 2007. Types of loess landslides and methods for their movement forecast. J. Eng. Geol. 15 (4), 500–506 (in Chinese). Li, T.L., Wang, C.Y., Li, P., 2013. Loess deposit and loess landslides on the Chinese loess plateau. In: Wang, F., et al. (Eds.), Progress of Geo-Disaster Mitigation Technology in Asia. Environmental Science and Engineering. Springer-Verlag, Berlin Heidelberg, pp. 235–261. Liu, T.S., 1985. Loess and the Environment. Science Press, Beijing (in Chinese). Peng, J.B., Fan, Z.J., Wu, D., Zhuang, J.Q., Da, I.F.C., Chen, W.W., 2015. Heavy rainfall triggered loess-mudstone landslide and subsequent debris flow in Tianshui, China. Eng. Geol. 186 (2), 79–90. Peng, J.B., Lin, H.Z., Wang, Q.Y., Zhuang, J.Q., Cheng, Y.X., Zhu, X.H., 2014. The critical issues and creative concept in mitigation research of loess geological hazards. J. Eng. Geol. 22 (4), 684–691 (in Chinese). Sasitharan, S., Robertson, P.K., Sego, D.C., Morgenstern, N.R., 1993. Collapse behavior of sand. Can. Geotech. J. 30, 569–577. Sassa, K., Fukuoka, H., Wang, G., Ishikawa, N., 2004. Undrained dynamic-loading ringshear apparatus and its application to landslide dynamics. Landslides 1 (1), 7–19. Sassa, K., Wang, G., Fukuoka, H., 2003. Performing undrained shear tests on saturated sands in a new intelligent type of ring shear apparatus. Geotech. Test. J. ASTM 26 (3), 257–265. Scheidegger, A.E., 1973. On the prediction of the reach and velocity of catastrophic landslides. Rock Mech. 5, 231–236. Skempton, A.W., 1985. Residual strength of clays in landslides, folded strata and the laboratory. Geotechnique 35 (1), 3–18. Wang, G., Zhang, D., Furuya, G., Yang, J., 2014b. Pore-pressure generation and fluidization in a loess landslide triggered by the 1920 Haiyuan earthquake, China: a case study. Eng. Geol. 174, 36–45. Wang, G.H., 2000. An Experimental Study on the Mechanism of Fluidized Landslide: With Particular Reference to the Effect of Grain Size and Fine-Particle Content on the Fluidization Behavior of Sands. Phd. Thesis. Kyoto University. Wang, G.L., Li, T.L., Xing, X.L., Zou, Y., 2015. Research on loess flow-slides induced by rainfall in July 2013 in Yan'an, NW China. Environ. Earth Sci. 73 (12), 7933–7944. Wang, J.J., Liang, Y., Zhang, H.P., Wu, Y., Lin, X., 2014a. A loess landslide induced by excavation and rainfall. Landslides 11 (1), 141–152. Wilson, J.P., Gallant, J.C., 2000. Terrain Analysis. Principles and Applications. Wiley, New York. Xu, L., Dai, F., Tu, X., Tham, L.G., Zhou, Y., Iqbal, J., 2014. Landslides in a loess platform, North-West China. Landslides 11 (6), 993–1005. Xu, L., Dai, F.C., Gong, Q.M., Tham, L.G., Min, H., 2012. Irrigation-induced loess flow failure in Heifangtai Platform, North-West China. Environ. Earth Sci. 66, 1707–1713. Xu, L., Dai, F.C., Kang, G.L., Tham, L.G., Tu, X.B., 2009. Analysis of some special engineeringgeological problems of loess landslide. Chin. J. Geotech. Eng. 31 (2), 288–292 (in Chinese). Xu, Z.J., Lin, Z.G., Zhang, M.S., 2007. Loess in China and loess landslides. Chin. J. Rock Mech. Eng. 26 (7), 1297–1312 (in Chinese). Zhang, D., Wang, G., 2007. Study of the 1920 Haiyuan earthquake-induced landslides in loess (China). Eng. Geol. 94 (1), 76–88. Zhang, D., Wang, G., Luo, C., Chen, J., Zhou, Y., 2009. A rapid loess flowslide triggered by irrigation in China. Landslides 6 (1), 55–60. Zhang, F.Y., Wang, G.H., Kamai, T., Chen, W.W., Zhang, D.X., Yang, J., 2013. Undrained shear behavior of saturated loess at different concentrations of sodium chlorate solution. Eng. Geol. 155 (1), 67–69. Zhang, M., Yin, Y.P., 2013. Dynamics, mobility-controlling factors and transport mechanisms of rapid long-runout rock avalanches in China. Eng. Geol. 167 (1), 37–58. Zhang, M.S., Liu, J., 2010. Controlling factors of loess landslides in western China. Environ. Earth Sci. 59, 1671–1680. Zhuang, J.Q., Cui, P., Wang, G.H., Chen, X.Q., Iqbal, J., Guo, X.J., 2015. Rainfall thresholds for the occurrence of debris flows in the Jiangjia Gully, Yunnan Province, China. Eng. Geol. 195, 335–346. Zhuang, J.Q., Peng, J.B., 2014. A coupled slope cutting-a prolonged rainfall-induced loess landslide: a 17 October 2011 case study. Bull. Eng. Geol. Environ. 73 (4), 997–1011. Zhuang, J.Q., Peng, J.B., Wang, G.H., Javed, I., Wang, Y., Li, W., Xu, Q., Zhu, X.H., 2016. Prediction of rainfall-induced shallow landslides in the Loess Plateau, Yan'an, China, using the TRIGRS model. Earth Surf. Process. Landf. http://dx.doi.org/10.1002/esp. 4050.