Original Paper Landslides DOI 10.1007/s10346-014-0545-2 Received: 30 April 2014 Accepted: 16 December 2014 © Springer-Verlag Berlin Heidelberg 2015
Fujun Niu I Jing Luo I Zhanju Lin I Jianhong Fang I Minghao Liu
Thaw-induced slope failures and stability analyses in permafrost regions of the Qinghai-Tibet Plateau, China
Abstract The distribution of permafrost-related slope failures along the Qinghai-Tibet Highway from Wuddaoliang to Fenghuoshan correlates with ice content, slope gradient, and ground temperature. Slope failures are of two types. (1) Retrogressive thaw slumps result from icy permafrost being exposed by either man-induced excavation or fluvial-thermal erosion and undercutting of basal slopes. (2) Active-layer-detachment failures are caused by thaw of icy permafrost at the active layer-permafrost interface. After initial failure, active-layer-detachment failures can lead to retrogressive thaw-slumping and localized surficial landslide. Common trigger mechanisms for failure include high summer air temperatures and heavy summer precipitation. A third possible trigger mechanism for slope failure is earthquake occurrence. A geotechnical slope stability analysis was undertaken for an active-layer-detachment failure that had progressed into a retrogressive thaw slump. A safety factor (Fs) of 1.24 for the natural slope was determined using in situ tested strength parameters. However, the slope would lose stability when either the groundwater level over the permafrost table exceeded 1.42m or seismic acceleration reached, or exceeded, 0.03g. Keywords Permafrost . Active-layer-detachment failure . Retrogressive thaw slumping . Stability assessment . Qinghai-Tibet Plateau Introduction Most permafrost on the Qinghai-Tibet Plateau (QTP) is continuous and characterized by warm ground temperatures (−1 to 0 °C) (Zhou et al. 2000). The region has experienced noticeable warming during the past 40–50 years (Cheng and Wu 2007); for example, both mean annual ground temperature (MAGT) and active-layer thickness have increased in the last two decades (Jin et al. 2006, 2008). These changes can lead to slope instability and failure. Such failures generally occur in the active layer in summer. They are triggered by the formation of a shear zone near the base of the active layer that results from thaw of the ice-rich zone at the active layer-permafrost interface. Thawing is usually attributed to either high summer air temperatures or extreme precipitation events. The QTP is also an earthquake-prone region and seismic activity can accelerate the possibility of slope instability. For example, the Kunlun Mountain earthquake (magnitude 8.1) on November 14, 2001 attracted significant attention from engineers and scientists. It follows therefore that an assessment of slope stability is essential for engineering planning and implementation. Here, we describe thaw-induced slope failures that have occurred within a 10-km-wide zone along the Qinghai-Tibet Highway (QTH), from Wudaoliang (35° 13′ N, 93° 4′ E) in the north to Fenghuo Mountain (34° 43′ N, 92° 53′ E) in the south (Fig. 1). The zone includes the Hoh Xil Hill region, the Beiluhe Basin region, and the Fenghuo Mountain region. Approximately 85 % of the
study area is higher than 4500 m above sea level (ASL). The QTH, the Qinghai-Tibet Railway (QTR), and other linear engineering facilities are located along this corridor (Fig. 2). The objectives are to (1) summarize the distribution and morphological characteristics of thaw-induced slope failures, (2) investigate the thermal regime and failure development of a typical failure, (3) determine shear strength parameters in the basal zone of the active layer using in situ direct shear tests, and (4) assess slope stability in both its natural state and under the influence of an earthquake. Study area Data from the meteorological station in Wudaoliang town (Fig. 2) shows that both mean annual air temperature (MAAT) and precipitation have increased during the past 30–50 years (Fig. 3). The MAAT in the study area is lower than −4 °C, while the extreme low is about −30 °C at the end of January and highest is approximately 25 °C in July. The annual precipitation, about 300 mm, mostly falls between May and August. The active-layer thickness ranges between 1.5 and 3.0 m and near-surface permafrost may contain ice that exceeds 20 % by volume; this characterizes about 80 % of the permafrost in the study area (Liu and Wu 2000; Wu et al. 2002, 2004). The MAGT monitored at a depth of 15 m varies between −3.5 and −0.5 °C, with lowest values in Fenghuoshan and highest in the Beiluhe Basin. Soil in the study area is mainly loose alpine steppe and alpine meadow; these support a sparse vegetation cover that is dominated by monotone community physiognomy, characterized by low stature (10 to 15 cm), short growth period, and low biomass (Niu et al. 2011). Thaw-induced slope failures Types and morphological characteristics A total of 42 slope failures were recognized using SPOT-5 imagery and field investigation within the 10-km lateral zone along the QTH from Wudaoliang to Fenghuo Mountain pass. The failures are classified as being either active-layer-detachment slides (failures) or retrogressive thaw slumps (see French 2007). They are well known in the North American literature and have been described from forested, tundra, and high Arctic environments (Burn and Lewkowicz 1990; Lewkowicz and Harris 2005a, b). Retrogressive thaw slumps are triggered by the thaw of permafrost and the melt of exposed ground ice. This then leads to the retreat of a steep frozen back scarp, or headwall. Thaw slumps along the Qinghai-Tibet Engineering Corridor (QTEC) are usually initiated following excavation of material for use in construction of local infrastructures (Ma et al. 2006; Niu et al. 2012). They generally have a width of 20 to 90 m and a length of 40 to 150 m and occur at concave slope segments. Most are arc-shaped and usually consist of a series of slumped arcuate blocks. Compression Landslides
Original Paper
Fig. 1 Location of the study area on the Qinghai-Tibet Plateau. Permafrost underlies 75 % of the total area of the plateau. The study site is in the continuous permafrost zone. (Source: Environmental and Ecological Science Data Center for West China, National Natural Science Foundation of China. http://westdc.westgis.ac.cn/data)
ridges in their lateral and toe zones cannot be observed, and the sliding distances are often only a few meters. For example, Fig. 4a, b shows a typical thaw slump on a NW facing 7° slope developed at K3035 mile of the QTH. The actively retreating headwall is shown in Fig. 4c. Behind the headwall, a series of fissures can be observed. Once slumping has occurred, newly collapsed material pushes earlier collapsed blocks and forms a draped and superposed structure in the failure zone (Fig. 4d). The thaw depth in the failure area is about 2.2 m, slightly higher than the 2.0 m measured in undisturbed natural ground. In the case of active-layer-detachment failures, the active layer may slide over the permafrost table for distances as much as several tens of meters; they constitute localized surficial landslides. As such, these can be larger than thaw slumps, with widths of 80 to 160 m and lengths of 120 to 280 m. The mean slope angle at failure sites was 7.6°, much lower than that for landslides in nonpermafrost regions of China (Huang 2007). These types of slope failure are usually arc-shaped or bell-shaped, and often with vertical headwalls and compression ridges across the toe zone. After failure, and due to thaw of the icy layers at the active layerpermafrost interface, the headwall may develop into a retrogressive thaw slump. Geometric measurement Morphometric variables including horizontal width and length, slope angle, and length-to-width ratio were calculated for each slope failure using satellite imagery validated with field observations. Both the widths and lengths of slope failures showed considerable variability. The mean width and length of the thawinduced landslides were 1.5–2.0 times larger than those of the thaw slumps. The length-to-width ratios of the features were commonly between 1 and 3, which demonstrates that the slope failures in the study area were dominantly compact in morphology rather than elongated (Fig. 5a). The long axis orientation was defined as the direction toward the head scarp of the feature. The slope failures are commonly Landslides
oriented in the NW direction (Fig. 5b), indicating that they developed on SE facing slopes. Since ground ice content does not vary significantly with aspect, the main reason for the observed distribution is probably the greater amount of solar radiation received by SE facing slopes. However, other studies (Lewkowicz and Harris 2005b; Wang et al. 2009) from the Mackenzie Valley, Canada, have shown that the majority of slope failures are concentrated on NE facing slopes where higher ice contents are found, indicating that in this environment, solar insolation was not a primary factor influencing the occurrence of slope failures. Spatial distribution of thaw-induced slope failures The spatial distribution of slope failures appears to be closely related to ice content, slope gradient, and ground temperature (Fig. 6). As regards to ice content, statistical analysis shows that the vast majority (94 %) of slope failures occur in areas of icy permafrost or where massive ground ice occurs. As regards to slope gradient, failures occur predominantly on slopes of moderate angle (Fig. 6); for example, about 88 % of the failures occur on slopes with gradients of between 6° and 10°. However, the occurrence of slope failures in permafrost-free regions is positively correlated with slope gradient (Akgun 2012; Günther et al. 2013). As regards to ground temperature, about 40 % of failures occur in areas where the MAGT is above −0.5 °C, 32 % occur in areas with MAGT of between −1.0 and −0.5 °C, and 28 % occur in areas where the MAGT is lower than −1.0 °C. The slope failure at K3035W A typical active-layer-detachment failure that developed into a localized surficial landslide was selected for detailed study. It was first examined on October 12, 2010. It is located on the west side of K3035 mile of QTH (34° 59′ N, 92° 58′ E). To distinguish it from the K3035 thaw slump, we refer to it as the K3035W landslide. An aerial view, taken on March 27, 2010
Fig. 2 Topographic map of the study area
(Fig. 4f), shows tensional fissures (red dash line). However, the landslide occurred sometime before October 12, 2010, resulting in wheel ruts visible in the landslide mass that had moved downslope 28 m (Fig. 4g). Therefore, the landslide probably occurred at the end of September 2010 when the thaw depth reached its maximum. The size of this failure is about 230 m in length from the back scarp to the edge of the compressed area, and 90 m wide in its middle part. A 1.8-mhigh headwall formed after failure occurred; this initiated retrogressive thaw slumping; retreat of about 9 m between 2011 and 2013 was monitored. The total volume of material moved was 22,500 m3, several shear fissures formed on the surface (Fig. 4e), and compression ridge of about 30–50 cm high formed at the toe zone (Fig. 4h). According to borehole data obtained in September 2012, the thaw depth was 1.8 m in the undisturbed natural conditions, 1.0 m in the failure plane, and 2.1 m in the main displaced mass.
Thermal regime In August 2012, thermistors were installed within three 15-m-deep boreholes (No. 1, 3, and 4) and one 10-m-deep borehole (No. 2) along the major axis of the K3035W landslide (Fig. 7). Borehole No. 1 is located in undisturbed terrain, borehole No. 2 in recently disturbed terrain close to the foot of back scarp, borehole No. 3 in the area between the scarp and the sliding mass at a distance of 30 m from the back scarp, and borehole No. 4 was drilled in the main sliding mass. Each thermistor string consisted of 20 thermistors at 0.5-m interval from the surface down to 10 m and 5 thermistors at 1.0-m interval below 10-m depth. The temperature measurement accuracy is ±0.02 °C. The data acquisition system is made up of a CR3000 data logger and a solar panel recharging a 12-V battery (Fig. 7). Figure 8 plots MAGT against depth from September 2012 to September 2013 for the four boreholes. Due to the thermal disturbance caused by the landslide, the MAGT in the sliding mass area Landslides
Original Paper
Fig. 3 Climate changes in the study area in the past 50 years. a MAAT in the past 50 years; b mean annual precipitation in the past 30 years
was slightly higher than that in the undisturbed area, and this difference mainly existed between the 2–6-m depth. The maximum differences in MAGT were about 1.3 °C (between the scar area and undisturbed terrain), and 0.5 °C (between recently disturbed terrain and undisturbed terrain). Such differences were caused by the thinner active layer in the failure area than that in the undisturbed area during the initial period of the landslide occurrence. In order to further investigate the thermal regime of the landslide, ground temperature isotherms as a function of depth and time, from September 2012 to October 2013, are plotted for the undisturbed terrain (Fig. 9a), the recently disturbed terrain (Fig. 9b), the scar area (Fig. 9c), and the sliding mass (Fig. 9d). As might be expected, the depth to the permafrost table in the recently disturbed area and the scar area was significantly less than that in undisturbed terrain; in 2012, the thaw penetration depth reached only about 100 cm in recently disturbed terrain and the scar area. However, the thickness of the active layer increased in 2013 and reached 150 cm in depth. With time, it is probable that the thickness of the active layer will eventually reach, or even exceed, the active-layer thickness in the adjacent undisturbed terrain. Failure process In order to investigate the retrogressive nature of the initial activelayer-detachment failure (K3035W landslide), 3D laser images of the area were acquired for October 2011, June 2012, October 2012, June 2013, and October 2013. Based on these images, morphological changes between 2011 and 2013 are shown in Fig. 10. The total collapsed area was 670 m2 from October 2011 to October 2012, while it was only 65 m2 from October 2012 to October 2013. This indicates that the rate of headwall retreat decreased in 2013. This was confirmed by a number of manual transects that monitored Landslides
retreat of the headwall from October 2011 to October 2013. Table 1 shows that the maximal headwall retreat from October 2011 to October 2012 was 7.4 m at transect No. 4, while maximal retreat from October 2012 to October 2013 was only 1.6 m, also at the same transect. In general, both the laser imagery and the transect monitoring indicate that the rate of headwall retreat was slower in the second year. However, more data over a longer time period are required before any average rate of headwall retreat can be stated. Geotechnical properties of the active layer Grain size analyses conducted on material from within the active layer indicated that the clay content varied significantly with depth. The highest amount, 48.7 %, was present in the layer closest to the permafrost table (Fig. 11). This is consistent with studies of active-layer-detachment failures in northern Canada (Lewkowicz and Harris 2005a). Therefore, high clay content near the permafrost table may be another important condition facilitating landslide initiation in permafrost regions. Atterberg limits and plasticity indices of soil samples were also measured (Table 2). Results indicate that liquid limits are consistent with plasticity indices. The maximal liquid limit and plasticity index was also found in soil closest to the permafrost table; specific values are 32.12 % and 12.25, respectively. In addition, the peak shear strengths of undisturbed soil samples collected from depths of 0.5, 1.0, and 1.7 m and the strength parameters tested by the direct shear test are given in Table 2. Shear strength of the basal zone Active-layer-detachment failures are generally caused by low shear strength in the basal zone of active layer. Knowledge of the strength parameters of the material that comprises the active layer
Fig. 4 Photographs of slope failures in the study area. a A typical thaw slump developed at K3035 mileage of the QTH; b satellite image of the thaw slump on March 27, 2010 (acquired from Google Earth: http://www.google.com/earth); c the head scarp of the thaw slump; d the draped and superposed structure in the slumped material; e a typical active-layer-detachment slide located at the west side of K3035 mileage of the QTH; f satellite image of the failure site on March 27, 2010 (acquired from Google Earth); g satellite image of the failure on October 12, 2012 (acquired from Google Earth); h compression ridges developed at the toe zone of the failure. The red arrow represents the direction of mass movement
is essential for accurate stability assessment. Therefore, in situ shear tests were conducted at a location near the headwall of the
K3035W landslide. The apparatus used, shown in Fig. 12, consisted of vertical loading devices, horizontal loading devices, and
Fig. 5 The morphological characteristics of the slope failures in the study area. a Histograms of the length-to-width ratios, b long axis orientation (0°=north, 90°=east, 180°=south, 270°=west)
Landslides
Original Paper
Fig. 6 Statistical relations between the distribution of slope failures and the ice content, slope gradient, and MAGT (measured at a depth of 15 m from the year 2005 to 2013). Permafrost is classified as ice-poor when ice content is less than 20 %, ice-rich when ice content is between 20 and 50 %, and massive ground ice when ice content is higher than 50 % (Niu et al. 2002)
Fig. 7 Plan and profile views of the K3035W landslide. a Plan view, delineated landslide regions, and thermistor locations, and b slope profile and borehole stratigraphy
Landslides
Fig. 8 Mean annual ground temperature profiles from boreholes at the K3035W landslide site, September 2012 to September 2013
displacement measurement devices. The vertical and horizontal loads were applied by two lifting jacks with a maximal output force of 10 kN. The strength levels were measured by two pressure gauges, which were calibrated in the Lanzhou Jiaoda Engineering
Measurement Co., Ltd. before the testing. The vertical and horizontal displacements were measured by four dial indicators with a precision of 0.01 mm. Two sets of tests, including eight samples, were undertaken. The dimensions of each sample were 50×50 cm in length and 30 cm in height. In order to reduce disturbance, each sample was manually exposed by shovel after removal of the top soil layers. The exposed frozen surface around each sample (i.e., the permafrost table) was covered by a 2-cm-thick layer of sand to reduce quick thawing and to prevent the shear box from plugging into the icy surface. The small inter-space between the sample and the shear box was also filled with fine sand. In addition, the ice content at the active layer-permafrost interface beneath each sample was measured after testing. The volumetric ice contents from beneath of all eight samples varied from 75 to 80 %; this guaranteed little influence of ice content on the test results. During testing, five levels were loaded at 5-min time intervals. Vertical displacement was measured every 5 min. Horizontal stress was applied when vertical displacement did not exceed 0.01 mm in 5 min. Before horizontal loading started, the maximal shearing strength was evaluated and averaged. The peak horizontal stress was recorded when the sample slipped, and residual stress was acquired at the time when horizontal displacement kept a constant value. The strength parameters in the basal zone of active layer were obtained based on the Mohr-Coulomb theory (Fig. 13) and are listed in Table 3. In order to guarantee the test result accuracy, we used the mean values of the shear strength parameters for subsequent slope stability calculation.
Fig. 9 Two-dimensional ground temperature fields for the boreholes at the K3035W landslide. a Undisturbed area away from the slide, b recently disturbed area, c scar area, and d displaced mass area. The thick red line represents the 0 °C isotherm
Landslides
Original Paper
Fig. 10 The location of the head scarp at different times. The transects (No. 1–8) for monitoring the head scarp retreat are also shown. The green numbers refer to the distance from the scanner to each transect
Table 1 Head scarp retreat distances along eight transects (see Fig. 10 for the location of each transect)
Date (year-month)
Retrogressive distance (m) No.1 No.2
No.3
No.4
No.5
No.6
No.7
No.8
2011-10
0
0
0
0
0
0
0
0
2012-06
0.5
1.0
0.9
0.5
0.4
0
1.3
0
2012-10
1.3
3.1
2.5
7.4
2.3
3.0
3.2
3.5
2013-06
1.3
3.1
2.5
7.9
2.5
3.0
3.2
3.5
2013-10
1.5
3.1
2.5
9.0
3.5
3.1
3.2
3.5
Fig. 11 a the stratigraphy of the active layer and top of permafrost, and b grain size distribution of soils from the test sites
Landslides
Table 2 Geotechnical properties of the active layer at the K3035W landslide site
Depth (m)
Moisture content (%)
Liquid limit (%) –
Plasticity index
Shear strength φ
c
0.5
17.02
–
28.2
0
1.0
18.18
23.54
6.98
24.1
20.6
1.5
23.65
27.32
8.29
–
1.7
36.65
32.12
12.25
33.5
– 28.7
Fig. 12 Pictures of the in situ strength tests. a A sample on the surface of the massive ground ice, and b test setup
Slope stability in natural conditions Most thaw instability studies in permafrost regions adopt an infinite slope model because the thaw of icy permafrost will lead to high pore water pressure at the base of active layer (McRoberts and Morgenstern 1974). This is generally considered to be the main trigger for slope failure in permafrost regions (Savigny et al. 1995). Therefore, slope stability should be assessed using effective stress analysis. Slope stability can be estimated from the following equation, assuming that seepage direction is parallel to the ground ice surface (Niu et al. 2005): Fs ¼
c0 þ f½ð1−mÞγ þ mðγ sat −γ w Þgzcos2 βtanφ0 ½ð1−mÞγ þ mγ sat zsinβcosβ
ð1Þ
where Fs is the safety factor, c′ is effective cohesion, φ′ is effective frictional angle in the basal zone of active layer, β is the angle of slope, γsat is unit weight of saturated soil, γ is unit weight of the soil above groundwater level, γw is unit weight of water, z is vertical depth of the slip surface, and m is the ratio of the height
of the water table above the slip surface to the depth of the slip surface (z). In undisturbed terrain at the site of the K3035W failure, the angle of slope is 7.5°, the unit weight of saturated soil and soil above the groundwater level are 21.66 and 18.84 kN/m3, respectively, the depth to the permafrost table is 1.8 m, and the level of the water table above the permafrost table was 0.8 m at the end of September 2013. Using the strength parameters of the basal zone obtained from the in situ shear tests (Table 3), the calculated safety factor (Fs) of the slope was 1.24 when thaw depth reached its maximum, indicating that the slope is stable in its natural status. Three factors trigger the instability of slopes in permafrost regions. First, extreme high air temperature in summer, or forest fires that reduce the thickness of insulating layer, can lead to an increase in pore water pressure near the permafrost table (Savigny et al. 1995; Lewkowicz and Harris 2005a). Second, persistent rainfall can elevate the water table in the warm season. The calculated result indicates that the safety factor would reach its critical value when the water table level increased to 1.42 m. Third, the
Table 3 In situ shear strength parameters in the basal zone of active layer (φ′ and φr′ are effective peak and residual angle of friction, c′ and cr′ are peak and residual cohesion)
φ′ (°)
c′ (kPa)
φr′ (°)
cr′ (kPa)
Test 1
10.8
0.51
9.3
0
Test 2
11.2
0.55
9.8
0
Mean
11.0
0.53
9.6
0
Landslides
Original Paper
Fig. 13 Horizontal stress-displacement curves for a test 1 and b test 2
occurrence of an earthquake will result in an instantaneous increase in the sliding force and pore water pressure. This last trigger factor is discussed below. Slope instability under the influence of earthquake In order to assess slope instability on the QTP under an earthquake influence, the Fs in the conditions of static equilibrium is given as follows: Fs ¼
c0 þ f½ð1−mÞγ þ mγ sat cos2 β−a½ð1−mÞρ þ mρsat sinβcosβ−mγ w cos2 βÞgztanφ ½ð1−mÞγ þ mγ sat zsinβcosβ þ a½ð1−mÞρ þ mρsat zcos2 β
ð2Þ where the Fs, c′, φ′, β, γsat, γ, γw, z, and m have been mentioned above, a is the seismic acceleration, and ρsat and ρ are the densities of, respectively, saturated soil and soil above groundwater level (2.21 and 1.92 g/cm3 in the K3035W landslide site). Based on Eq. 2, the critical value of seismic acceleration for Fs=1 is 0.03g (g=9.8 m/s2) in the K3035W failure site. According to the Seismic Ground Motion Parameter Zonation Map of China (GB 18306-2001), the seismic peak ground acceleration is 0.1g in the study area; this is sufficient to promote slope instability of a site with the same conditions as the K3035W failure site. Furthermore, the above calculation does not consider the influence of pore water pressure; this will increase when an earthquake occurs and will result in more unstable conditions. Therefore, the slope may reach its failure threshold even when seismic acceleration is less than 0.03g. Conclusions Thaw-induced slope failures are widespread on the QTP, especially along the QTEC. The distribution of failures is closely related to the occurrence of icy permafrost, slope gradients of between 6° and 10°, and mean annual ground temperatures (MAGT) that are higher than −0.5 °C. Slope failures in the study area are classified as being either thaw slumps or active-layerdetachment failures that incorporate retrogressive thaw slumping of the headwall. Monitoring of a typical failure indicates that the MAGT and depth to the permafrost table in the failure area are lower than similar values in undisturbed terrain. Headwall retreat is highest in the initial years after failure. The shear strength parameters in the basal zone of the active layer were determined by in situ shear tests. Slope stability analysis for the K3035W failure site was Landslides
undertaken using in situ tested strength parameters. The slope at the K3035W site appears to be stable in its undisturbed state but could reach failure threshold when either the groundwater level in the active layer is higher than 1.42 m or earthquake acceleration reaches, or exceeds, 0.03g. Acknowledgments This work was supported by the Western Project Program of the Chinese Academy of Sciences (KZCX2-XB3-19), the State Key Development Program of Basic Research of China (973 Plan, 2012CB026101), and the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (Grant No. 41121061). The authors are indebted to Professor Fuchu Dai, Institute of Geology and Geophysics, Chinese Academy of Sciences, and Dr. Sergey Marchenko, University of Alaska Fairbanks, for constructive comments. Professor Hugh French, Emeritus Professor, University of Ottawa, undertook final editing.
References Akgun A (2012) A comparison of landslide susceptibility maps produced by logistic regression, multi-criteria decision, and likelihood ratio methods: a case study at İzmir, Turkey. Landslides 9(1):93–106 Burn CR, Lewkowicz AG (1990) Retrogressive thaw slumps. Can Geogr 34(3):273–276 Cheng GD, Wu TH (2007) Responses of permafrost to climate change and their environmental significance, Qinghai-Tibet Plateau. J Geophys Res 112:F02S03. doi:10.1029/2006JF00063 French HM (2007) The periglacial environment, 3rd edn. Wiley, UK, p 458 Günther A, Reichenbach P, Malet JP, Eeckhaut MVD, Hervás D, Dashwood C, Guzzetti F (2013) Tier-based approaches for landslide susceptibility assessment in Europe. Landslides 10(5):529–546 Huang RQ (2007) Large-scale landslides and their sliding mechanisms in China since the 20th century. Chin J Rock Mech Eng 26(3):433–454 (in Chinese) Jin HJ, Zhao L, Wang SL, Jin R (2006) Modes of degradation and thermal regimes of permafrost along the Qinghai-Tibet Highway. Sci China D 49(11):1170–1183, in Chinese Jin HJ, Yu QH, Wang SL (2008) Changes in permafrost environments along the QinghaiTibet engineering corridor induced by anthropogenic activities and climate warming. Cold Reg Sci Technol 53(3):317–333 Lewkowicz AG, Harris C (2005a) Morphology and geotechnique of active-layer detachment failures in discontinuous and continuous permafrost, northern Canada. Geomorphology 69(1):275–297 Lewkowicz AG, Harris C (2005b) Frequency and magnitude of active-layer detachment failures in discontinuous and continuous permafrost, northern Canada. Permafrost Periglac 16(1):115–130
Liu YZ, Wu QB (2000) Study on ground temperature field in permafrost regions of Qinghai-Tibet Plateau. Highway 2(1):4–8 (in Chinese) Ma W, Niu FJ, Satoshi A, Jin DW (2006) Slope instability phenomena in permafrost regions of Qinghai-Tibet Plateau, China. Landslides 3(3):260–264 McRoberts EC, Morgenstern NR (1974) Stability of slopes in frozen soil, Mackenzie Valley, N.W.T. Can Geotech J 11(4):554–573 Niu FJ, Zhang JM, Zhang JZ (2002) Engineering geological characteristics and evaluations of permafrost in Beiluhe testing field of Qinghai- Tibetan Railway. J Glaciol Geocryol 24(3):264–269 (in Chinese) Niu FJ, Cheng GD, Ni WK, Jin DW (2005) Engineering-related slope failure in permafrost regions of the Qinghai-Tibet Plateau. Cold Reg Sci Technol 42(3):215–225 Niu FJ, Lin ZJ, Liu H, Lu JH (2011) Characteristics of thermokarst lakes and their influence on permafrost in Qinghai-Tibet Plateau. Geomorphology 132(3):222–233 Niu FJ, Luo J, Lin ZJ, Ma W, Lu JH (2012) Development and thermal regime of a thaw slump in the Qinghai-Tibet plateau. Cold Reg Sci Technol 83–84:131–138 Savigny W, Logue C, MacInnes K (1995) Forest fire effects on slopes formed in ice-rich permafrost soils Mackenzie Valley, Northwest Territories. 48th Canadian Geotechnical Conference, Vancouver. Canadian Geotechnical Society, Vancouver, pp 989–998 Wang BL, Paudel B, Li HQ (2009) Retrogression characteristics of landslides in finegrained permafrost soils, Mackenzie Valley, Canada. Landslides 6(2):121–127 Wu QB, Liu YZ, Tong CJ (2002) Interactions between the permafrost and engineering environments in the cold regions. J Eng Geol 8(3):281–287 (in Chinese) Wu QB, Cheng GD, Ma W (2004) The impact of climate warming on Qinghai-Tibetan Railroad. Sci China D 47(Supp I):122–130, in Chinese
Zhou YW, Guo DX, Qiu GQ (2000) Permafrost in China. Science Press, Beijing, pp 106–404 (in Chinese)
F. Niu : J. Luo : Z. Lin : M. Liu State Key Laboratory of Frozen Soil Engineering, Cold and Arid Regions Environmental and Engineering Research Institute, Chinese Academy of Sciences, No. 320, West Donggang Road, Lanzhou, Gansu 730000, China F. Niu Key Laboratory of Highway Construction & Maintenance Technology in Permafrost Region, Ministry of Transport, CCCC First Highway Consultants Co., Ltd, Xi’an, 710075, China J. Luo ()) : M. Liu University of Chinese Academy of Sciences, Beijing, 100049, China e-mail:
[email protected] J. Fang Qinghai Research Institute of Transportation, Xining, 810001, China
Landslides