In seasonally frozen regions, snowmelt-induced landslides are becoming a common geomorphological process. However, their failure mechanism are not yet well understood. This study focuses on the Zeketai River Basin in the Ili Basin of Ten-zan, China, and attempts to explain the deformation process and failure mechanism. The results revealed that the 242 loess landslides showed obvious season distribution and aspect differentiation. The landslide probability density-area distribution was well in line with the three-parameter inverse-gamma probability distribution. Long-term snowmelt infiltration and repeated soil freezing-thawing affected the slope surface and subsurface hydrothermal and mechanical equilibrium. The combined effect of static liquefaction of sliding surface and sliding liquefaction at slope toe was a hybrid mechanism for slope deformation and failure. On exposed landslide slope thicker snowpack and faster snowmelt contributed to persistent deformation of this landslide. In the future, climate warming would increase snowmelt-driven slope failures in seasonally frozen regions. More careful consideration and discussion regarding the thermal-hydraulic and mechanical processes and early warning of snowmelt-driven landslides should be strengthened.
Research Article
Response of snowmelt-driven loess landslides to climate change in Ten- zan, north-west China
https://doi.org/10.21203/rs.3.rs-4267047/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Loess landslide
Snowmelt
Freezing-thawing
Liquefaction
Failure mechanism
The landslides were rapid geological and geomorphological processes that had an important impact on the evolution of regional topography (Broeckx et al., 2020). The continued increase of landslides in mountainous areas was closely linked to climate change (Durand et al.,2009; Huggel et al., 2012; Saez et al.,2013; Hinds et al.,,2019). Most landslides were mainly triggered through precipitation around the world (Suradi et al., 2016; Huang et al., 2020), but slope instabilities in relation to the snowmelt process influenced by temperature were also more frequent (Saez et al., 2013; Shan et al., 2014; Subramanian et al., 2018). Snowmelt-induced landslides driven by extreme climate had gradually represented a common slope instability event in seasonally frozen regions (Cardinali et al. 2000; Guzzetti et al. 2003; Kawagoe et al. 2009; Trandafir et al., 2015). For example, a snowmelt event associated with an unusually warm 10-day period from late May through early June in 1983 triggered about 150 landslides in northern Utah (Anderson et al. 1984; Wieczorek et al. 1989). The 4233 landslide events once occurred in the Umbria region of central Italy was caused by sudden warming on 1 January 1997(Cardinali et al. 1999). The 1448 landslides in Niigata, Japan were triggered by abnormal warming during the snowmelt period from 2005 to 2006 (Akiyama et al.,2006; Matsuura et al. 2008). In recent years, some scholars had studied the impact of snowmelt on landslides (Nakazato et al., 2013; Kimura et al., 2016; Miao et al., 2019). Many studies showed that snowmelt was the main contributor to landslide disasters in frozen areas (Matsuura et al., 2013; Subramanian et al., 2020), and the occurrence of landslides were strongly linked to the gradual buildup of a snowmelt-generated perched water table within the slope (Ishikawa & Miura, 2011; Trandafir et al.,2015; Mori et al., 2017). Previous studies had focused on landslide events in other regions triggered by high summer temperatures and heavy precipitation, and their formation mechanisms were better understood (Xu et al.,2012; Zhuang et al., 2017). However, landslides in frozen areas were closely related to Winter blizzards, soil freeze-thaw cycles and snowmelt water infiltration (Kawamura & Miura, 2013; Subramanian et al., 2017). The triggering mechanism and evolution process of snowmelt-induced landslides was not well understood due to its complex thermos-hydraulic coupling processes.
Loess landslides were the product of dynamic geological processes in the evolution of unique loess landscape, which were controlled by factors such as topography, climate, lithology, earthquakes and human activities (Peng et al., 2018). The Ili valley was a typical seasonal frozen area with rain and snow, which was called a ‘wet island’ in the arid region of northwest China. The loess was widely distributed in the Ili basin of northwest China, and the loess origin was categorized as the "mountain source area-river transport" (MR) mode (Li, et al.,2020; 2021). Recently the landslide disasters during the snowmelt period had been increasing in the Ili loess area with a warming climate (Xian et al., 2022). Thus, it was an ideal area to gain a deeper understanding of the snowmelt-induced slope instability process.
Combining remote sensing data regarding landslide distribution, statistics concerning historical landslide events, multi-scale meteorological data analysis and loess physical-mechanical tests, the paper aims to ascertain the basic characteristics and spatio-temporal distribution of the loess landslides in the study area, to determine the influencing factors and evolution process, to explore the thermal-hydraulic-mechanical coupling mechanism, and to discuss the impact of climate change on snowmelt-driven landslides in the future. This study will make an important contribution to determine the critical period of slope deformation, establish early warning thresholds for landslide occurrence and promote effective management decisions to control landslide deformation and disaster mitigation.
2.1. Study area
The study area was located in the Zeketai River Basin with a drainage area of about 190km2 in the Ili Valley of Tianshan, China, which was distributed on the southern slope of the Avulalle Mountain (Fig. 1a). The length of the Zeketai River was 26km with a channel slope of 100–150‰, and the annual average runoff was about 21,830m3. The study area was a westerly controlled seasonal freezing zone. The annual average temperature and precipitation were around 8.3℃ and 477.5mm respectively observed in the Xinyuan Meteorological Station (928.2m above sea level). The freezing period lasted up to 5 months from November to March of the next year, and the temperature rose sharply from − 5.9℃ to 11.5℃ from February to April. The maximum depth of snow cover and frozen soil was about 100cm and 120cm, respectively. Additionally, according to data from the Tianshan Snow and Avalanche Research Station of the Chinese Academy of Sciences (1,776m above sea level), the annual average temperature and precipitation were around 1.4℃ and 807.3mm, respectively. From March to May the temperature rose sharply from − 3.3℃ to 9.1℃. The freezing period lasted up to 6 months from November to April of the following year. The maximum thickness of snow cover was about 120cm and the maximum depth of frozen soil was approximately 160cm. As the altitude increased, both temperature and precipitation in the study area showed obvious gradient changes.
The Zeketai River Basin was divided into two major geomorphic units (Fig. 1b). The geomorphic unit at elevations of 960–2,342m was a low-middle mountain structural denudation landform with the slope of 25–40°. The stratum lithology was mainly composed of granite porphyry, diorite, glutenite and mudstone. The surface was covered by a loess layer with a thickness of 5-30m. The geomorphic section with an elevation of 857∼960m was a piedmont alluvial plain and hilly landform, mostly comprising a single layer of loess with a maximum thickness of 80m. In the study area the loess was mainly distributed at an elevation of 800–2200m. Most of the loess is Quaternary Late Pleistocene Malan loess, and a small proportion was the Middle Pleistocene Lishi loess. Large amounts of loess provided an important material base for loess landslides.
2.2. Landslide investigation
Combined with statistical yearbooks, site surveys and disaster record of the highway maintenance department, the spatio-temporal distribution data of the landslide events with accurate records from 1990 to 2010 were obtained.
We catalogued the landslide events to obtain their location, scale and scope by using high-resolution images (resolution = 30m) and Google Earth image interpretation. The spatial distribution characteristics (including the slope, aspect and elevation) of the landslides were quantitatively analyzed based on digital elevation models (DEMs) from ASTER GDEM.
Site investigations were conducted to recognize landslide information by using a 1:50,000 topographic map and a 1:200,000 geological map, including geomorphic characteristics, stratigraphic structure, underground water and slope deformation, etc.
The undisturbed and remolded soil samples were taken from landslide sliding surface. The parameters of undisturbed loess such as water content, density, liquid plastic limit, and plastic index were measured. The triaxial shear tests of remolded soils with different moisture contents (10%,15%,20%,25% and 30%) were used to obtain axial compressive pressure and shearing strength at different confining pressures (100kpa,200kpa,300kpa and 400kpa) under freeze-thaw cycles (0,1,3,5,10,20,30).
2.3. Climate data
The meteorological and snow observation data was provided from the Xinyuan Meteorological Station (83°18′E, 43°27′N, 928.2m) and the Tianshan Snow and Avalanche Research Station of the Chinese Academy of Sciences (43°33′06″N, 84°00′28″E, 1,776m). The distance between the two stations installed in the Ili River Valley is about 100km. The monthly average temperature and precipitation data from 1990 to 2010 was obtained from the Xinyuan Meteorological Station. The dataset was from the Tianshan Snow and Avalanche Research Station of the Chinese Academy of Sciences (Li et al., 2022), including monthly snow depth, soil water content, soil surface temperature, atmospheric temperature and net radiation data from 2005 to 2010 as well as daily average snow depth, temperature and net radiation data during the snowmelt period in March.
2.4. Data analysis
The spatio-temporal of landslides was analyzed based on mathematical and statistical analysis, and the frequency-area relationship expressed by the three-parameter inverse-gamma probability density function proposed by Malamud et al. (2004). Comparison of landslide activity with climate change during the 1990–2010 were statistically analyzed. The response feedback of snow depth to atmospheric temperature and net radiation was explored, and further to analyze its influence on soil humidity and surface temperature based on meteorological data from 2005 to 2010. The tests about the axial compressive strength and shearing strength were used to analyze soil structure damage and strength reduction under freeze-thaw cycles, and further reveal the sliding liquefaction and punching and shearing damages at the foot of the loess slope.
3.1. Landslide distribution
(1) Spatio-temporal distribution
The 242 loess slope failures with total area of 1.6km2 were identified in the Zeketai River Basin (Fig. 1b). About 90% landslides were small to medium in size with an area range from 500m2 to 15,000m2. Landslides with an area larger than 15,000m2 accounted for about 7% (17), and only 3% (7) had an area smaller than 500m2. Deep landslides (more than 2m in depth) accounted for about 88%, and the sliding surfaces of 70% landslides were located at the contact surface of loess and underlying bedrock.
The loess landslides showed significant temporal distribution and topographic differentiation (Fig. 1c). Firstly, the distribution of the loess landslides exhibited an obvious aspect differentiation (Fig. 1d). 61% (148) of loess landslides occurred on the northward slope (the northern, north-western and north-eastern slopes). Only 36 (15%) landslides occurred on the south-eastern, southern and south-western slopes. Secondly, 40 loess landslides were recorded historically during 1990 to 2010, and an average of 2 landslides per year occurred. The occurrence frequency of landslides showed an increasing trend (Fig. 1e). Additionally, the loess landslides presented a clear seasonal distribution (Fig. 1f). 26 loess landslides occurred in Spring (from March to May), which far exceeded the average monthly frequency of 3.3 events.
(2) Frequency-area distribution
The relationship between the landslide probability density and area distribution was quantified statistically using logarithmic coordinates, and 500m2 was set as the area increment in this landslide catalogue. The maximum and minimum area of the landslides were 4.2×104m2 and 148.02 m2, and 80% of the landslide area was concentrated in the range of 1000–10000 m2. As the landslide area increased to a certain range (1,000–3,500m2), the probability density reached its maximum value (Fig. 1g). The frequency-area relationship in this landslide catalogue could also be expressed by the three-parameter inverse-gamma probability density function proposed by Malamud et al. (2004), as Eq. (1).
Where, 
was gamma function of parameter \(\rho\), \(\rho =2.5\), \({\alpha }=1.48\times {10}^{-2}{km}^{2}\), \(s=2.12\times {10}^{-3}{km}^{2}\), \({r}^{2}=0.906.\)
3.2. Climate-driven landslide events
During 1990 to 2010, 65% of the total 40 landslides occurred due to abnormal warming in Spring (Figs. 2a-c). Except for the extreme precipitation event in 2002, there was no clear correlation between precipitation and slope failure. In addition, the relationship between landslide frequency and precipitation was also not significant at other months, except for March-May (Fig. 2d). For example, a total of 12 landslides occurred in April at a rainfall of 77.2 mm, whereas only 2 landslides occurred when precipitation reached 74.7mm in September. It was obvious that the snowmelt infiltration during the melting season played an important role before the onset of slope failure.
3.3. Snow accumulation and melting
Snow depth varied continuously during the snow accumulation and melting process from October to April during 2005–2010 (Fig. 3). About 70% of the snow melting in the Tianshan Mountains occurred in Spring, and 75% of the snow cover accumulated in Winter (Hu et al., 2016). Snow depth and net radiation showed a strong negative linear relationship with a correlation coefficient of R2 = 0.8724 (Figs. 3a-b). The maximum snow depth occurred in February, and snowfall began to melt in March. The snow depth and atmospheric temperature also showed a corresponding relationship, but the change in snow depth appeared to be a hysteresis relative to temperature variation (Figs. 3c-d). The relationship between daily air temperature and snow depth in March was non-linear with a correlation coefficient of R2 = 0.755. Based on the data analysis in March from 2005 to 2010, a functional relationship among snow depth, net radiation and air temperature were established as Eq. (2) (Fig. 3e). The snow depth in March decreased linearly with increased net radiation and air temperature.
z = p1 + p2*x + p3*y^3 (2)
where, p1 = 0.8226, p2=-0.00498, p3=-0.00016, correlation coefficient R2 = 0.944, z represented snow depth (m), x represented net radiation (w/m2), and y represented air temperature (℃).
The snow depth and soil physical parameters including the surface temperature and water content maintained a negative peak-valley relationship (Figs. 3f-g). The water content of the soil reached a peak value of about 40% in late March, and the peak value of soil surface temperature appeared at about 6°C in April. Expect for the extreme blizzard in 2010, the minimum soil moisture content fluctuated around 20% annually. As the snow depth decreased, the soil water content and the surface temperature both increased non-linearly (Fig. 3h). The surface temperature began to exchange heat with the atmosphere when the snow depth reached 64–73cm. The exponential increase in soil water content was closely related to the snowmelt process from March to April. The volumetric water content of the soil increased to its maximum exactly at peak snowmelt period.
3.4. Soil freezing and thawing
The axial compressive strength of loess with different water contents decreased gradually after freeze-thaw cycles (Figs. 4a-d). During the first 10 freeze-thaw cycles, the strength reduction of loess was greater with an average strength loss of 20%, especially for the first freeze-thaw cycle, the axial stress decreased more significantly with a maximum strength loss of 12%. When 30 freeze-thaw cycles of loess were finished, the average total loss of strength reached about 25%.
In addition, the axial compressive strength of loess decreased with the increasing water content. The magnitude of strength reduction was greatest with 20–35% losses when the water content was 20–25%. When the water contents ranged between 10–20%, the change of internal friction angle was not obvious, and the angles of internal friction appeared to overlap each other with increasing freeze-thaw cycles (Fig. 4e). However, when the water content exceeded 25%, the angle of internal friction decreased significantly. The loess cohesion decreased with increasing water content and freeze-thaw cycles (Fig. 4f). During the first 5 freeze-thaw processes, the maximum decrease in loess cohesion occurred.
The relationship between the probability density and area distribution of landslides in the study area fitted well with the three-parameter inverse gamma distribution. For small landslides, the area distribution exhibited an exponential flip, whereas for medium-large landslides, the area distribution decayed with a negative power-law relationship with an exponent β = -3.5, which was greater than the decay exponent β = -2.4 for the earthquake or rainfall triggered landslides (Malamud et al., 2004a,b; Xu et al. 2014). Cardinali et al. (2000) found that in the Umbria region of central Italy the 4233 snowmelt-triggered landslides occurred with a decay exponent of β = -2.4. Stark et al. (2001) found the same power-law distribution pattern for large landslides using a double Pareto distribution, with a power exponent of -2.11 for the medium and large-scale landslides in Taiwan, and corresponding power exponents of -2.44 and − 2.48 for medium and large-scale landslides in New Zealand, respectively. Eeckhaut et al. (2007) found that β values between 1.4 and 3.5 were also reasonable for individual landslide events. In this landslide event, the anomaly in the decay index β was believed to relate to incomplete historical landslides occurring in multiple periods during many years, and the recorded landslides were not induced by a single rapid snowmelt event. In addition, the number and size of snowmelt-induced loess landslides also limited the power-law index that controlled the area distribution of large landslides.
The loess landslides in the study area were very closely related to snowmelt driven by anomalous warming during the melting season. Some studies also found snowmelt-driven landslide events were very clearly linked to snowmelt driven by anomalous spring warming(Durand et al.,2009; Saez et al., 2013; Xian et al.,2022). A well-developed yield zone along the sliding surface with progressive rise in the perched water table above the toe of the slide mass may explain the snowmelt-induced slope deformation mechanism prior to the onset of catastrophic failure (Wieczorek 1996; Ishikawa & Miura, 2011; Trandafir et al., 2015; Mori et al., 2017; Subramanian et al., 2020). However, the freeze-thaw cycles, as another hydro-thermophysical process in the seasonal freezing zone, also had a significant impact on slope deformation and instability. The seasonal permafrost experienced repeated freeze-thawing cycles during melting season (Norikazu and Hiroaki ,1999; Nadim et al.,2006). The effect of freeze-thaw action on slope mainly included the freezing and stagnation effect of groundwater inside the slope and the freeze-thaw cycle in the surface layer of the slope. In the freezing period, the soil frozen layer confined infiltrating water within the slope to form frozen stagnant water, which not only expanded the range of soil softening, but also raised the water table to generate hyperstatic pore water pressure at the sliding surface. And then, static liquefaction of the soil on the slip surface occurred under the combined effect of increased overlying slip weight and reduced strength of the soil, leading to slope creeping deformation. Static liquefaction due to shearing or increased water content could be observed after some landslides (Zhang et al. 2009; Xu et al. 2012). During the melting period, the soil strength at the foot of loess slope decreased from peak to residual strength after freeze-thaw cycles, and original frozen seepage points at the slope foot thawed to form large drainage channels, resulting in rapid drainage of large amounts of frozen stagnant water within the slope to form strong dynamic water pressures. The low axial compressive strength and shear resistance of loess after freeze-thaw cycles caused the saturated loess at the toe to present high mobility and sliding liquefaction, and the strong hydrodynamic pressure caused the softening collapse and punching and shearing damages at the foot of the loess slope. The combined effect of static liquefaction of sliding surface and sliding liquefaction at slope toe was a hybrid mechanism for loess slope deformation and failure.
Atmosphere-surface interactions affected snowmelt and soil freezing and thawing process (Subramanian et al., 2020), and long-term periodic snowmelt infiltration and repeated soil freezing and thawing both influenced the slope surface and subsurface hydrothermal and mechanical equilibrium (Xian et al.,2022). Slope was disturbed to form exposed slope after the landslides occurred, and numerous surface cracks formed at the back edge of the landslide. The loss of vegetation and ground disturbance not only increased snowpack, but also reduced shear strength and increased groundwater conductivity and connectivity (Huggel, 2009; Patton et al., 2019). Furthermore, the topographic depression formed by the landslide resulted in a thicker snowpack (Essery & Pomeroy, 2004). The thicker snowmelt and the resulting freezing and thawing reinforced the impact on the hydrological and mechanical balance of the exposed slope, further contributed to persistent deformation of this landslide. Long-term snowmelt runoff had a greater cumulative impact on soil seepage (Schulz et al., 2009; Hinds et al., 2019). Snow within the landslide persisted longer after snow on the adjacent hillslope had melted (Patton et al.,2021). Thicker snowpack and faster snowmelt on exposed landslide slopes would cause original landslide to present multi-phase, multi-stage and long-term failure processes.
With global warming, the climate in Northwest China is becoming warming and humidification with a high frequency of anomalous warming and extreme precipitation events (Yao et al., 2019;Yao et al., 2022), and snowmelt-induced landslides in seasonally frozen areas will become a common disaster phenomenon. In the paper, a hybrid mechanism for loess slope deformation and failure was proposed considering snowmelt infiltration and soil freezing and thawing cycles, which would contribute to the establishment and implementation of an early warning system to reduce the snowmelt-induced landslide risk (Fig. 5). An effective warning system for snowmelt induced landslides should incorporate various environmental factors such as temperature and rain-on-snow events, as well as variations in the characteristics of the snowpack such as density and thickness (Martelloni et al. 2013). Snow depth is an important feedback indicator of snow melting and accumulation, so the research on factors affecting snow depth can help predict snowmelt landslide trends. In the future, long-term observation and monitoring of snow depth and its influencing factors should be strengthened.
A total of 242 loess slope failures were identified in the Zeketai River Basin of Ili basin, China. The frequency-area relationship could be expressed by the three-parameter inverse-gamma probability density function. The 242 loess landslides showed significant season distribution and aspect differentiation. 61% of landslides were scattered on the northward slope, and 65% of recorded landslides (40 events) occurred in Spring during 1990–2010. The snowmelt-driven landslides were very closely related to abnormal warming during the melting season. Long-term periodic snowmelt infiltration and repeated soil freezing and thawing both influenced the surface and subsurface hydrothermal and mechanical equilibrium of slope instability. The combined effect of static liquefaction of sliding surface and sliding liquefaction at slope toe was a hybrid mechanism for slope deformation and failure. With future global warming, snowmelt-driven landslides driven by abnormal warming will become a more common geological disaster in cold and high-altitude areas. The atmospheric indicators in relation to snow depth change could be used as an effective tool for predicting future snowmelt-driven landslides. In the future, more careful consideration and discussion regarding the thermal-hydraulic and mechanical processes and early warning of snowmelt-driven landslides should be strengthened based on long-term field observations, indoor modeling experiments tests and numerical simulations, etc..
Author Contribution
xueli wei wrote the main manuscript text, and yu xian prepared figures.All authors reviewed the manuscript.
Acknowledgements
This research was supported by the Central Guidance on Local Science and Technology Development Fund of Sichuan Province (Grant No. 2022ZYD0022), Natural Science Foundation of Sichuan Province(Grant No.2023NSFSC2086)and Scientific Research Fund of Chengdu University of Information and Technology (Grant No. KYTZ2022126).
- Akiyama, K., Kurihara, J., Ogawa, K., 2006. Characteristics of slope failure induced by the Mid Niigata prefecture earthquake and the snow melting after the snow term. Disaster mitigation of debris flows, slope failures and landslides. Universal Academy Press Inc, Tokyo.
- Anderson L.R., Keaton J.R., Saarinen T., Wells W.G.,1984. The Utah landslides, debris flows, and floods of May–June 1983. National Research Council, Washington, DC, 96.
- Broeckx J., Rossi M., Lijnen K., Campforts B., Poesen J., Vanmaercke M., 2020. Landslide mobilization rates: A global analysis and model. Earth Sci. Rev. 201, 102972.
- Cardinali M., Ardizzone F., Galli M., Guzzetti F., Reichenbach P.,2000. Landslides triggered by rapid snow melting: the December 1996–January 1997 event in Central Italy. In: Claps P, Siccardi F (eds) Proceedings of the 1st Plinius Conference on Mediterranean Storms, Maratea, Italy, October 1999, European Geophysical Society. Bios Publisher, Cosenza, 439–448.
- Cardinali M., Ardizzone F., Galli M., Guzzetti F., Reichenbach P., 2000. Landslides triggered by rapid snow melting: the December 1996–January 1997 event in Central Italy. In: Proceedings of 1st EGS Plinius conference, 14–16 Oct 1999. Bios Publisher, Cosenza.439–448.
- Durand Y., Laternser M., Giraud G., Etchevers P., Lesaffre B., Mérindol L., 2009. Reanalysis of climate in the French Alps (1958–2002). J. Appl. Meteorol. Clim. 48, 429–449.
- Eeckhaut M. V. D., Poesen J., Govers G., Verstraeten G., Demoulin A., 2017.Characteristics of the size distribution of recent and historical landslides in a populated hilly region. Earth and Planetary Science Letters.256, 588-603.
- Essery, R., & Pomeroy, J.,2004. Vegetation and topographic control of wind-blown snow distributions in distributedand aggregated simulations for an Arctic Tundra basin. Journal of Hydrometeorology. 5, 735–744.
- Guzzetti F, Reichenbach P, Cardinali M, Ardizzone F, Galli M.,2003. The impact of landslides in the Umbria region, central Italy. Nat Hazards Earth Syst Sci. 3,469–486.
- Hinds, E., Lu, N., Mirus, B., et al., 2019. Effects of Infiltration Characteristics on Spatial-Temporal Evolution of Stability of an Interstate Highway Embankment. J.Geotech. Geoenvironmental Eng. 145(9), 05019008.
- Hu W., Liu H., Wang H., Zhao W., 2016. Analysis of the terrain effect on snow cover accumulating and melting in the Tianshan Mountains. J. Glaciol. Geocryol. 38(5), 1227-1232. (in Chinese) .
- Huang Q., Xu X., Kulatilake, P. et al.,2020. Formation mechanism of a rainfall triggered complex landslide in southwest China. J. Mt. Sci. 17, 1128–1142.
- Huggel C.,2009. Recent extreme slope failures in glacial environments: Effects of thermal perturbation. Quaternary Science Reviews. 28, 1119–1130.
- Huggel C., Clague J.J., and Korup O., 2012.Is climate change responsible for changing landslide activity in high mountains? EARTH SURF PROC LAND. 37, 77–91.
- Ishikawa T., Miura S., 2011. Influence of freeze-thaw action on deformation-strength characteristics and particle crushability of volcanic coarse-grained soils. Soils Found.51(5), 785-799.
- Kawagoe S., Kazama S., Sarukkalige P.R.,2009.Assessment of snowmelt triggered landslide hazard and risk in Japan. Cold Reg Sci Technol. 58,120–129.
- Kawamura S., Miura S., 2013. Rainfall-induced failures of volcanic slopes subjected to freezing and thawing. Soils Found. 53(3), 443-461.
- Kimura T., Katsura S., Maruyama K., ISHIDA K., 2016. Topographic features of source and transfer-deposition areas of long-travelling landslides induced by snowmelt. J. Jpn. Landslide Soc. 53(2),31–42.
- Li L.H., Hao J.S., Wang H.C., 2022.The snow observation dataset from the Tianshan Station for Snow Cover and Avalanche Research(1967-2021) [Dataset]. https://www.doi.org/10.12072/ncdc.SNOW.db2348.2022
- Li, Y., Song, Y., Fitzsimmons, K.E., Chen, X., Prud'homme, C., Zong, X., 2020. Origin of loess deposits in the North Tian Shan piedmont, Central Asia. Palaeogeography, Palaeoclimatology, Palaeoecology. 109972.
- Li, Y., Song, Y.G., 2021. Discussion of the paper “Loess genesis and worldwide distribution” by Yanrong Li, Wenhui Shi, Adnan Aydin, et al. Earth-Science Reviews. 221,103151.
- Malamud B.D., Turcotte D.L., Guzzetti F., Reichenbach P., 2004. Landslide inventories and their statistical properties. Earth Surf. Proc. Land. 29(6), 687-711.
- Martelloni G, Segoni S, Lagomarsino D, Fanti R, Catani F.,2013. Snow accumulation/melting model (SAMM) for integrated use in regional scale landslide early warning systems. Hydrol Earth Syst Sci. 17,1229–1240.
- Matsuura S., Asano S., Okamoto T.,2008. Relationship between rain and/or meltwater, pore-water pressure and displacement of a reactivated landslide. Eng. Geol.101(1-2), 49-59.
- Matsuura S., Okamoto T., Asano S., Matsuyama K., 2013. Characteristics of meltwater and/or rainfall regime in a snowy region and its effect on sediment-related disasters. B. Eng. Geol. Environ.72(1), 119-129.
- Miao F., Wu Y., Li L., Liao K, Zhang L., 2019. Risk assessment of snowmelt‑induced landslides based on GIS and an effective snowmelt model.Nat. Hazards. 97,1151–1173.
- Mori A., Subramanian S. S., Ishikawa T., Komatsu M., 2017. A Case Study of a Cut Slope Failure Influenced by Snowmelt and Rainfall. Procedia Eng. 189, 533-538.
- Nakazato H., Shoda D., Inoue K., Suzuki H.,2013. A case study of behavior observation of landslide induced by snowmelt after an earthquake. Earthquake-induced landslides. Springer, Berlin.341–345.
- Norikazu M., Hiroaki S., 1999. Rockfall activity from an alpine cliff during thawing periods. Geomorphology. 3(28), 309-328.
- Nadim F., Kjekstad O., Peduzzi P., Herold C., Jaedicke C.,2006. Global landslide and avalanche hotspots. Landslides. 3,159–173.
- Patton, A. I., Rathburn, S. L., & Capps, D. M., 2019. Landslide response to climate change in permafrost regions. Geomorphology. 340, 116–128.
- Patton, A. I., Rathburn, S. L., Capps, D. M., McGrath, D., & Brown, R. A.,2021. Ongoing landslide deformation in thawing permafrost. Geophysical Research Letters. 48, e2021GL092959.
- Peng J., Wang S., Wang Q., Zhuang J., Huang W., Zhu X., Leng Y., Ma P.,2018. Distribution and genetic types of loess landslides in China. J ASIAN EARTH SCI.170,329-350.
- Saez J. L., Corona C., Stoffel M., Berger F.,2013. Climate change increases frequency of shallow spring landslides in the French Alps. Geology.41(5), 619-622.
- Schulz W.H., McKenna J..P, Kibler J.D., Biavati G.,2009. Relations between hydrology and velocity of a continuously moving landslide—evidence of pore-pressure feedback regulating landslide motion? Landslides.6,181–190.
- Shan W., Guo Y., Wang F., Marui H., Strom A.,2014.Landslides in cold regions in the context of climate change. Environmental science and engineering, Springer.
- Stark C.P., Hovius N., 2001. The characterization of landslide size distributions. Geophysical Research Letters.28(6): 1091-1094.
- Subramanian S. S., Fan X., Yunus A. P., van Asch T., Scaringi G., Xu Q., Dai L., Ishikawa T., Huang R., 2020.A sequentially coupled catchment-scale numerical model for snowmelt-induced soil slope instabilities. J. Geophys. Res. Earth Surf. 125,e2019JF005468.
- Subramanian S. S., Ishikawa T., Tokoro T., 2017. Stability assessment approach for soil slopes in seasonal cold regions. Eng. Geol. 221, 154-169.
- Subramanian S. S., Ishikawa T., Tokoro T.,2018. An early warning criterion for the prediction of snowmelt induced soil slope failures in seasonally cold regions. Soils Found. 58(3), 582-601.
- Suradi, M., Fourie, A.B. & Saynor, M.J. 2016.An experimental and numerical study of a landslide triggered by an extreme rainfall event in northern Australia. Landslides. 13, 1125–1138.
- Trandafir A.C., Ertugrul O.L., Giraud R.E., McDonald G., 2015.Geomechanics of a snowmelt-induced slope failure in glacial till. Environ. Earth Sci. 73(7),3709-3716.
- Trandafir, A.C., Ertugrul, O.L., Giraud, R.E. et al., 2015.Geomechanics of a snowmelt-induced slope failure in glacial till. Environ Earth Sci. 73, 3709–3716.
- Wieczorek G.F., Lips E.W., Ellen S.,1989. Debris flows and hyperconcentrated floods along the Wasatch Front, Utah, 1983 and 1984. Bull Assoc Eng Geol. 26(2),191–208.
- Xian Y., Wei X.L.,Zhou H.B.,et al.,2022.Snowmelt‑triggered reactivation of a loess landslide in Yili, Xinjiang, China: mode and mechanism. Landslides.19,1843-1860.
- 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 C., Xu X., Yao X., Dai F.,2014. Three (nearly) complete inventories of landslides triggered by the May 12, 2008 Wenchuan Mw 7.9 earthquake of China and their spatial distribution statistical analysis. Landslides. 11(3),441–461.
- Yao, T., Xue, Y., Chen, D., et al.,2019. Recent third pole’s rapid warming accompanies cryospheric melt and water cycle intensification and interactions between monsoon and environment: Multidisciplinary approach with observations, modeling, and analysis. Bull. Am. Meteorol. Soc. 100 (3),423–444.
- Yao J.Q, Chen Y.N., Guan X.F., Zhao Y., Chen J., Mao W.Y.,2022. Recent climate and hydrological changes in a mountain – basin system in Xinjiang, China. Earth-Science Reviews. 226,103957.
- 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.
- Zhuang J., Peng J., Wang G., Javed I., Wang Y., Li W.,2017. Distribution and characteristics of landslide in Loess Plateau: A case study in Shaanxi province. Eng Geol.236,89-96.
No competing interests reported.