Abstracts:Freeze-thaw (F-T) cycle is an important external factor affecting the mechanical properties of saline soils in cold regions. In this study, the effects of F-T cycles (0, 1, 5, 10, 30, 60 and 120) and salt contents (0, 0.5, 1.0, 2.0 and 3.0%) on the shear properties of saline soil in western Jilin Province of northeastern China, including stress-strain behavior, peak undrained shear strength (τ u ), resilient modulus (E R ), cohesion (c u ) and internal friction angle (φ u ), were investigated by conducting unconsolidated-undrained (UU) triaxial compression tests. The results demonstrated that, when the salt content was the same, the τ u , E R and c u basically showed a decreasing-steady-decreasing trend with increasing F-T cycles, and the dynamic evolutions of soil microstructure were mainly responsible for such variations. The Gouy-Chapman diffuse double layer together with matrix suction theories were employed to analyze the mechanism of the effects of salt content on the c u and φ u at different freeze-thaw cycles; the thicker diffuse double layer induced by higher sodium ion contents and the decreasing matrix suction were deemed as the main reasons that led the c u to continuous reduction; the φ u was believed to be successively affected by the enhanced lubrication effect and the salt crystallization process, and there was a threshold salt content (2.0%) which was influenced by the number of F-T cycles. Finally, by categorizing the different experimental combinations into relative undamaged, single-factor damaged and two-factor damaged states concerning the τ u , an empirical mathematical equation of high reliability (R ²  > 0.985) was established to describe the combining effects of F-T and salinity on the changes in τ u .

Abstracts:Time-lapse photography was employed to monitor a snowbank at 3640 m above sea level in the Uinta Mountains, Utah, USA. The snowbank forms against a 35-m high, east-facing escarpment and is nourished by wind redistribution of snow from an alpine plateau. The snowbank is capped by a large cornice, and its central core persists through the summer in most years. An automated, solar-powered digital camera was deployed between October 2016 and June 2017, and programmed to capture 5 photographs each day between 8:00 and 16:00, local time. Although cold temperatures affected the batteries during the winter, a total of 812 photographs were collected, for an overall average of 3.4 per day. These images were combined in an animation displaying the growth of the snowbank and associated cornice over the course of the winter. The debris fields resulting from nineteen cornice fall avalanches were noted in the sequence of photographs. Data collected at a nearby automated weather station reveal that 12 avalanches (between early December and early April) were preceded by significant increases in snow depth, snow water equivalent, and precipitation, with sustained windspeeds above the winter average. In contrast, six of the last seven events (between late April and early June) occurred in the absence of new snowfall, but were associated with rapid rises in temperature and notable decreases in snow water equivalent. The average interval between recorded avalanches is 10 days, with a maximum of 29 days and a minimum of 2 h. Recurrence intervals were shorter in December/January, in late March/early April, and in late April/early May. Only one avalanche occurred in the 56 days between 27 January and 24 March. Time-lapse photography is a powerful tool for monitoring nival processes.

Abstracts:In a permafrost environment, supra-permafrost water is an important local factor affecting the shallow ground thermal regime and also plays a significant role in geotechnical problems. In this study, the impacts of supra-permafrost water ponding and drainage on the thermal regime and settlement characteristics of a railway embankment in continuous permafrost zone, the interior of the Qinghai-Tibet Plateau were investigated using 13-year records (2003–2016) of field measured data. Results showed that, because of the great latent heat of the ponded water alongside the embankment, a zero curtain layer (ZCL) as much as 5.5 m thick developed in and beneath the embankment after embankment construction in 2003. A three-layered thermal regime developed in the embankment and the subgrade. Although crushed rock revetments were installed 5 years after embankment construction, no soil cooling occurred beneath the ZCL and the ZCL remained stable in thickness. Meanwhile, embankment settlement developed quickly and nearly linearly from 2003 to 2012, with a cumulative settlement approaching approximately 160 mm at both the sunny and the shady embankment shoulders. Compression and creep of warm and ice-rich permafrost layer (WIPL) beneath the ZCL are considered as the main contributors to this significant settlement. Following drainage of the ponded water during 2012, the ZCL cooled quickly and completely frozen after two cold seasons. The underlying permafrost table moved upwards into the embankment and the WIPL cooled with time. Embankment settlement also slowed significantly. However, an asymmetrical temperature field developed in the embankment and the subgrade because of the sunny-shady embankment slope effect, leading to differential settlement between the sunny and the shady embankment shoulders.

Abstracts:Expected runout distances and related return periods are the most important parameters needed for zoning in terrain prone to snow avalanching. Hazard mapping procedures usually allocate areas of land to zones with a different degree of danger based on return periods estimated for given snow volumes in the starting zone or with statistical/dynamical models. On forested avalanche paths, dendrogeomorphology has a great potential to add critical input data to these calculations in terms of recurrence intervals or return periods. However, quite paradoxically, recurrence interval maps of snow avalanches have only rarely been retrieved from tree-ring analysis and mostly represent the inverse of the mean frequency of avalanches that could be retrieved locally rather than the return period. The purpose of this study therefore was to propose a consistent approach for tree-ring based recurrence interval mapping of snow avalanche events. On the basis of 71 snow avalanches retrieved from 2570 GD growth disturbances identified in 307 larch trees from three avalanche paths located in the vicinity of Täsch (Canton of Valais, Swiss Alps), we first followed the classical approach used in dendrogeomorphology and derived recurrence interval maps through interpolation from recurrence intervals observed at the level of individual trees. We then applied an expert delineation of the spatial extent of past events based on the location of disturbed trees. Our results show that the second step improved representation of expected patterns of recurrence intervals that typically increase as one moves down the centerline of the avalanche path. Despite remaining limitations and uncertainties precluding from direct use of our maps for hazard mapping purpose, these results suggest that dendrogeomorphic time series of snow avalanches can yield valuable information for the assessment of recurrence intervals of avalanches on forested paths for which only very limited or no historical data exists, and that this data can be obtained independently from meteorological data or numerical modeling.

Abstracts:Existing grounded ice or floating ice in late winter represents the distinct thermal state of the lake bottom, which regulates the evolution processes of thermokarst lakes. The aim of this study was to identify either grounded ice or floating ice using ground penetrating radar (GPR) and analyze the factors that influence identification. Densely spaced GPR measurements were collected using a pair of 900 MHz antennas in a thermokarst lake on the northeast Qinghai–Tibet Plateau. Numerical simulations were used to evaluate the factors that influence the accuracy and reliability of the GPR detection. Results show that the ice thickness was less than 1.0 m and indicated a spatial difference on the lake. The normalized amplitudes of reflected waves (NAR) were used for discriminating the grounded ice and floating ice. The floating ice was identified by high NAR (higher than 0.3), and the grounded ice was discriminated by low NAR (lower than 0.2). Numerical simulations showed that the sandwiched thin water and frozen sediment layer beneath the lake ice can smooth the amplitude contrasts between the grounded ice and floating ice. The air/water bubbles in ice and undulating ice bottom interfaces scatter the reflection amplitudes significantly. The factors that influence the GPR reflection amplitudes apart from the permittivity contrasts, can lead to notable uncertainties in identifying grounded and floating ice based on NAR.