Climate warming has aggravated the occurrence of thaw settlement in permafrost region, but the associated risk has not been precisely assessed or understood. This study applied four machine learning models to explore and compare the spatial distribution of thaw settlement risk in the Wudaoliang-Tuotuohe region, Qinghai-Tibet Plateau, namely, naive Bayesian, k-nearest neighbor, logistic model tree and random forest models. A total of 853 thaw settlement locations and 12 conditioning factors were used to train and validate the above four models. The results indicated that random forest model performed best with the highest accuracy. The risk map produced by random forest model implied that about 76.55% of thaw settlements were located in very high-risk regions, which only occupied 6.85% of study area. The volume ice content, active layer thickness and thawing degree days were the main factors leading thaw settlement. By further comparing the performances between random forest model and other three models, the overestimated and underestimated risk regions (Beiluhe and Tuotuohe basins), and imbalanced conditioning factors (altitude and slope angle) were determined. In contrast with similar studies, this research performed better in model construction and accuracy. The results can help designers to implement precautionary measures in thaw settlement risk management.

Climate warming could exacerbate the occurrence of thaw settlement hazard in the permafrost regions of the Qinghai-Tibet Plateau (QTP), which would threaten the stability of engineering infrastructure in cold regions. The risk associated with permafrost settlement, valuable for the regional sustainable development, remains poorly assessed or understood on the QTP. In this study, three common Geo-hazard indices were used to assess the settlement risks in the permafrost regions of the QTP, including the settlement index, the risk zonation index, and the allowable bearing capacity index. However, large spatial differences existed in simulating the risk maps by using the abovementioned Geo-hazard indices. Hence, we developed a combined index (I-c) by integrating the three indices to reduce the uncertainty of the simulations. The results indicated that the ground ice is a critical factor for assessing the settlement risk in permafrost regions. We also applied the Ic to assess the settlement risk along the Qinghai-Tibet Railway (QTR). The proportion of low-risk area along the QTR would be the highest (45.38%) for the future periods 2061-2080 under Representative Concentration Pathway 4.5. The medium-risk area combined with the high-risk area would be accounted for more than 40%, which were located at the boundary of the present permafrost regions. Therefore, the corresponding adaptation measures should be taken to reduce the potential economic losses caused by the high-risk regions to the infrastructure. Overall, the results would present valuable references for engineering design, construction and maintenance, and provide insights for early warning and prevention of permafrost thaw settlement hazard on the QTP. (C) 2021 Elsevier B.V. All rights reserved.

Climate warming could exacerbate the occurrence of thaw settlement hazard in the permafrost regions of the Qinghai-Tibet Plateau (QTP), which would threaten the stability of engineering infrastructure in cold regions. The risk associated with permafrost settlement, valuable for the regional sustainable development, remains poorly assessed or understood on the QTP. In this study, three common Geo-hazard indices were used to assess the settlement risks in the permafrost regions of the QTP, including the settlement index, the risk zonation index, and the allowable bearing capacity index. However, large spatial differences existed in simulating the risk maps by using the abovementioned Geo-hazard indices. Hence, we developed a combined index (I-c) by integrating the three indices to reduce the uncertainty of the simulations. The results indicated that the ground ice is a critical factor for assessing the settlement risk in permafrost regions. We also applied the Ic to assess the settlement risk along the Qinghai-Tibet Railway (QTR). The proportion of low-risk area along the QTR would be the highest (45.38%) for the future periods 2061-2080 under Representative Concentration Pathway 4.5. The medium-risk area combined with the high-risk area would be accounted for more than 40%, which were located at the boundary of the present permafrost regions. Therefore, the corresponding adaptation measures should be taken to reduce the potential economic losses caused by the high-risk regions to the infrastructure. Overall, the results would present valuable references for engineering design, construction and maintenance, and provide insights for early warning and prevention of permafrost thaw settlement hazard on the QTP. (C) 2021 Elsevier B.V. All rights reserved.

In this article, we consider the problem of thermal response of the near-surface ice-rich permafrost to the effects of linear infrastructure and current climate change. First, we emphasize the scientific and practical significance of the study and briefly describe permafrost conditions and related hazards in the study area. Then we present a mathematical model which accounts for the actual process of soil thawing and freezing and consists of two nonlinear equations: heat conduction and moisture transfer. Numerical calculations were made to predict temperature and moisture conditions in the railroad embankment, taking into account solar radiation, snow cover, rainfall infiltration, and evaporation from the surface. The numerical results indicate that moisture migration and infiltration play the primary role in the development of frost heaving and thaw settlement. During winter, the frost-heave extent is monotonously increased due to pore moisture migration to the freezing front. Strong volume expansion (dilatation) is observed near the surface of the active layer with the onset of the warm season and meltwater infiltration. Settlement of the upper layers of the soil occurs in the summer months (June-August) when there is intense evaporation due to drying. Autumn rains stop the process of thaw settlement by increasing the soil moisture. The above processes are repeated cyclically every year. A frozen core shifts to the shaded side of the embankment under the influence of variations in the solar radiation. Over time, the total moisture content of the frozen core is increased which increases differential heaving and negatively affects the stress-strain state in the embankment. The quantitative and qualitative characteristics of the processes of frost heaving and thaw settlement are obtained in the annual and long-term cycles.

The permafrost regions currently occupy about one quarter of the Earth's land area. Climate-change scenarios indicate that global warming will be amplified in the polar regions, and could lead to a large reduction in the geographic extent of permafrost. Development of natural resources, transportation networks, and human infrastructure in the high northern latitudes has been extensive during the second half of the twentieth century. In areas underlain by ice-rich permafrost, infrastructure could be damaged severely by thaw-induced settlement of the ground surface accompanying climate change. Permafrost near the current southern margin of its extent is degrading, and this process may involve a northward shift in the southern boundary of permafrost by hundreds of kilometers throughout much of northern North America and Eurasia. A long-term increase in summer temperatures in the high northern latitudes could also result in significant increases in the thickness of the seasonally thawed layer above permafrost, with negative impacts on human infrastructure located on ice-rich terrain. Experiments involving general circulation model scenarios of global climate change, a mathematical solution for the thickness of the active layer, and digital representations of permafrost distribution and ice content indicates potential for severe disruption of human infrastructure in the permafrost regions in response to anthropogenic climate change. A series of hazard zonation maps depicts generalized patterns of susceptibility to thaw subsidence. Areas of greatest hazard potential include coastlines on the Arctic Ocean and parts of Alaska, Canada, and Siberia in which substantial development has occurred in recent decades.