Article
Engineering, Geological
Jie Zhang, Zipeng Wang, Jinzheng Hu, Shihao Xiao, Wenyu Shang
Summary: This paper introduces a Bayesian machine learning-based method to evaluate the effect of model and observational uncertainties on slope failure time (SFT) prediction. A comprehensive slope failure database is compiled and the application of the method is illustrated with an example. Verification studies show that the method outperforms the traditional inverse velocity method and the maximum likelihood method. This study provides an effective tool for predicting SFT.
JOURNAL OF ROCK MECHANICS AND GEOTECHNICAL ENGINEERING
(2022)
Article
Engineering, Geological
Jie Zhang, Hong-zeng Yao, Zi-peng Wang, Ya-dong Xue, Lu-lu Zhang
Summary: This paper assesses the performances of linear and non-linear inverse velocity (INV) methods in predicting slope failure time. It is found that non-linear INV methods may encounter pitfalls such as saddle points and ill-conditioned Hessian matrices. On the other hand, linear INV methods are more stable and accurate, making them a preferred choice in future applications.
GEORISK-ASSESSMENT AND MANAGEMENT OF RISK FOR ENGINEERED SYSTEMS AND GEOHAZARDS
(2023)
Article
Environmental Sciences
Jia-zhu Wang, Neng-pan Ju, Yong-bo Tie, Yong-jian Bai, Hua Ge
Summary: This study presents a simple framework for identifying the onset of acceleration (OOA) in landslides by analyzing monitoring velocity data. The framework includes three steps: selection of the absolute value of velocity, reliable area identification, and OOA identification. A new parameter based on exponential moving average (EMA) is developed to identify the landslide OOA. The framework is applied to three historical case studies, demonstrating good correspondence between the accuracy rate and the coefficient of determination (R-2). The predicted failure time based on the identified OOA points shows high R-2 and accuracy.
JOURNAL OF MOUNTAIN SCIENCE
(2023)
Article
Engineering, Multidisciplinary
Zhou XiaoPing, Ye Teng
Summary: The study proposed the INSRA method to estimate the time-of-failure of unstable slopes, improving the accuracy and effectiveness of forecasting.
SCIENCE CHINA-TECHNOLOGICAL SCIENCES
(2021)
Article
Environmental Sciences
Yan Du, Lize Ning, Santos D. Chicas, Mowen Xie
Summary: This paper proposes conditions for using the INVM based on the TSR to discriminate stepped landslides. It analyzes the real-time changes of TSR and discusses the changes of Delta TSR after the uniform deformation phase of the landslide. The results show that TSR reaches the extreme value one day earlier than the landslide deformation velocity, indicating its potential to reduce false alarms in risk assessment.
ENVIRONMENTAL EARTH SCIENCES
(2023)
Article
Engineering, Geological
Xinli Hu, Shuangshuang Wu, Guangcheng Zhang, Wenbo Zheng, Chang Liu, Chuncan He, Zhongxu Liu, Xuyuan Guo, Han Zhang
Summary: Empirical and numerical methods are commonly used for landslide movement prediction due to their ability to forecast failure time and consider influential factors. This study introduces an integrated prediction model using the Verhulst inverse function and random forest algorithm to fully consider landslide kinematics and external factors. Results demonstrate significant improvement in prediction accuracy compared to individual models, with a decrease in error rates and increased feasibility for predicting movement of other landslides.
ENGINEERING GEOLOGY
(2021)
Article
Engineering, Geological
Eduardo E. Alonso
Summary: This paper analyzes the dynamic behavior of landslides with a well-defined failure surface and discusses procedures for identifying velocity and runout once stability is lost. By contrasting case studies with theoretical developments, it is evident that phenomena such as creeping motion and rapid sliding are crucial in landslide research.
Article
Geosciences, Multidisciplinary
Han Du, Danqing Song
Summary: This study improves the accuracy of slope failure prediction by comparing the traditional IVM method and a modified IVM method. The effects of noise on the prediction results are assessed, and a two-level alert procedure combining different filters is proposed.
Article
Metallurgy & Metallurgical Engineering
Lei Cheng, Qing Guo, Wei Yu, Ying Han, Qingwu Cai
Summary: The mathematical features of the classical theta projection method and its modified forms based on creep curves of a low-cost heat-resistant steel are analyzed. The prediction accuracy is compared using a detailed data processing method and the relative error on different creep stages. Results show that the 3-parameter method improves balanced residual errors but sacrifices accuracy on primary creep, while the 5-parameter method accurately predicts primary creep characteristics with reduced deviation in the tertiary creep stage. The 5-parameter method also has a larger reliable prediction region than the 3-parameter method, showing good consistency at stress levels above 134 MPa. The study provides theoretical guidance for the selection of different theta projection methods.
STEEL RESEARCH INTERNATIONAL
(2023)
Article
Geochemistry & Geophysics
Hiroyuki Noda, Chengrui Chang
Summary: Forecasting the acceleration of slow landslides to catastrophic failure is crucial for improving landslide warnings. A previous study used a rate-and-state-dependent friction law to analyze the creep behavior of landslides and found that the acceleration follows a power law with an exponent value of 2. Further analysis showed that the slip rate affects the exponent value differently depending on the type of friction law used.
EARTH AND PLANETARY SCIENCE LETTERS
(2023)
Article
Geosciences, Multidisciplinary
Shuo Zhang, Tong Jiang, Xiangjun Pei, Runqiu Huang, Qiang Xu, Yushan Xie, Xuwei Pan, Longxiao Zhi
Summary: The forecast of failure time and the definition of early warning threshold for unstable slope are crucial for preventing landslide disaster and reducing losses. A mathematical model is used to accurately describe the creep behavior of the slope, and a forecasting method for creep landslide is proposed. An improved tangential angle criterion is examined to enhance the accuracy of the forecast. The study finds that the initial creep stage is essential for predicting the failure time.
FRONTIERS IN EARTH SCIENCE
(2022)
Article
Engineering, Environmental
Zhiyong Fu, Jingjing Long, Wenqiang Chen, Changdong Li, Haikuan Zhang, Wenmin Yao
Summary: Prediction models for landslide displacement with step-like behavior in the reservoir area were established based on machine learning algorithms, with a hybrid reliability model proposed to improve evaluation accuracy. The models decomposed displacement into trend and periodic components using CEEMD algorithm, with RF and BPNN algorithms employed for prediction. Results showed both CEEMD-RF and CEEMD-BPNN models accurately predicted displacement, with CEEMD-RF model showing more reliable predictive ability under different scenarios. The proposed failure probability could more accurately and comprehensively evaluate the model's predictive ability compared to existing indices.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2021)
Article
Engineering, Geological
Longfei Zhang, Lei Gui, Yang Wang, Jizhixian Liu, Ying Cao
Summary: This study systematically explores the instability issues of downhill progressive failure in long slopes and proposes a unified analytical model to predict triggering thresholds and failure modes. The findings of three practical cases show good consistency with field observations and existing studies. This work provides important insights for the risk assessment and decision-making of downhill progressive landslides.
CANADIAN GEOTECHNICAL JOURNAL
(2023)
Article
Environmental Sciences
Chunfeng Ye, Heping Xie, Fei Wu, Cunbao Li
Summary: This study investigated the influence of moisture content on the creep behavior of sandstone. It was found that the creep stress had a greater impact on creep deformation and rate, while moisture content played a dominant role in creep time-to-failure. The power-law creep model was found to be the most accurate in predicting sandstone creep deformation.
ENVIRONMENTAL EARTH SCIENCES
(2022)
Article
Environmental Sciences
Katsuo Sasahara
Summary: This study measured groundwater level and surface displacement in a sandy model slope under repeated rainfall to examine the effect of repeated pore pressure loading and unloading on slope deformation. The velocity had small fluctuations even immediately before failure, and positive and negative accelerations occurred due to velocity fluctuations. The velocity increased with a rise in groundwater level and approached its ultimate value before failure. Surface displacement increased with both rising and falling groundwater levels, as well as under a constant groundwater level. The relationship between velocity and acceleration derived from surface displacement was linear on a logarithmic scale and unique for each stage with increasing and decreasing velocities due to groundwater level changes. A new relationship was established in this study for any velocity trend and a method for predicting the time of failure was proposed based on the velocity-acceleration relationship.
ENVIRONMENTAL EARTH SCIENCES
(2022)
