From spatio-temporal landslide susceptibility to landslide risk forecast

The literature on landslide susceptibility is rich with examples that span a wide range of topics.However,the component that pertains to the extension of the susceptibility framework toward space–time modeling is largely unexplored.This statement holds true,particularly in the context of landslide risk,where few scientific contributions investigate risk dynamics in space and time.This manuscript proposes a modeling protocol where a dynamic landslide susceptibility is obtained via a binomial Generalized Additive Model whose inventories span nine years(from 2013 to 2021).For the analyses,the data cube is organized with a mapping unit consisting of 26,333 slope units repeated over an annual temporal unit,resulting in a total of 236,997 units.This phase already includes several interesting modeling experiments that have rarely appeared in the landslide literature(e.g.,variable interaction plots).However,the main innovative effort is in the subsequent phase of the protocol we propose,as we used climate projections of the main trigger(rainfall)to obtain future estimates of yearly susceptibility patterns.These estimates are then combined with projections of urban settlements and associated populations to create a dynamic risk model,assuming vulnerability=1.Overall,this manuscript presents a unique example of such a modeling routine and offers a potential standard for administrations to make informed decisions regarding future urban development.

Effective and Sustainable Flood and Landslide Risk Reduction Measures:An Investigation of Two Assessment Frameworks

Natural events such as floods and landslides can have severe consequences.The risks are expected to increase,both as a consequence of climate change and due to increased vulnerabilities,especially in urban areas.Although preventive measures are often cost-effective,some measures are beneficial to certain values,while some may have negative impacts on other values.The aim of the study presented here was to investigate two frameworks used for assessing the effectiveness and sustainability of physical and nonphysical flood and landslide risk reduction measures.The study is based on literature,available information from authorities and municipalities,expert knowledge and experience,and stakeholder views and values.The results indicate that the risks for suboptimization or maladaptation are reduced if many aspects are included and a broad spectrum of stakeholders are involved.The sustamability assessment tools applied here can contribute to a more transparent and sustainable risk management process by assessing strategies and interventions with respect to both short- and long-term perspectives,including local impacts and wider environmental impacts caused by climate change,for example.The tools can also cover social and economic aspects.The assessment tools provide checklists that can support decision processes,thus allowing for more transparent decisions.

Reservoir-induced landslides and risk control in Three Gorges Project on Yangtze River,China

The Three Gorges region in China was basically a geohazard-prone area prior to construction of the Three Gorges Reservoir （TGR）. After construction of the TGR, the water level was raised from 70 m to 175 m above sea level （ASL）, and annual reservoir regulation has caused a 30-m water level difference after impoundment of the TGR since September 2008. This paper first presents the spatiotemporal distribution of landslides in six periods of 175 m ASL trial impoundments from 2008 to 2014. The results show that the number of landslides sharply decreased from 273 at the initial stage to less than ten at the second stage of impoundment. Based on this, the reservoir-induced landslides in the TGR region can be roughly classified into five failure patterns, i.e. accumulation landslide, dip-slope landslide, reversed bedding landslide, rockfall, and karst breccia landslide. The accumulation landslides and dip-slope landslides account for more than 90%. Taking the Shuping accumulation landslide （a sliding mass volume of 20.7 × 106 m^3） in Zigui County and the Outang dip-slope landslide （a sliding mass volume of about 90 × 106 m^3） in Fengjie County as two typical cases, the mechanisms of reactivation of the two landslides are analyzed. The monitoring data and factor of safety （FOS） calculation show that the accumulation landslide is dominated by water level variation in the reservoir as most part of the mass body is under 175 m ASL, and the dip-slope landslide is controlled by the coupling effect of reservoir water level variation and precipitation as an extensive recharge area of rainfall from the rear and the front mass is below 175 m ASL. The characteristics of landslide-induced impulsive wave hazards after and before reservoir impoundment are studied, and the probability of occurrence of a landslide-induced impulsive wave hazard has increased in the reservoir region. Simulation results of the Ganjingzi landslide in Wushan County indicate the strong relationship between landslide-induced surge and water variation with high potential risk to shipping and residential areas. Regarding reservoir regulation in TGR when using a single index, i.e. 1-d water level variation, water resources are not well utilized, and there is also potential risk of disasters since 2008. In addition, various indices such as 1-d, 5-d, and 10-d water level variations are proposed for reservoir regulation. Finally, taking reservoir-induced landslides in June 2015 for example, the feasibility of the optimizing indices of water level variations is verified.

Landslide identification using machine learning

Landslide identification is critical for risk assessment and mitigation.This paper proposes a novel machinelearning and deep-learning method to identify natural-terrain landslides using integrated geodatabases.First,landslide-related data are compiled,including topographic data,geological data and rainfall-related data.Then,three integrated geodatabases are established;namely,Recent Landslide Database(Rec LD),Relict Landslide Database(Rel LD)and Joint Landslide Database(JLD).After that,five machine learning and deep learning algorithms,including logistic regression(LR),support vector machine(SVM),random forest(RF),boosting methods and convolutional neural network(CNN),are utilized and evaluated on each database.A case study in Lantau,Hong Kong,is conducted to demonstrate the application of the proposed method.From the results of the case study,CNN achieves an identification accuracy of 92.5%on Rec LD,and outperforms other algorithms due to its strengths in feature extraction and multi dimensional data processing.Boosting methods come second in terms of accuracy,followed by RF,LR and SVM.By using machine learning and deep learning techniques,the proposed landslide identification method shows outstanding robustness and great potential in tackling the landslide identification problem.

Landslide disaster prevention and mitigation through works in Hong Kong

Hong Kong has a high concentration of developments on hilly terrain in close proximity to man-made slopes and natural hillsides.Because of the high seasonal rainfall,these man-made slopes and natural hillsides would pose a risk to the public as manifested by a death toll of 470 people due to landslides since the late 1940s.In 1977,the Government of the Hong Kong SAR embarked on a systematic programme,known as the Landslip Preventive Measure(LPM)Programme,to retroft substandard man-made slopes.From 1977 to 2010,about 4500 substandard government man-made slopes have been upgraded through engineering works.During the period,the Programme had evolved progressively in response to Government’s internal demand for continuous improvement and rising public expectation for slope safety.In 2010,the Government implemented the Landslip Prevention and Mitigation(LPMit)Programme to dovetail with the LPM Programme,with the focus on retroftting the remaining moderate-risk substandard man-made slopes and mitigating systematically the natural terrain landslide risk pursuant to the"react-to-known"hazard principle.This paper presents the evolution of the LPM and LPMit Programmes as well as the insight on landslide prevention and mitigation through engineering works.

Calculation of landslide occurrence probability in Taiwan region under different ground motion conditions

In this study,Bayesian probability method and machine learning model are used to study the real occurrence probability of earthquake-induced landslide risk in Taiwan region.The analyses were based on the 1999 Taiwan Chi-Chi Earthquake,the largest earthquake in the history in this Region in a hundred years,thus can provide better control on the prediction accuracy of the model.This seismic event has detailed and complete seismic landslide inventories identified by polygons,including 9272 seismic landslide records.Taking into account the real earthquake landslide occurrence area,the difference in landslide area and the non-sliding/sliding sample ratios and other factors,a total of 13,656,000 model training samples were selected.We also considered other seismic landslide influencing factors,including elevation,slope,aspect,topographic wetness index,lithology,distance to fault,peak ground acceleration and rainfall.Bayesian probability method and machine learning model were combined to establish the multi-factor influence of earthquake landslide occurrence model.The model is then applied to the whole Taiwan region using different ground motion peak accelerations(from 0.1 g to 1.0 g with 0.1 g intervals)as a triggering factor to complete the real probability of earthquake landslide map in Taiwan under different peak ground accelerations,and the functional relationship between different Peak Ground Acceleration and their predicted area is obtained.

Population amount risk assessment of extreme precipitation-induced landslides based on integrated machine learning model and scenario simulation

In this study,the future landslide population amount risk(LPAR)is assessed based on integrated machine learning models(MLMs)and scenario simulation techniques in Shuicheng County,China.Firstly,multiple MLMs were selected and hyperparameters were optimized,and the generated 11 models were crossintegrated to select the best model to calculate landslide susceptibility;by calculating precipitation for different extreme precipitation recurrence periods and combining the susceptibility results to assess the landslide hazard.Using the town as the basic unit,the exposure and vulnerability of the future landslide population under different Shared Socioeconomic Pathways(SSPs)scenarios in each town were assessed,and then combined with the hazard to estimate the LPAR in 2050.The results showed that the integrated model with the optimized random forest model as the combination strategy had the best comprehensive performance in susceptibility assessment.The distribution of hazard classes is similar to susceptibility,and with an increase in precipitation,the low-hazard area and high-hazard decrease and shift to medium-hazard and very high-hazard classes.The high-risk areas for future landslide populations in Shuicheng County are mainly concentrated in the three southwestern towns with high vulnerability,whereas the northern towns of Baohua and Qinglin are at the lowest risk class.The LPAR increased with the intensity of extreme precipitation.The LPAR differs significantly among the SSPs scenarios,with the lowest in the“fossil-fueled development(SSP5)”scenario and the highest in the“regional rivalry(SSP3)”scenario.In summary,the landslide susceptibility model based on integrated machine learning proposed in this study has a high predictive capability.The results of future LPAR assessment can provide theoretical guidance for relevant departments to cope with future socioeconomic development challenges and make corresponding disaster prevention and mitigation plans to prevent landslide risks from a developmental perspective.

The relationship between three-dimensional coseismic displacement and distribution of coseismic landslides

In mountainous areas,landslides induced by destructive earthquakes are one of the main causes of human casualties,which is an important link in the chain of earthquake hazards.Earthquake-triggered landslides are mainly controlled by three factors,namely seismic property,topography,and geology.Many studies have been conducted on these controlling factors of earthquake-triggered landslides.However,little is known about the effect of coseismic displacement on the distribution of landslides under different slope aspects and slope angles,hindering our understanding of the mechanism of inducing landslides by the combination of surface displacement and slope geometry at the local scale and leading to controversial opinions about the abnormal number of earthquake-triggered landslides in several cases.Here,we took the 2008 Wenchuan M_(w) 7.9 earthquake in China,the 2015 Gorkha M_(w) 7.8 earthquake in Nepal,and the 2016 Kaikōura M_(w) 7.8 earthquake in New Zealand as examples to investigate the relationship between the distribution of large earthquake-triggered landslides and the three-dimensional (3D)coseismic displacement field.We divided the landslide-prone area around the epicenter into regular grids and calculated the 3D coseismic displacement in each grid according to the radar satellite images and slip distribution model.Then,the 3D coseismic displacement was projected to two coordinate systems related to the slope where the landslides were located for statistical analysis.We determined that the surface uplift perpendicular to the slope is more likely to induce landslides,particularly when combined with large slope angles.Meanwhile,the number of landslides will be significantly reduced where the subsidence occurs.Regardless of uplift or subsidence,landslides are more likely to occur when the direction of coseismic horizontal displacement is far from the slope.The larger the slope angles are,the greater the effects of horizontal displacement and slope aspect.A dominant slope aspect also exists for earthquake-triggered landslides,which is different from the mean slope aspect calculated from the background topography.This dominant aspect angle is related to the focal mechanism and striking angle of surface rupture.These results indicate that we can simulate the 3D coseismic displacement field from known fault location and earthquake mechanism and combine the topographic data for landslide risk assessment in earthquake-prone mountainous areas to minimize the damage caused by possible earthquake-triggered landslides.