Modelling of shallow landslides with machine learning algorithms

This paper introduces three machine learning(ML)algorithms,the‘ensemble'Random Forest(RF),the‘ensemble'Gradient Boosted Regression Tree(GBRT)and the Multi Layer Perceptron neural network(MLP)and applies them to the spatial modelling of shallow landslides near Kvam in Norway.In the development of the ML models,a total of 11 significant landslide controlling factors were selected.The controlling factors relate to the geomorphology,geology,geo-environment and anthropogenic effects:slope angle,aspect,plan curvature,profile curvature,flow accumulation,flow direction,distance to rivers,water content,saturation,rainfall and distance to roads.It is observed that slope angle was the most significant controlling factor in the ML analyses.The performance of the three ML models was evaluated quantitatively based on the Receiver Operating Characteristic(ROC)analysis.The results show that the‘ensemble'GBRT machine learning model yielded the most promising results for the spatial prediction of shallow landslides,with a 95%probability of landslide detection and 87%prediction efficiency.

Estimation of rainfall thresholds for shallow landslides in the Sierra Madre Oriental, northeastern Mexico

Landslides induced by prolonged rainfalls are frequent mass movements along the northeastern portion of the Sierra Madre Oriental in Mexico,causing significant damage to infrastructure.In this work,we have studied the connection between rainfall and landslides in the Santa Rosa Canyon,a catchment located in the northeastern Mexico.A landslide database triggered by major storms and hurricanes that have hit the region over the past 30 years was analyzed.A total of 92 rainfall events in the Santa Rosa Canyon were studied to determine the amount of precipitation needed to trigger shallow landslides.For each event the duration(D,in hours)and the cumulated rainfall event(E,in mm)were determined by using historical rainfall data from weather stations located near the study area.We have proposed an ED threshold for rainfall-induced landslides with durations 0.5<D<120 hours to address the conditions that trigger these events in the study area.On analyzing the obtained threshold,it has been established that almost 60 mm of a daily rainfall accumulation is required to trigger shallow landslides in the study area.This estimation is consistent with other calculations made for areas close to the Santa Rosa Canyon.Finally,we validated the predictive capability of the threshold with a different set of rainfall data that did not result in landslides performing a straightforward receiver operating characteristic analysis.A good approach was obtained,especially for rainfall events with daily measurements.Results could be used as input information in the design of a landslide early warning system for the northeastern Mexico,and replicated for other landslide prone areas in the region.

Estimation of Shallow Landslide Susceptibility Incorporating the Impacts of Vegetation on Slope Stability

This study aimed to develop a physical-based approach for predicting the spatial likelihood of shallow landslides at the regional scale in a transition zone with extreme topography.Shallow landslide susceptibility study in an area with diverse vegetation types as well as distinctive geographic factors(such as steep terrain,fractured rocks,and joints)that dominate the occurrence of shallow landslides is challenging.This article presents a novel methodology for comprehensively assessing shallow landslide susceptibility,taking into account both the positive and negative impacts of plants.This includes considering the positive efects of vegetation canopy interception and plant root reinforcement,as well as the negative efects of plant gravity loading and preferential fow of root systems.This approach was applied to simulate the regional-scale shallow landslide susceptibility in the Dadu River Basin,a transition zone with rapidly changing terrain,uplifting from the Sichuan Plain to the Qinghai–Tibet Plateau.The research fndings suggest that:(1)The proposed methodology is efective and capable of assessing shallow landslide susceptibility in the study area;(2)the proposed model performs better than the traditional pseudo-static analysis method(TPSA)model,with 9.93%higher accuracy and 5.59%higher area under the curve;and(3)when the ratio of vegetation weight loads to unstable soil mass weight is high,an increase in vegetation biomass tends to be advantageous for slope stability.The study also mapped the spatial distribution of shallow landslide susceptibility in the study area,which can be used in disaster prevention,mitigation,and risk management.

Automated machine learning for rainfall-induced landslide hazard mapping in Luhe County of Guangdong Province,China

Landslide hazard mapping is essential for regional landslide hazard management.The main objective of this study is to construct a rainfall-induced landslide hazard map of Luhe County,China based on an automated machine learning framework(AutoGluon).A total of 2241 landslides were identified from satellite images before and after the rainfall event,and 10 impact factors including elevation,slope,aspect,normalized difference vegetation index(NDVI),topographic wetness index(TWI),lithology,land cover,distance to roads,distance to rivers,and rainfall were selected as indicators.The WeightedEnsemble model,which is an ensemble of 13 basic machine learning models weighted together,was used to output the landslide hazard assessment results.The results indicate that landslides mainly occurred in the central part of the study area,especially in Hetian and Shanghu.Totally 102.44 s were spent to train all the models,and the ensemble model WeightedEnsemble has an Area Under the Curve(AUC)value of92.36%in the test set.In addition,14.95%of the study area was determined to be at very high hazard,with a landslide density of 12.02 per square kilometer.This study serves as a significant reference for the prevention and mitigation of geological hazards and land use planning in Luhe County.

Spatial prediction of rainfall-induced shallow landslides using hybrid integration approach of Least-Squares Support Vector Machines and differential evolution optimization:a case study in Central Vietnam

This study represents a hybrid intelligence approach based on the differential evolution optimization and Least-Squares Support Vector Machines for shallow landslide prediction,named as DE-LSSVMSLP.The LSSVM is used to establish a landslide prediction model whereas the DE is adopted to search the optimal tuning parameters of the LSSVM model.In this research,a GIS database with 129 historical landslide records in the Quy Hop area(Central Vietnam)has been collected to establish the hybrid model.The receiver operating characteristic(ROC)curve and area under the curve(AUC)were used to assess the performance of the newly constructed model.Experimental results show that the proposed model has high performances with approximately 82%of AUCs on both training and validating datasets.The model’s results were compared with those obtained from other methods,Support Vector Machines,Multilayer Perceptron Neural Networks,and J48 Decision Trees.The result comparison demonstrates that the DE-LSSVMSLP deems best suited for the dataset at hand;therefore,the proposed model can be a promising tool for spatial prediction of rainfall-induced shallow landslides for the study area.

Characteristics of root-permeated soil under simple-shear and direct-shear conditions

The simple-shear condition is closer to reality than the direct-shear condition for simulating the mechanical behavior of vegetated soil slope under shallow failure.However,study on simple-shear characteristics for vegetated slope is still insufficient,and there lacks intuitive comparison of characteristics between these two shear conditions.In this study,large-scale simple-shear and direct-shear experiments were conducted on soil permeated by roots of Amorpha fruticosa to investigate the shear strength and stiffness.The stress-displacement relationship of each sample was obtained and further normalized to unify the influence of root content.The results reveal that the direct-shear condition overestimates the shear strength of root-permeated soils(by 41%)and thus the estimation of slope stability based on the parameters of direct-shear condition is not conservative.Furthermore,the initial stiffness of root-permeated soil under simple-shear condition is 34%lower than that under direct-shear condition.The higher strength and stiffness under direct-shear condition are caused by the following reasons:the shear plane does not have the lowest strength,the shear area is decreasing,and the shear zone is thinner.The significant deformation(lower stiffness)revealed by the simple-shear condition facilitates the application of early warning for vegetated shallow landslides.

Effects of solar radiation and fine roots on suction of Amorpha fruticose-vegetated soil

The thickness of shallow landslides is generally less than 2 m,which is of the same order of magnitude as the growth range of vegetation roots.Vegetation roots can improve the stability of shallow soil through mechanical and hydraulic effects.Therefore,the landslide process is closely related to the plant roots growing on the slope surface.Plant roots play a dominant role in the regulation of soil suction through solar radiation induced transpiration.However,little is known about the correlation between cumulative solar radiation and soil suction.Moreover,the specific effects of fine roots on the suction distribution are not clear in most previous studies.In this study,a vegetated soil of a drought-tolerant and water-tolerant shrub,namely Amorpha fruticose,was adopted.The suction and volumetric water content of bare and vegetated soils were monitored under natural conditions for 4 months.The results demonstrate that there is a nearly linear relationship between cumulative solar radiation and suction ranging from zero to 100 kPa.Regarding the modeling of the soil-plant-atmosphere interactions,this relationship could serve a significant role in calculating the root water uptake under given solar radiation conditions.In addition,higher suctions were observed at the lower layer of the vegetated soil than those at the middle layer,which is different from the results of vegetated soil from previous investigations.This is due to the fact that the root area index(RAI)of fine roots at the lower layer is twice that of the middle layer.Importantly,the higher concentration of fine roots at the lower layer of vegetated soil sample resembles the root distribution of shrub near the soil-bedrock interface on shallow bedrock landslides.The fine roots would increase soil suction through transpiration,and hence reduce the permeability and increase shear strength of landslides.Eventually,these new findings serve as a preliminary step on the evaluation of the stability of this common type of landslides.

Untangling the influence of soil moisture on root pullout property of alfafa plant

Root pullout property of plants was of key importance to the soil reinforcement and the improvement of slope stability. To investigate the influence of soil moisture on root pullout resistance and failure modes in soil reinforcement process, we conducted pullout tests on alfalfa(Medicago sativa L.) roots at five levels(40, 30, 20, 10 and 6 kPa) of soil matric suction, corresponding to respectively 7.84%, 9.66%, 13.02%, 19.35% and 27.06% gravimetric soil moisture contents. Results showed that the maximal root pullout force of M. sativa decreased in a power function with increasing soil moisture content from 7.84% to 27.06%. Root slippage rate increased and breakage rate decreased with increasing soil moisture content. At 9.66% soil moisture content, root slippage rate and breakage rate was 56.41% and 43.58%, respectively. The threshold value of soil moisture content was about 9.00% for alfalfa roots in the loess soil. The maximal pullout force of M. sativa increased with root diameter in a power function. The threshold value of root diameter was 1.15 mm, because root slipping force was greater than root breaking force when diameter >1.15 mm, while diameter ≤1.15 mm, root slipping force tended to be less than root breaking force. No significant difference in pullout forces was observed between slipping roots and breaking roots when they had similar diameters. More easily obtained root tensile force(strength) is suggested to be used in root reinforcement models under the condition that the effect of root diameter is excluded as the pullout force of breaking roots measured in pullout tests is similar to the root tensile force obtained by tensile tests.

Physically-based approach to analyze rainfall-triggered landslide using hydraulic gradient as slide direction

An infinite slope stability numerical model driven by the comprehensive physically-based integrated hydrology model(InHM) is presented.In this approach,the failure plane is assumed to be parallel to the hydraulic gradient instead of the slope surface.The method helps with irregularities in complex terrain since depressions and flat areas are allowed in the model.The present model has been tested for two synthetic single slopes and a small catchment in the Mettman Ridge study area in Oregon,United States,to estimate the shallow landslide susceptibility.The results show that the present approach can reduce the simulation error of hydrological factors caused by the rolling topography and depressions,and is capable of estimating spatial-temporal variations for landslide susceptibilities at simple slopes as well as at catchment scale,providing a valuable tool for the prediction of shallow landslides.