Prediction Of Rainfall-induced Landslides Research Articles

Advancements, Challenges, and Future Directions in Rainfall-Induced Landslide Prediction: A Comprehensive Review

Advancements, Challenges, and Future Directions in Rainfall-Induced Landslide Prediction: A Comprehensive Review

Investigation of Model Uncertainty in Rainfall-Induced Landslide Prediction under Changing Climate Conditions

Climate change can exacerbate the occurrence of extreme precipitation events, thereby affecting both the frequency and intensity of rainfall-induced landslides. It is important to study the threat of rainfall-induced landslides under future climate conditions for the formulation of disaster prevention and mitigation policies. Due to the complexity of the climate system, there is great uncertainty in the climate variables simulated by a global climate model (GCM), which will be further propagated in landslide prediction. In this study, we investigate the spatial and temporal trends of future landslide hazards in China under climate change, using data from a multi-model ensemble of GCMs based on two scenarios, RCP4.5 and RCP8.5. The uncertainty characteristics are then estimated based on signal-to-noise ratios (SNRs) and the ratio of agreement in sign (RAS). The results show that the uncertainty of landslide prediction is mainly dominated by the GCM ensemble and the RCP scenario settings. Spatially, the uncertainty of landslide prediction is high in the western areas of China and low in the eastern areas of China. Temporally, the uncertainty of landslide prediction is evolving, with characteristics of high uncertainty in the near future and characteristics of low uncertainty in the distant future. The annual average SNRs in the 21st century are 0.44 and 0.50 in RCP4.5 and RCP8.5, respectively, and the RAS of landslide prediction in Southeastern China is only 50–60%. This indicates that more than half of the patterns show trends that are opposite to those of the ensemble, suggesting that their landslide change trends are not universally recognized in the pattern ensemble. Considering the uncertainty of climate change in landslide prediction can enable studies to provide a more comprehensive picture of the possible range of future landslide changes, effectively improving the reliability of landslide hazard prediction and disaster prevention.

Machine learning and artificial intelligence models development in rainfall-induced landslide prediction

<span lang="EN-US">In Indonesia, rainfall is one crucial triggering factor for landslides. This paper aims to build landslide event prediction models using several machine learning and artificial intelligence algorithms. The algorithms were trained with two different methods. The input of the algorithms was precipitation data obtained from the global satellite mapping of precipitation satellite observation, and the target was landslide event occurrence data obtained from the Indonesian National Board for Disaster Management. Each algorithm provided some model candidates with different parameter settings for each method. As a result, there were 52 and 72 model candidates for both methods. The best model was then chosen from each method. The result shows that the model generated by generalized linear model was the best model for the first method and deep learning for the second one. Furthermore, the best models at each method gained 0.828 and 0.836 for the area under receiver operating characteristics curve, and their log-loss were 0.156 and 0.154. The second method, which used input data transformation, provided better performance.</span>

Insight from a Physical-Based Model for the Triggering Mechanism of Loess Landslides Induced by the 2013 Tianshui Heavy Rainfall Event

Rainfall-induced landslides pose a significant threat to human life, destroy highways and railways, and cause farmland degradation in the Loess Plateau. From 19 June 2013 to 26 July 2013, continuous and heavy rainfall events occurred in the Tianshui area, Gansu Province. This strong rainfall process included four short-term serious rainfall events and long-term intermittent rainfall, triggering many shallow loess landslides. To improve our understanding of this rainfall process as the triggering mechanism of the loess landslides, we conducted the physical-based spatiotemporal prediction of rainfall-induced landslides. By utilizing precipitation data recorded every 12 h from the rain gauge stations and 51 soil samples from within a 50 km radius of the study area, we predicted 1000 physical-based model-calculated pictures of potential landslides, and the slope failure probability (Pf) of the study area was obtained by Monte Carlo simulations. The model was validated by the actual landslide data of the 2013 heavy rainfall event, and the effects of the precipitation process and the trigger mechanism on the landslides were discussed. The results showed that the fourth rainfall event had the best prediction ability, while the third event had the second-best prediction ability. There was a solid linear link between the antecedent precipitation (Pa) and the predicted landslide area (Pls) based on the fitting relationship, indicating that antecedent rainfall may play a significant role in the occurrence of landslides in the region. By comparing the distribution of the predicted results of the four heavy rainfall events with the actual landslide, we observed that the first two rainfall processes may not have been the main reason for slope failure, contributing only to prepare for the landslides in the later period. The superposition of the fourth and third rainfall events finally determined the spatial distribution characteristics of the landslide induced by the 2013 heavy rainfall event.

Influencing factors on the simulation of rainfall-induced landslide prediction based on case study

Influencing factors on the simulation of rainfall-induced landslide prediction based on case study

Ensemble machine learning models based on Reduced Error Pruning Tree for prediction of rainfall-induced landslides

ABSTRACT In this paper, we developed highly accurate ensemble machine learning models integrating Reduced Error Pruning Tree (REPT) as a base classifier with the Bagging (B), Decorate (D), and Random Subspace (RSS) ensemble learning techniques for spatial prediction of rainfall-induced landslides in the Uttarkashi district, located in the Himalayan range, India. To do so, a total of 103 historical landslide events were linked to twelve conditioning factors for generating training and validation datasets. Root Mean Square Error (RMSE) and Area Under the receiver operating characteristic Curve (AUC) were used to evaluate the training and validation performances of the models. The results showed that the single REPT model and its derived ensembles provided a satisfactory accuracy for the prediction of landslides. The D-REPT model with RMSE = 0.351 and AUC = 0.907 was identified as the most accurate model, followed by RSS-REPT (RMSE = 0.353 and AUC = 0.898), B-REPT (RMSE = 0.396 and AUC = 0.876), and the single REPT model (RMSE = 0.398 and AUC = 0.836), respectively. The prominent ensemble models proposed and verified in this study provide engineers and modelers with insights for development of more advanced predictive models for different landslide-susceptible areas around the world.

Rainfall-Induced Landslide Prediction Using Machine Learning Models: The Case of Ngororero District, Rwanda.

Landslides fall under natural, unpredictable and most distractive disasters. Hence, early warning systems of such disasters can alert people and save lives. Some of the recent early warning models make use of Internet of Things to monitor the environmental parameters to predict the disasters. Some other models use machine learning techniques (MLT) to analyse rainfall data along with some internal parameters to predict these hazards. The prediction capability of the existing models and systems are limited in terms of their accuracy. In this research paper, two prediction modelling approaches, namely random forest (RF) and logistic regression (LR), are proposed. These approaches use rainfall datasets as well as various other internal and external parameters for landslide prediction and hence improve the accuracy. Moreover, the prediction performance of these approaches is further improved using antecedent cumulative rainfall data. These models are evaluated using the receiver operating characteristics, area under the curve (ROC-AUC) and false negative rate (FNR) to measure the landslide cases that were not reported. When antecedent rainfall data is included in the prediction, both models (RF and LR) performed better with an AUC of 0.995 and 0.997, respectively. The results proved that there is a good correlation between antecedent precipitation and landslide occurrence rather than between one-day rainfall and landslide occurrence. In terms of incorrect predictions, RF and LR improved FNR to 10.58% and 5.77% respectively. It is also noted that among the various internal factors used for prediction, slope angle has the highest impact than other factors. Comparing both the models, LR model’s performance is better in terms of FNR and it could be preferred for landslide prediction and early warning. LR model’s incorrect prediction rate FNR = 9.61% without including antecedent precipitation data and 3.84% including antecedent precipitation data.

How much does the rainfall temporal resolution affect rainfall thresholds for landslide triggering?

In many areas of the world, the prediction of rainfall-induced landslides is usually carried out by means of empirical rainfall thresholds. Their definition is complicated by several issues, among which are the evaluation and quantification of diverse uncertainties resulting from data and methods. Threshold effectiveness and reliability strongly depend on the quality and quantity of rainfall measurements and landslide information used as input. In this work, the influence of the temporal resolution of rainfall measurements on the calculation of landslide-triggering rainfall thresholds is evaluated and discussed. For the purpose, hourly rainfall measurements collected by 172 rain gauges and geographical and temporal information on the occurrence of 561 rainfall-induced landslides in Liguria region (northern Italy) in the period 2004–2014 are used. To assess the impact of different temporal resolutions on the thresholds, rainfall measurements are clustered in increasing bins of 1, 3, 6, 12 and 24 h. A comprehensive tool is applied to each dataset to automatically reconstruct the rainfall conditions responsible for the failures and to calculate frequentist cumulated event rainfall–rainfall duration (ED) thresholds. Then, using a quantitative procedure, the calculated ED thresholds are validated. The main finding of the work is that the use of rainfall measurements with different temporal resolutions results in considerable variations of the shape and the validity range of the thresholds. Decreasing the rainfall temporal resolution, thresholds with smaller intercepts, higher slopes, shorter ranges of validity and higher uncertainties are obtained. On the other hand, it seems that the rainfall temporal resolution does not influence the validation procedure and the threshold performance indicators. Overall, the use of rainfall data with coarse temporal resolution causes a systematic underestimation of thresholds at short durations, resulting in relevant drawbacks (e.g. false alarms) if the thresholds are implemented in operational systems for landslide prediction.

Application of hydrological modelling for temporal prediction of rainfall-induced shallow landslides

In some rainfall-triggered landslides, intensity-duration thresholds can have limited prediction ability; therefore, investigation of alternative approaches that can be used for temporal prediction of rainfall-induced landslides is needed. This paper presents a methodology for predicting rainfall-induced shallow landslides based on a lumped conceptual hydrological model. The production storage level during the rainfall event and the rainfall sum during the event are used for landslide prediction. Based on these two hydro-meteorological variables a threshold is defined that could be used for rainfall-induced landslides prediction as part of an early warning system. The presented methodology is tested using the meso-scale Selška Sora River catchment case study in western Slovenia where 20 active landslides from the Slovenian National Landslide Database are used to calibrate and evaluate the methodology performance. The results are compared to three different (i.e. local, regional, and global) intensity-duration thresholds. The results of the presented approach are superior in terms of several goodness-of-fit criteria compared to tested local and global ID thresholds. Because only daily rainfall, evapotranspiration, and discharge data are needed to calibrate the selected hydrological model and only daily rainfall and evapotranspiration to run the model, the presented approach could also be useful for data-scarce areas where detailed physically based landslide prediction models that require many data cannot be constructed. Moreover, we have also derived the probabilistic version of the proposed threshold for triggering of shallow landslides using copula functions.

Development of elastic wave velocity threshold for rainfall-induced landslide prediction and early warning

The main aim of this paper is to define an elastic wave velocity threshold for rainfall-induced landslide prediction and early warning. To achieve this goal, several key questions need to be addressed, such as what threshold of wave velocity should be employed for issuing early warning? And how such thresholds should be selected? In an attempt to answer these questions, a total of 32 centrifuge experiments were conducted on model slopes. We first determined the critical deformation prior to initiation of shear failure. Then, the corresponding wave velocities at the critical deformation were compiled to establish a threshold of wave velocity to be used for early warning of landslides. Finally, a large-scale slope model test was performed to confirm the proposed threshold of elastic wave velocity. Based on the wave monitoring results, a warning is proposed to be issued at normalized elastic wave velocity of 0.9, and the slope toe is preferable for wave monitoring.

Spatial Prediction of Rainfall-Induced Landslides Using Aggregating One-Dependence Estimators Classifier

In this study, the spatial prediction of rainfall-induced landslides at the Pauri Gahwal area, Uttarakhand, India has been done using Aggregating One-Dependence Estimators (AODE) classifier which has not been applied earlier for landslide problems. Historical landslide locations have been collated with a set of influencing factors for landslide spatial analysis. The performance of the AODE model has been assessed using statistical analyzing methods and receiver operating characteristic curve technique. The predictive capability of the AODE model has also been compared with other popular landslide models namely Support Vector Machines (SVM), Radial Basis Function Neural Network (ANN-RBF), Logistic Regression (LR), and Naive Bayes (NB). The result of analysis illustrates that the AODE model has highest predictability, followed by the SVM model, the ANN-RBF model, the LR model, and the NB model, respectively. Thus AODE is a promising method for the development of better landslide susceptibility map for proper landslide hazard management.

Estimating actual and potential bare soil evaporation from silty pyroclastic soils: Towards improved landslide prediction

The estimation of evaporative fluxes and their effects on soil suction is assuming a prominent role in the field of interpretation and early-warning prediction of rainfall-induced landslides. Evaporation models refer essentially to sands or plastic (silty and clayey) soils. Models validated specifically for non-plastic silty pyroclastic soils, usually characterized by very high porosity, are instead unavailable. This deficit arises although silty pyroclastic covers are widely spread across the world, increasingly involved in rainfall-induced landslides and recognized showing particular hydrological behaviour. A number of questions may be raised about the issue: (i) may any evaporative models be reliably extended to silty pyroclastic soils?; (ii) what atmospheric variables need to be monitored at least to reliably predict evaporation fluxes in these soils?; and (iii) how accurate evaporation estimations are if they are referred to silty pyroclastic covers for early warning purposes? This study addresses these questions by assessing the capabilities of several simplified models in estimating evaporative (potential and actual) fluxes for silty pyroclastic soils. To this aim, a large-scale lysimeter, consisting in a silty pyroclastic layer exposed to the atmosphere and comprehensively monitored for both weather forcing and hydrological soil variables, is adopted. It provides a dataset of observations suitable to calibrate and validate the selected evaporation models. Moreover, the availability of weather observations makes it possible to define the minimum set of equipment required to attain reliable evaporation estimation. This study shows that: (i) the adoption of a literature-based calibration can produce misleading estimates of actual evaporation, whereas the model performances after a lysimeter-based recalibration are satisfactory; (ii) reducing the weather measurements can induce an overestimation of predicted fluxes up to 50%; and (iii) the investigated models quite accurately predict water out-coming fluxes while running in early warning predictions.

Enhancing the accuracy of rainfall-induced landslide prediction along mountain roads with a GIS-based random forest classifier

Along mountain roads, rainfall-triggered landslides are typical disasters that cause significant human casualties. Thus, to establish effective mitigation measures, it would be very useful were government agencies and practicing land-use planners to have the capability to make an accurate landslide evaluation. Here, we propose a machine learning methodology for the spatial prediction of rainfall-induced landslides along mountain roads which is based on a random forest classifier (RFC) and a GIS-based dataset. The RFC is used as a supervised learning technique to generalize the classification boundary that separates the input information of ten landslide conditioning factors (slope, aspect, relief amplitude, toposhape, topographic wetness index, distance to roads, distance to rivers, lithology, distance to faults, and rainfall) into two distinctive class labels: ‘landslide’ and ‘non-landslide’. Experimental results with a cross validation process and sensitivity analysis on the RFC model parameters reveal that the proposed model achieves a superior prediction accuracy with an area under the curve of 0.92. The RFC significantly outperforms other benchmarking methods, including discriminant analysis, logistic regression, artificial neural networks, relevance vector machines, and support vector machines. Based on our experimental outcome and comparative analysis, we strongly recommend the RFC as a very capable tool for spatial modeling of rainfall-induced landslides.

Elastic wave velocity monitoring as an emerging technique for rainfall-induced landslide prediction

Landslides are recurring phenomena causing damages to private property, public facilities, and human lives. The need for an affordable instrumentation that can be used to provide an early warning of slope instability to enable the evacuation of vulnerable people, and timely repair and maintenance of critical infrastructure is self-evident. A new emerging technique that correlates soil moisture changes and deformations in slope surface by means of elastic wave propagation in soil was developed. This approach quantifies elastic wave propagation as wave velocity. To verify its applicability, a series of fixed and varied slope model tests, as well as a large scale model test, were conducted. Analysis of the results has established that the elastic wave velocity continuously decreases in response of moisture content and deformation, and there was a distinct surge in the decrease rate of wave velocity with failure initiation, soil deformation was thus envisaged to have more significant effect on elastic wave velocity than water content. It is proposed that a warning be issued at switch of wave velocity decrease rate. Based on these observations, expected operation of the elastic wave velocity monitoring system for landslide prediction in the field application is presented. Consequently, we conclude that the elastic wave velocity monitoring technique has the potential to contribute to landslide prediction.

Spatial prediction of rainfall-induced landslides for the Lao Cai area (Vietnam) using a hybrid intelligent approach of least squares support vector machines inference model and artificial bee colony optimization

The main objective of this study is to produce a landslide susceptibility map for the Lao Cai area (Vietnam) using a new hybrid intelligent method based on least squares support vector machines (LSSVM) and artificial bee colony (ABC) optimization, namely LSSVM-BC. LSSVM and ABC are state-of-the-art soft computing techniques that have been rarely utilized in landslide susceptibility assessment. LSSVM is adopted to develop landslide prediction model whereas ABC was used to optimize the prediction model by identifying an appropriate set of the LSSVM hyper-parameters. To establish the hybrid intelligent method, a GIS database with ten landslide-influencing factors and 340 landslide locations that occurred mainly during the last 20-years was constructed. These historical landslide locations were collected from the existing inventories that sourced from (i) five landslide projects carried out in this study areas before and (ii) interpretations of SPOT satellite images with resolution of 2.5 m. The study area was geographically split into two different parts, with landslides located in the first part was used for building models whereas the other landslides in the second part was used for the model validation. Performance of the LSSVM-BC model was assessed using the receiver operating characteristic (ROC) curve and area under the curve (AUC). Result shows that the prediction power of the model is good with the area under the curve (AUC) = 0.900. Experiments have pointed out the prediction power of the LSSVM-BC is better than that obtained from the popular support vector machines. Therefore, the proposed model is a promising tool for spatial prediction of landslides at the study area. The landslide susceptibility map is useful for landuse planning for the Lao Cai area.

A Comparative Study of Least Square Support Vector Machines and Multiclass Alternating Decision Trees for Spatial Prediction of Rainfall-Induced Landslides in a Tropical Cyclones Area

The objective of this study is to explore and compare the least square support vector machine (LSSVM) and multiclass alternating decision tree (MADT) techniques for the spatial prediction of landslides. The Luc Yen district in Yen Bai province (Vietnam) has been selected as a case study. LSSVM and MADT are effective machine learning techniques of classification applied in other fields but not in the field of landslide hazard assessment. For this, Landslide inventory map was first constructed with 95 landslide locations identified from aerial photos and verified from field investigations. These landslide locations were then divided randomly into two parts for training (70 % locations) and validation (30 % locations) processes. Secondly, landslide affecting factors such as slope, aspect, elevation, curvature, lithology, land use, distance to roads, distance to faults, distance to rivers, and rainfall were selected and applied for landslide susceptibility assessment. Subsequently, the LSSVM and MADT models were built to assess the landslide susceptibility in the study area using training dataset. Finally, receiver operating characteristic curve and statistical index-based evaluations techniques were employed to validate the predictive capability of these models. As a result, both the LSSVM and MADT models have high performance for spatial prediction of landslides in the study area. Out of these, the MADT model (AUC = 0.853) outperforms the LSSVM model (AUC = 0.803). From the landslide study of Luc Yen district in Yen Bai province (Vietnam), it can be conclude that the LSSVM and MADT models can be applied in other areas of world also for and spatial prediction. Landslide susceptibility maps obtained from this study may be helpful in planning, decision making for natural hazard management of the areas susceptible to landslide hazards.

Type of Slope Failure Identified from Pore Water Pressure and Moisture Content Measurements

Rainfall-induced landslides occur in many parts of the world and causing a lot of the damages. For effective prediction of rainfall-induced landslides the comprehensive understanding of the failure process is necessary. Under different soil and hydrological conditions experiments were conducted to investigate and clarify the mechanism of slope failure. The failure in model slope was induced by sprinkling the rainfall on slope composed of sandy soil in small flume. Series of tests were conducted in small scale flume to better understand the failure process in sandy slopes. The moisture content was measured with advanced Imko TDR (Time Domain Reflectrometry) moisture sensors in addition to measurements of pore pressure with piezometers. The moisture content increase rapidly to reach the maximum possible water content in case of higher intensity of rainfall, and higher intensity of the rainfall causes higher erosion as compared to smaller intensity of the rainfall. The controlling factor for rainfall-induced flowslides was density of the slope, rather than intensity of the rainfall and during the flowslide the sudden increase in pore pressure was observed. Higher pore pressure was observed at the toe of the slope as compared to upper part of the slope.

An algorithm for the objective reconstruction of rainfall events responsible for landslides

In Italy, rainfall-induced landslides are recurrent phenomena that cause societal and economic damage. Thus, assessing the rainfall conditions responsible for landslides is important and may contribute to reducing risk. The prediction of rainfall-induced landslides relies primarily on empirical rainfall thresholds. However, the thresholds are affected by uncertainties that limit their use in operational warning systems. A source of uncertainty lies in the characterization of the rainfall events responsible for landslides. Objective criteria for the definition of rainfall events are lacking. To overcome the problem, we propose an algorithm that reconstructs the rainfall events, identifies the rainfall conditions that have resulted in landslides, and measures the duration and the cumulated rainfall for the events. The algorithm is independent from the local settings and uses a reduced set of parameters to account for different physical settings and operational conditions. We tested the algorithm in Sicily, Italy, with rainfall and landslide information between January 2002 and December 2012. The rainfall conditions responsible for landslides identified by the algorithm were compared against results obtained manually. The algorithm was proven capable of accurately reconstructing most (87.7 %) of the rainfall events. For each landslide, the algorithm identified a variable number of rainfall conditions responsible for the failures, which are equally likely triggers of the landslide. This opens the possibility of evaluating the uncertainty introduced by different criteria to determine the rainfall events responsible for landslides. Use of the algorithm shall contribute to reducing the uncertainty in the definition of landslide-triggering rainfall events, to compiling large catalogues of rainfall events with landslides and to determining reliable rainfall thresholds for possible landslide occurrence.

A simulation chain for early prediction of rainfall-induced landslides

A new simulation chain for early prediction of rainfall-induced landslides in unsaturated soils is presented. It includes a special computational weather code for forecasting the evolution of the synoptic weather and its changes due to interaction with the Earth’s surface (rainfall pattern), and a hydro-mechanical code to analyse rainfall effects on slope stability by computing degree of saturation and pore pressure changes due to rainwater infiltration. The linkage between these two numerical codes is ensured by an interface with the aim of bringing the data provided by the first code, which operates at basin or slope scale. The simulation chain can work in computational times that may be considered suitable for civil protection operations.

Analysis of the relationship of landslide occurrence with rainfall: a case study of Wudu County, China

Landslides are a significant hazard in many parts of the world and represent an important geohazard in China. Rainfall is the primary triggering agent for landslides and often used for prediction slope failures. However, the relationship between rainfall and landslide occurrences is very complex. Great efforts have been made on the study of regional rainfall-induced landslide forecasting models in recent years; still, there is no commonly accepted method for rainfall-induced landslide prediction. In this paper, the quantitative antecedent soil water status (ASWS) model is applied to investigate the influence of daily and antecedent rainfall on the triggering of landslides and debris flows. The study area is Wudu County in Gansu Province, an area which exhibits frequent landslide occurrences. The results demonstrate a significant influence of high intensity rainfall events on landslide triggering. Still, antecedent rainfall conditions are very important and once a threshold of approximately 20 mm is exceeded, landslides and debris flows can occur even without additional rainfall. The study presented could also facilitate the implementation of a regional forecasting scheme once additional validation has been carried out.