Landslide Hazard Prediction Research Articles

Landslide Susceptibility Assessment in Vietnam Using Support Vector Machines, Decision Tree, and Naïve Bayes Models

The objective of this study is to investigate and compare the results of three data mining approaches, the support vector machines (SVM), decision tree (DT), and Naïve Bayes (NB) models for spatial prediction of landslide hazards in the Hoa Binh province (Vietnam). First, a landslide inventory map showing the locations of 118 landslides was constructed from various sources. The landslide inventory was then randomly partitioned into 70% for training the models and 30% for the model validation. Second, ten landslide conditioning factors were selected (i.e., slope angle, slope aspect, relief amplitude, lithology, soil type, land use, distance to roads, distance to rivers, distance to faults, and rainfall). Using these factors, landslide susceptibility indexes were calculated using SVM, DT, and NB models. Finally, landslide locations that were not used in the training phase were used to validate and compare the landslide susceptibility maps. The validation results show that the models derived using SVM have the highest prediction capability. The model derived using DT has the lowest prediction capability. Compared to the logistic regression model, the prediction capability of the SVM models is slightly better. The prediction capability of the DT and NB models is lower.

Calibration and resolution effects on model performance for predicting shallow landslide locations in Taiwan

In this paper we optimise the spatially explicit prediction of landslide hazard, landslide triggering and subsequent movement downslope of materials for a mountainous catchment in Taiwan. The location prediction is optimised by subsequently adding three location parameters: rainfall distribution, land-use classes and DEM derived slopes. Then the three most important model parameters are calibrated to find the best prediction for both stable and unstable areas. The landslides predicted by the LAPSUS-LS model are compared with a landslide inventory to validate the output. The optimal model settings for the calibration area are then applied to a validation area. Results show that model performance can be improved by adding the spatial distribution of rainfall and by stratifying according to land-use classes. Landslide prediction is better with fine resolution DEMs, mainly because the local topography is smoothed in coarser resolutions. Although in general the amount of landslides is over-predicted, the overall performance indicates that the model is able to capture the important factors determining landslide location. Additional spatially distributed data such as regolith or soil depth and regeneration rates of the legacy effect can further enhance the model's prediction.

Combining neural network with fuzzy, certainty factor and likelihood ratio concepts for spatial prediction of landslides

The landslide studies can be categorized as pre- and postdisaster studies. The predisaster studies include spatial prediction of potential landslide zones known as landslide susceptibility zonation (LSZ) mapping to identify the areas/locales susceptible to landslide hazard. The LSZ maps provide an assessment of the safety of existing habitations and infrastructural/functional elements and help plan further developmental activities in the hilly regions. Landslides are one of the natural geohazards that affect at least 15% of land area of India. Different types of landslides occur frequently in geodynamical active domains of the Himalayas. In India, various techniques have been developed and adopted for LSZ mapping of different regions. However, the technique for LSZ mapping is not yet standardized. The present research is an attempt in this direction only. In our earlier work (Kanungo et al. 2006), a detailed study on conventional, artificial neural network (ANN)- black box-, fuzzy set-based and combined neural and fuzzy weighting techniques for LSZ mapping in Darjeeling Himalayas has been documented. In this paper, other techniques such as combined neural and certainty factor concept along with combined neural and likelihood ratio techniques have been assessed in comparison with combined neural and fuzzy technique for the preparation of LSZ maps of the same study area in parts of Darjeeling Himalayas. It is observed from the present study that the LSZ map produced using combined neural and fuzzy approach appears to be the most accurate one as in this case only 2.3% of the total area is found to be categorized as very high susceptibility zone and contains 30.1% of the existing landslide area. This approach can serve as one of the key objective approaches for spatial prediction of landslide hazards in hilly terrain.

Evaluation of slope stability with topography and slope stability analysis method

In existing researches connected with landslide, natural slope was regard as a parallel slope, but it was found that many landslides were occurred on the concave and convex slopes besides parallel slopes. This study was performed to evaluate the effect of shape and length of slope and soil depth for slope stability analysis and to compare infinite slope analysis with circular failure analysis for grid by grid analysis in prediction of GIS-based landslide hazard. As grid size is increased, average slope and curvature were decreased and the decreasing tendency was prominent in mountainous areas. Comparing infinite slope analysis and circular failure analysis for various degree of slope and soil depth, safety factor by infinite slope analysis is larger than that by circular failure analysis. Result showed that the factor safety depends on the shape of slope which is composed of two continuous slopes, if grid size is small (such as 5 m×5 m). So after due consideration for soil depth and topography, proper slope stability analysis must be performed.

GIS-based landslide hazard zonation model and its application

Multi-source earth observation data are highly desirable in nowaday landslide hazard prediction modeling. Landslide Hazard Zonation (LHZ) is a very important content of landslide hazard prediction modeling. In this paper, take Wan county for instance, we investigate the potentials of derivation from hybrid data sets to study landslide hazard zonation based on the ordinal scale relative weighting-rating technique. LHZ is then performed with chosen factor layers including: buffer map of thrusts, lithology, slope angle and relative relief calculated from DEM, NDVI, buffer map of drainage and Lineaments extracted from the digital satellite scene (TM). Then the Landslide Hazard Index (LHI) value is calculated and landslide hazard zonation is decided by slicing LHI histogram. The statistics demonstrate that High stable slope zone, Stable slope zone, Quasi-stable slope zone, Relatively unstable slope zone, Unstable slope zone and Defended slope zone account for 2.20%, 14.02%, 39.88%, 28.27%, 12.17% and 3.47% respectively. Then, GPS deformation control points on the landslide bodies are used to verify the validity of the LHZ technique.

Knowledge-Based Landslide Susceptibility Zonation System

The landslide susceptibility of a region is important for socioeconomic considerations and engineering applications. Thus, an automated system for mapping of landslide susceptibility could be of significant benefit for society. In this paper, a knowledge-based landslide susceptibility zonation (LSZ) system has been proposed. The system consists of input, understanding, expert, and output modules. The input module accepts thematic images of contributing factors for landslides. The understanding module interprets input images to extract relevant information as required by the expert module. The expert module consists of knowledge base and inference strategy to categorize a region into different landslide intensities. Finally the output module provides a LSZ map. It is a pixel-based system and provides output having the scale same as that of the input maps. The system has been tested to prepare a landslide susceptibility map for the Tehri-Garhwal region in India's lower Himalayas, and further validated with studies for two other different regions. The proposed system provides output commensurate with that provided by experts. The categories of hazard zones have a discrepancy as little as 6.2% in high hazard zones and near to 1.5% and 4% in moderate and low hazard zones, respectively. The high hazard zones in the LSZ maps from the proposed system are supersets of that obtained by experts (i.e., the proposed system provides safer LSZ map). Thus, it can be concluded that the proposed system can be used for preparation of LSZ maps. In the future, the methodology may be extended for real time assessment and prediction of landslide hazards.

Regional Landslide Hazard Warning and Risk Assessment

Regional landslide hazard prediction and warning is still a difficult problem and hot topic in the research of landslide hazards. In the past decade, the researches mainly focused on the analysis of the combination of rainfall and geological environment. In this paper, we summarize the current studies of landslide hazard prediction and risk assessment, and propose that the combination of hazard prediction with risk management is not only the need for hazard prediction and prevention, but also the trend in the future. Basic theories are discussed from two aspects: the spatial prediction and time warning. Based on these theories, we set up a landslide hazard information management system and a real-time warning information releasing system, which is developed on MapGIS platform. Taking landslides as examples in Yongjia City, Zhejiang Province, during the period of typhoon Rananim of 2004, we have studied the spatial landslide hazard prediction, life vulnerability assessment and economic risk assessment.

Spatial Prediction of Landslide Hazard Using Logistic Regression and ROC Analysis

An empirical modeling of road related and non-road related landslide hazard for a large geographical area using logistic regression in tandem with signal detection theory is presented. This modeling was developed using geographic information system (GIS) and remote sensing data, and was implemented on the Clearwater National Forest in central Idaho. The approach is based on explicit and quantitative environmental correlations between observed landslide occurrences, climate, parent material, and environmental attributes while the receiver operating characteristic (ROC) curves are used as a measure of performance of a predictive rule. The modeling results suggest that development of two independent models for road related and non-road related landslide hazard was necessary because spatial prediction and predictor variables were different for these models. The probabilistic models of landslide potential may be used as a decision support tool in forest planning involving the maintenance, obliteration or development of new forest roads in steep mountainous terrain.

Spatial Prediction of Landslide Hazard Using Fuzzy k-means and Dempster-Shafer Theory

Landslide databases and input parameters used for modeling landslide hazard often contain imprecisions and uncertainties inherent in the decision-making process. Dealing with imprecision and uncertainty requires techniques that go beyond classical logic. In this paper, methods of fuzzy k-means classification were used to assign digital terrain attributes to continuous landform classes whereas the Dempster-Shafer theory of evidence was used to represent and manage imprecise information and to deal with uncertainties. The paper introduces the integration of the fuzzy k-means classification method and the Dempster-Shafer theory of evidence to model landslide hazard in roaded and roadless areas illustrated through a case study in the Clearwater National Forest in central Idaho, USA. Sample probabilistic maps of landslide hazard potential and uncertainties are presented. The probabilistic maps are intended to help decision-making in effective forest management and planning.

Integrating a fuzzy k -means classification and a Bayesian approach for spatial prediction of landslide hazard

A robust method for spatial prediction of landslide hazard in roaded and roadless areas of forest is described. The method is based on assigning digital terrain attributes into continuous landform classes. The continuous landform classification is achieved by applying a fuzzy k-means approach to a watershed scale area before the classification is extrapolated to a broader region. The extrapolated fuzzy landform classes and datasets of road-related and non road-related landslides are then combined in a geographic information system (GIS) for the exploration of predictive correlations and model development. In particular, a Bayesian probabilistic modeling approach is illustrated using a case study of the Clearwater National Forest (CNF) in central Idaho, which experienced significant and widespread landslide events in recent years. The computed landslide hazard potential is presented on probabilistic maps for roaded and roadless areas. The maps can be used as a decision support tool in forest planning involving the maintenance, obliteration or development of new forest roads in steep mountainous terrain.

Rain-induced failures on residual soil slopes

The hilly region of Western Ghats, India, is a site of frequent landslides, which have caused a severe loss of lives and property. The region is characterised by the presence of residual soils, rugged topography, high relief, and medium to steep slopes. During the heavy rainfall, the pore water pressure is built quickly in the residual soils.
 This study deals with a probabilistic stability analysis of rain-induced failures on residual soil slopes. A mass movement analysis is also attempted using the momentum transfer law for the Parmachi Landslide. Attention is given to the associated risk of sliding, incorporating important factors, which influence the failure of natural slopes. A landslide hazard prediction procedure is presented and its application to debris flows at Parmachi is shown as an example. The probability of failure is evaluated in conjunction with the consequences of failure. Decision analysis is also tried with the probabilistic slope stability analysis of the Parmachi Landslide. The evaluated risk and decision theory are used to suggest a decision-making process for landslide hazard mitigation.

Prediction and mapping of landslide hazard

The paper describes a mechanics-based methodology for prediction of landslide hazard on hillside slopes. The principal components are estimation of infiltration and groundwater response, and estimation of safety factor and failure probability. Predictions with the models are made by using the best estimates of the input parameters and the associated uncertainties. The uncertainties about the inputs and the sensitivities of the model outputs to uncertainties about the inputs are evaluated and translated into failure probability. Our objective was to provide a method that can be used to estimate landslide hazard over a large area prior to making detailed site investigations. The method was applied to slopes along the South Fork of Tilton River, in the Cascade Mountains of Washington State. Predicted hazards were compared with results of a landslide inventory. The data used in the examples were derived from published sources. However, the procedures developed are not limited to this condition and can readily incorporate additional information derived from more detailed investigations.Key words: failure probability, groundwater, hazard, landslides, mapping, slope stability.

Prediction and mapping of landslide hazard

Prediction and mapping of landslide hazard

Use of GIS Technology in the Prediction and Monitoring of Landslide Hazard

Technologies such as Geographical Information Systems (GIS) have raised great expectations as potential means of coping with natural disasters, including landslides. However, several misconceptions on the potential of GIS are widespread. Prominent among these is the belief that a landslide hazard map obtained by systematic data manipulation within a GIS is assumed to be more objective than a comparable hand-made product derived from the same input data and founded on the same conceptual model. Geographical data can now be handled in a GIS environment by users who are not experts in either GIS or natural hazard process fields. The reality of the successful application of GIS within the landslide hazard domain seems to be somewhat less attractive than current optimistic expectations. In spite of recent achievements, the use of GIS in the domain of prevention and mitigation of natural catastrophes remains a pioneering activity. Diffusion of the technology is still hampered by factors such as the difficulty in acquiring appropriate raw data, the intrinsic complexity of predictive models, the lack of efficient graphical user interfaces, the high cost of digitisation, and the persistence of bottlenecks in hardware capabilities. In addition, researchers are investing more in tuning-up hazard models founded upon existing, often unreliable data than in attempting to initiate long-term projects for the acquisition of new data on the causes of catastrophic events. Governmental institutions are frequently involved in risk reduction projects whose design and implementation appear to be governed more by political issues than by technical ones. There is an unfortunate general tendency to search for data which can be collected at low cost rather than attempting to capture the information which most readily explains the causes of a disaster. If the technical, cultural, economic and political reasons for this unhealthy state cannot be adequately tackled, the International Decade for Natural Disaster Reduction will probably come to an end without achieving significant advances in the prediction and control of natural disasters.

Automatic extraction of watershed characteristics using spatial analysis techniques with application to groundwater mapping

Spatial analysis techniques were developed using a raster-based image processing system to derive geometric and morphologic characteristics of watersheds from a digital elevation model. The watershed data layers were overlaid with digital soils information for two 7 1′ 2 quadrangles in the Cascade Mountains of Washington. The data layers were then integrated with a ground-water prediction model and maps of groundwater level zones were prepared from the digital data sets. This information can be useful in assessing the stability of slopes in landslide hazard prediction.

GIS-based quantitative landslide hazard prediction modelling in natural hillslope, Agra Khola watershed, central Nepal

The influence of geological and geomorphological variables were spatially integrated to develop landslide hazard prediction model in the Agra Khola watershed of central Nepal where a large number of landslides triggered off due to extreme weather event of July 19-21, 1993. A quantitative technique of multivariate analysis was performed to predict elements or observations of landslides successfully into different hazard levels in the area. The predicted landslide hazard was validated and spatially relevancy of the prediction is established. The GIS-based prediction model possessed objectivity and reproducibility, and also improved the landslide hazard mapping in the natural hillslope. doi: 10.3126/bdg.v10i0.1421 Bulletin of the Department of Geology, Tribhuvan University, Kathmandu, Nepal, Vol. 10, 2007, pp. 63-70