Spatial data analysis and application of evidential belief functions to shallow landslide susceptibility mapping at Mt. Umyeon, Seoul, Korea

We generated high-quality shallow landslide susceptibility maps for Bijar County, Kurdistan Province, Iran, using Random Forest (RAF), an ensemble computational intelligence method and three meta classifiers—Bagging (BA, BA-RAF), Random Subspace (RS, RS-RAF), and Rotation Forest (RF, RF-RAF). Modeling and validation were done on 111 shallow landslide locations using 20 conditioning factors tested by the Information Gain Ratio (IGR) technique. We assessed model performance with statistically based indexes, including sensitivity, specificity, accuracy, kappa, root mean square error (RMSE), and area under the receiver operatic characteristic curve (AUC). All four machine learning models that we tested yielded excellent goodness-of-fit and prediction accuracy, but the RF-RAF ensemble model (AUC = 0.936) outperformed the BA-RAF, RS-RAF (AUC = 0.907), and RAF (AUC = 0.812) models. The results also show that the Random Forest model significantly improved the predictive capability of the RAF-based classifier and, therefore, can be considered as a useful and an effective tool in regional shallow landslide susceptibility mapping.

Download Full-text

Use of GIS and Remote Sensing for Landslide Susceptibility Mapping

Advances in Environmental Engineering and Green Technologies - Advanced Methodologies and Technologies in Engineering and Environmental Science ◽

10.4018/978-1-5225-7359-3.ch026 ◽

2019 ◽

pp. 384-398 ◽

Cited By ~ 2

Author(s):

Arzu Erener ◽

Gulcan Sarp ◽

Sebnem H. Duzgun

Keyword(s):

Remote Sensing ◽

Information Systems ◽

Spatial Data ◽

Landslide Susceptibility ◽

Geographical Information Systems ◽

Vegetation Index ◽

Susceptibility Mapping ◽

Landslide Susceptibility Mapping ◽

Gis And Remote Sensing ◽

Data Layers

In recent years, geographical information systems (GISs) and remote sensing (RS) have proven to be common tools adopted for different studies in different scientific disciplines. GIS is defined as a set of tools for the input, storage, retrieval, manipulation, management, modeling, analysis, and output of spatial data. RS, on the other hand, can play a role in the production of a data and in the generation of thematic maps related to spatial studies. This study focuses on use of GIS and RS data for landslide susceptibility mapping. Five factors including normalized difference vegetation index (NDVI) and topographic wetness index (TWI), slope, lineament density, and distance to roads were used for the grid-based approach for landslide susceptibility mappings. Results of this study suggest that geographic information systems can effectively be used to obtain susceptibility maps by compiling and overlaying several data layers relevant to landslide hazards.

Download Full-text

Shallow landslide prediction and analysis with risk assessment using a spatial model in the coastal region in the state of São Paulo, Brazil

Natural Hazards and Earth System Sciences Discussions ◽

10.5194/nhessd-1-5199-2013 ◽

2013 ◽

Vol 1 (5) ◽

pp. 5199-5236

Author(s):

P. I. M. Camarinha ◽

V. Canavesi ◽

R. C. S. Alvalá

Keyword(s):

Risk Assessment ◽

Spatial Data ◽

Urban Areas ◽

Coastal Region ◽

Spatial Prediction ◽

Shallow Landslide ◽

Susceptibility Mapping ◽

Landslide Susceptibility Mapping ◽

Landslide Risk ◽

Serra Do Mar

Abstract. In Brazil, most of the disasters involving landslide occur in coastal regions, with population density concentrated on steep slopes. Thus, different approaches have been used to evaluate the landslide risk, although the greatest difficulty is related to the scarcity of spatial data with good quality. In this context, four cities located on the southeast coast of Brazil – Santos, Cubatão, Caraguatatuba and Ubatuba – in a region with the rough reliefs of the Serra do Mar and with a history of natural disasters were evaluated. Spatial prediction by fuzzy gamma technique was used for the landslide susceptibility mapping, considering environmental variables from data and software in the public domain. To validate the susceptibility mapping results, it was overlapped with risk sectors provided by the Geological Survey of Brazil (CPRM). A positive correlation was observed between the classes most susceptible and the location of these sectors. The results were also analyzed from the categorization of risk levels provided by CPRM. To compare the approach with other studies using landslide-scar maps, correlated indexes were evaluated, which also showed satisfactory results, thus indicating that the methodology presented is appropriate for risk assessment in urban areas and can be replicated to municipalities that do not have risk areas mapped.

Download Full-text