A case study from Koyulhisar (Sivas-Turkey) for landslide susceptibility mapping by artificial neural networks

Yilmaz, Işık

doi:10.1007/s10064-009-0185-2

Cited by 148 publications

(62 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Deterministic models are completely based on mathematical relationships that depend on the physical laws in which the relation between resisting and driving forces can be calculated for the mass movements. The most important data that are required for the deterministic models are engineering characteristics of the rocks and soils, slope geometry, discontinuity characteristics, and hydrological conditions (Yilmaz 2009). The main problem with the deterministic models is the need for intensive data from individual slopes, which makes these methods effective for studying only small areas (Ayalew and Yamagishi 2005).…”

Section: Introductionmentioning

confidence: 99%

Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia

Youssef

Pourghasemi²,

Pourtaghi³

et al. 2015

Landslides

632

246

View full text Add to dashboard Cite

Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia Abstract The purpose of the current study is to produce landslide susceptibility maps using different data mining models. Four modeling techniques, namely random forest (RF), boosted regression tree (BRT), classification and regression tree (CART), and general linear (GLM) are used, and their results are compared for landslides susceptibility mapping at the Wadi Tayyah Basin, Asir Region, Saudi Arabia. Landslide locations were identified and mapped from the interpretation of different data types, including high-resolution satellite images, topographic maps, historical records, and extensive field surveys. In total, 125 landslide locations were mapped using ArcGIS 10.2, and the locations were divided into two groups; training (70%) and validating (25%), respectively. Eleven layers of landslide-conditioning factors were prepared, including slope aspect, altitude, distance from faults, lithology, plan curvature, profile curvature, rainfall, distance from streams, distance from roads, slope angle, and land use. The relationships between the landslide-conditioning factors and the landslide inventory map were calculated using the mentioned 32 models (RF, BRT, CART, and generalized additive (GAM)). The models' results were compared with landslide locations, which were not used during the models' training. The receiver operating characteristics (ROC), including the area under the curve (AUC), was used to assess the accuracy of the models. The success (training data) and prediction (validation data) rate curves were calculated. The results showed that the AUC for success rates are 0.783 (78.3%), 0.958 (95.8%), 0.816 (81.6%), and 0.821 (82.1%) for RF, BRT, CART, and GLM models, respectively. The prediction rates are 0.812 (81.2%), 0.856 (85.6%), 0.862 (86.2%), and 0.769 (76.9%) for RF, BRT, CART, and GLM models, respectively. Subsequently, landslide susceptibility maps were divided into four classes, including low, moderate, high, and very high susceptibility. The results revealed that the RF, BRT, CART, and GLM models produced reasonable accuracy in landslide susceptibility mapping. The outcome maps would be useful for general planned development activities in the future, such as choosing new urban areas and infrastructural activities, as well as for environmental protection.

show abstract

Section: Introductionmentioning

confidence: 99%

Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia

Youssef

Pourghasemi²,

Pourtaghi³

et al. 2015

Landslides

632

246

View full text Add to dashboard Cite

show abstract

“…Other different methods have been proposed by several investigators, including weights-of-evidence methods (Bonham-Carter1991; Neuhäuser and Terhorst 2007;Pradhan et al 2010d;Regmi et al 2010a;Pourghasemi et al 2012a, b), modified Bayesian estimation (Chung and Fabbri 1999), weighting factors, weighted linear combinations of instability factors (Ayalew et al 2004), landside nominal risk factors (Saha et al 2005), probabilistic-based frequency ratio model Fabbri 2003, 2005;Lee andPradhan 2006, 2007;Akgűn et al 2008;Pradhan et al 2011Pradhan et al , 2012, certainty factors (Pourghasemi et al 2012a), information values (Saha et al 2005), modified Bayesian estimation (Chung and Fabbri 1999). Among recent models for landslide susceptibility mapping, soft computing techniques such as neuro-fuzzy (Sezer et al 2011;Vahidnia et al 2010;Oh and Pradhan 2011), artificial neural networks (Bui et al 2012a; Lee 2009, 2010a, b;Pradhan et al 2010a, b, d;Pradhan and Buchroithner 2010;Pradhan and Pirasteh 2010;Pradhan2011a;Poudyalet al2010;Yilmaz 2009aYilmaz , b, 2010aChoiet al2012;Zarea et al 2012), fuzzy-logic (Akgu ¨n et al 2012;Bui et al 2012b;Ercanoglu and Gokceoglu 2002;Kanungo et al 2008;Pradhan 2010bPradhan , 2010cPradhan , 2011b…”

mentioning

confidence: 99%

Landslide susceptibility mapping using certainty factor, index of entropy and logistic regression models in GIS and their comparison at Mugling–Narayanghat road section in Nepal Himalaya

et al. 2012

View full text Add to dashboard Cite

Landslide susceptibility maps are vital for disaster management and for plan-ning development activities in the mountainous country like Nepal. In the present study, landslide susceptibility assessment of Mugling-Narayanghat road and its surrounding area is made using bivariate (certainty factor and index of entropy) and multivariate (logistic regression) models. At first, a landslide inventory map was prepared using earlier reports and aerial photographs as well as by carrying out field survey. As a result, 321 landslides were mapped and out of which 241 (75 %) were randomly selected for building landslide susceptibility models, while the remaining 80 (25 %) were used for validating the models. The effectiveness of landslide susceptibility assessment using GIS and statistics is based on appropriate selection of the factors which play a dominant role in slope stability. In this case study, the following landslide conditioning factors 2 were evaluated: slope gradient; slope aspect; altitude; plan curvature; lithology; land use; distance from faults, rivers and roads; topographic wetness index; stream power index; and sediment transport index. These factors were prepared from topographic map, drainage map, road map, and the geological map. Finally, the validation of landslide susceptibility map was carried out using receiver operating characteristic (ROC) curves. The ROC plot estimation results showed that the susceptibility map using index of entropy model with AUC value of 0.9016 has highest prediction accuracy of 90.16 %. Similarly, the susceptibility maps produced using logistic regression model and certainty factor model showed 86.29 and 83.57 % of prediction accuracy, respectively. Furthermore, the ROC plot showed that the success rate of all the three models performed more than 80 % accuracy (i.e. 89.15 % for IOE model, 89.10 % for LR model and 87.21 % for CF model). Hence, it is concluded that all the models employed in this study showed reasonably good accuracy in predicting the landslide susceptibility of Mugling-Narayanghat road section. These landslide suscepti-bility maps can be used for preliminary land use planning and hazard mitigation purpose.

show abstract

“…Statistical approaches are indirect methods of landslide hazard mapping which are essentially data driven in approach and involve determination of ratings/ranks of classes within variables based on spatial statistical relationship of landslides with variables. Statistical techniques of landslide hazard assessment can be easily implemented in a GIS as demonstrated by numerous workers such as Weight of Evidence (Lee et al 2002a;Mathew et al 2007;Neuhäuser and Terhorst 2007;Thiery et al 2007;Dahal et al 2008a, b;Song et al 2008;Regmi et al 2010;Piacentini et al 2012;Chen and Li 2014;Jebur et al 2014;Meten et al 2014;Sujatha et al 2014), artificial neural network (Yilmaz 2009a;Chauhan et al 2010b;Pradhan and Pirasteh 2010;Pradhan and Lee 2010a, b, c;Pradhan and Buchroithner 2010;Bui et al 2012b;Zare et al 2013;Conforti et al 2014), multiple and logistic regression (Dai and Lee 2002;Chau and Chan 2005;Ayalew and Yamagishi 2005;Mathew et al 2009;Chauhan et al 2010a;Bai et al 2010;Das et al 2010;Althuwaynee et al 2014a;Jebur et al 2014) and frequency ratio (Lee and Sambath 2006;Lee and Pradhan 2007;Yilmaz 2009b). Although artificial neural network-based technique is not purely statistical technique, but it is implemented as a statistical technique in the present context and, therefore, can be included in this category.…”

Section: Introductionmentioning

confidence: 99%

Spatial prediction of landslide susceptibility in parts of Garhwal Himalaya, India, using the weight of evidence modelling

Guri

Ray

Patel

2015

Environ Monit Assess

View full text Add to dashboard Cite

Garhwal Himalaya in northern India has emerged as one of the most prominent hot spots of landslide occurrences in the Himalaya mainly due to geological causes related to mountain building processes, steep topography and frequent occurrences of extreme precipitation events. As this region has many pilgrimage and tourist centres, it is visited by hundreds of thousands of people every year, and in the recent past, there has been rapid development to provide adequate roads and building infrastructure. Additionally, attempts are also made to harness hydropower by constructing tunnels, dams and reservoirs and thus altering vulnerable slopes at many places. As a result, the overall risk due to landslide hazards has increased many folds and, therefore, an attempt was made to assess landslide susceptibility using 'Weights of Evidence (WofE)', a well-known bivariate statistical modelling technique implemented in a much improved way using remote sensing and Geographic Information System. This methodology has dual advantage as it demonstrates how to derive critical parameters related to geology, geomorphology, slope, land use and most importantly temporal landslide distribution in one of the data scarce region of the world. Secondly, it allows to experiment with various combination of parameters to assess their cumulative effect on landslides. In total, 15 parameters related to geology, geomorphology, terrain, hydrology and anthropogenic factors and 2 different landslide inventories (prior to 2007 and 2008-2011) were prepared from high-resolution Indian remote sensing satellite data (Cartosat-1 and Resourcesat-1) and were validated by field investigation. Several combinations of parameters were carried out using WofE modelling, and finally using best combination of eight parameters, 76.5 % of overall landslides were predicted in 24 % of the total area susceptible to landslide occurrences. The study has highlighted that using such methodology landslide susceptibility assessment can be carried out in vast stretches of Himalaya in short time in order to assess the impact of development as well as climate change/variability. The resultant map can play a critical role in selecting areas for remedial measures for slope stabilisation as well planning for future development of the region.

show abstract

A case study from Koyulhisar (Sivas-Turkey) for landslide susceptibility mapping by artificial neural networks

Cited by 148 publications

References 32 publications

Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia

Landslide susceptibility mapping using random forest, boosted regression tree, classification and regression tree, and general linear models and comparison of their performance at Wadi Tayyah Basin, Asir Region, Saudi Arabia

Landslide susceptibility mapping using certainty factor, index of entropy and logistic regression models in GIS and their comparison at Mugling–Narayanghat road section in Nepal Himalaya

Spatial prediction of landslide susceptibility in parts of Garhwal Himalaya, India, using the weight of evidence modelling

Contact Info

Product

Resources

About