Advertisement

Arabian Journal of Geosciences

, Volume 7, Issue 2, pp 725–742 | Cite as

Application of frequency ratio, statistical index, and weights-of-evidence models and their comparison in landslide susceptibility mapping in Central Nepal Himalaya

  • Amar Deep Regmi
  • Krishna Chandra Devkota
  • Kohki Yoshida
  • Biswajeet Pradhan
  • Hamid Reza Pourghasemi
  • Takashi Kumamoto
  • Aykut Akgun
Original Paper

Abstract

The Mugling–Narayanghat road section falls within the Lesser Himalaya and Siwalik zones of Central Nepal Himalaya and is highly deformed by the presence of numerous faults and folds. Over the years, this road section and its surrounding area have experienced repeated landslide activities. For that reason, landslide susceptibility zonation is essential for roadside slope disaster management and for planning further development activities. The main goal of this study was to investigate the application of the frequency ratio (FR), statistical index (SI), and weights-of-evidence (WoE) approaches for landslide susceptibility mapping of this road section and its surrounding area. For this purpose, the input layers of the landslide conditioning factors were prepared in the first stage. A landslide inventory map was prepared using earlier reports, aerial photographs interpretation, and multiple field surveys. A total of 438 landslide locations were detected. Out these, 295 (67 %) landslides were randomly selected as training data for the modeling using FR, SI, and WoE models and the remaining 143 (33 %) were used for the validation purposes. The landslide conditioning factors considered for the study area are slope gradient, slope aspect, plan curvature, altitude, stream power index, topographic wetness index, lithology, land use, distance from faults, distance from rivers, and distance from highway. The results were validated using area under the curve (AUC) analysis. From the analysis, it is seen that the FR model with a success rate of 76.8 % and predictive accuracy of 75.4 % performs better than WoE (success rate, 75.6 %; predictive accuracy, 74.9 %) and SI (success rate, 75.5 %; predictive accuracy, 74.6 %) models. Overall, all the models showed almost similar results. The resultant susceptibility maps can be useful for general land use planning.

Keywords

Landslides Frequency ratio Statistical index Weights-of-evidence GIS Himalaya 

Introduction

The Himalayas are one of the most active mountains which have resulted from the collision of the Indian Plate and the Tibetan Plate some 50 million years ago. Since Nepal occupies a major portion of the Himalaya, it has always been recognized as the area prone to natural hazards, e.g., earthquakes and landslides. The annual loss of lives and property due to these natural hazards are significantly high in the Himalayan region. In Nepal itself, the average number of deaths from 1983 to 2007 is about 329 persons, whereas the annual loss of life is as high as 1,336 persons in 1993 (Dhital 2005). The death due to landslides alone is about 111 persons per year.

Intensive rainfall (446 mm in 24 h at Devghat, the confluence of Kali Gandaki and Trishuli River) of July 30 and 31, 2003 triggered hundreds of landslides, slope failures, rock falls, and debris falls along the highway and its surrounding region (Adhikari 2009) (Fig. 1). It was further aggravated by mass movements during the monsoon of 2006, blocking the traffic for several weeks (DWIDP 2009). This road section is one of the busiest strategic roads of Nepal, which connects the Terai with the capital and other central and western mountain districts. Thus, to maintain the traffic as well as to minimize the losses of life and property, landslide susceptibility assessment is essential for this road section.
Fig. 1

Landslide location map of the study area

Landslide susceptibility can be defined as the propensity of an area to generate landslides (Guzzetti et al. 2006). There exists several hazard modeling methodology in literature and can be categorized into heuristic, deterministic, and statistical approaches (van Westen et al. 1997; Aleotti and Chowdhury 1999; Guzzetti et al. 1999; Dai et al. 2002; Alexander 2008; Sharma and Kumar 2008; Wan et al. 2009; Wan and Lei 2009). A heuristic approach (Deoja et al. 1991; Anbalagan 1992; Nagarajan et al. 2000; Saha et al. 2002) is a direct or semidirect mapping methodology in which a relationship is established between the occurrence of slope failures and the causative factors. They are based on the assumption that the relationships between landslide susceptibility and the preparatory variables are known and are specified in the models. A set of variables is then entered into the model to estimate landslide susceptibility (Gupta and Joshi 1990). In this approach, the relative importance of each causative factor for slope instability is evaluated according to subjective expert’s knowledge or statistics of landslide distribution and analysis (van Westen et al. 1999). Therefore, assigning weight values and ratings to the variables is very subjective, and the results are often not reproducible.

In deterministic analysis, the landslide hazard is determined using slope stability models, resulting in the calculation of factor of safety. These models require a large amount of detailed input data, derived from laboratory tests. Hence, these models are applicable in small areas at large scales.

In the statistical approach, multivariate and bivariate statistical techniques are used for landslide susceptibility mapping throughout the world. Bivariate statistical analyses involve the idea of comparing a landslide inventory map with maps of landslide influencing parameters in order to rank the corresponding classes according to their role in landslide formation. Ranking is normally carried out using landslide densities. A variety of multivariate statistical approaches exist, but those commonly used to map landslide susceptibility include discriminant analyses and logistic regression (Lee 2007a; Pradhan 2010a). Different other methods have been proposed by several investigators, including weights-of-evidence (WoE) methods (Bonham-Carter 1991; Gokceoglu et al. 2005; Neuhäuser and Terhorst 2007; Mathew et al. 2007; Bui et al. 2008; Zhu and Wang 2009; Regmi et al. 2010a, b; Pourghasemi et al. 2012a), weighting factors, weighted linear combinations of instability factors (Ayalew et al. 2004), landside nominal risk factors (Gupta and Joshi 1990; Saha et al. 2005), probabilistic models (Carranza and Hale 2001, 2002; Chung and Fabbri 2003, 2005; Lee 2005; Lee and Pradhan 2006, 2007; Akgün et al. 2011; Pradhan et al. 2012; Mohammady et al. 2012; Pourghasemi et al. 2012b, d), certainty factors (Binaghi et al. 1998), information values (Saha et al. 2005; Wan et al. 2008), modified Bayesian estimation (Chung and Fabbri 1998). Additionally, there exists some other data mining techniques, such as neuro-fuzzy (Lee et al. 2006; Vahidnia et al. 2010; Oh and Pradhan 2011; Sezer et al. 2011; Bui et al. 2011), fuzzy logic (Ercanoglu and Gokceoglu 2002; Kanungo et al. 2006, 2008; Lee 2007b; Muthu et al. 2008; Pradhan and Lee 2009; Pradhan 2010b, c; Pradhan 2011a, b; Pourghasemi et al. 2012c), artificial neural networks (Kanungo et al. 2006; Melchiorre et al. 2008; Chen et al. 2009; Pradhan and Lee 2009, 2010a, b; Pradhan et al. 2010a, b; Pradhan and Buchroithner 2010; Pradhan and Pirasteh 2010; Poudyal et al. 2010; Yilmaz 2009a, b; Yilmaz 2010a, b; Pradhan 2011c; Zare et al. 2012; Bui et al. 2012a), support vector machine (Yao et al. 2008; Yilmaz 2010a, b; Pourghasemi et al. 2012f; Pradhan 2012; Bui et al. 2012b), decision tree methods (Saito et al. 2009; Yeon et al. 2010; Nefeslioglu et al. 2010a, b; Pourghasemi et al. 2012e; Pradhan 2012), Shannon’s entropy (Devkota et al. 2012; Pourghasemi et al. 2012d, g), general additive model (Xiao et al. 2010), evidential belief function (Althuwaynee et al. 2012; Bui et al. 2012c), and spatial decision support system (Wan 2009). Understanding the differences between the proposed approaches is not always simple. The main differences are the rigor of the approach (Chung and Fabbri 1998) and the method used to estimate the prior probability of landslide occurrence.

There are many literatures available related to the landslide studies from Nepal Himalaya (Deoja et al. 1991; Dhital et al. 1991; Dangol et al. 1993; Dangol 2000; Pantha et al. 2008; Regmi et al. 2012a, b). These studies basically deal with the loss of life and property, physical properties of the landslides, and effect of geological setting in the formation of these landslides, along with recommendations for preventive measures and landslide susceptibility mapping using the geographic information system (GIS). The main purpose of this paper is to develop landslide susceptibility maps of the Mugling–Narayanghat road corridor and its surrounding region using frequency ratio (FR), statistical index (SI), and WoE models in order to find the better model that is more accurate in landslide susceptibility mapping in the study area.

The study area

The Mugling–Narayanghat Highway is 36 km long and is located in a mountain range between Narayanghat and Mugling of Chitwan Districts, Narayani Zone, Central Nepal. The study area lies within longitude 84°25′30″ E to 84°35′30″ E and latitude 27°52′ N to 27°46′ N (Fig. 1) that falls within the topographical map 2784-03C (Mugling) and 2784-02D (Jugedi Bajar). It covers an area of about 138.14 km2. The minimum and maximum altitudes of the area vary between 200 m at Jugedi Bajar and 1,540 m to the north of Bespani.

Subtropical to temperate climate prevail in the study area, with the winter temperatures ranging in between 6 and 25 °C, while that of summer ranging from 25 to 40 °C. The monthly maximum temperature and daily rainfall records from the nearby Bharatpur station during 2002–2006 have given the highest maximum temperature of 41.2 °C in May 2004 and a mean annual rainfall of 2,650 mm (DWIDP 2009). The months of April, May, and June are the hottest, with average maximum temperatures of 37.8, 39.3, and 38.6 °C, respectively. As in other parts of Nepal, summer monsoon is dominant from June to September’s end. The distribution of the drainage network of the study area is shown in Fig. 1. Trishuli River is the main river system in the study area. From chainage 10 (Ch10) to chainage14 (Ch14), the Trishuli River runs toward east to southeast and flows in the Siwaliks rocks. Within this section, there are three major tributaries of the river, i.e., Das Khola at Ch12 + 600, Khahare Khola at Ch11 + 300, and Jugedi Khola at Ch10 + 300 (Fig. 1). Ch14 to Ch36 of the highway is located in the Lesser Himalayan Zone. Along this section, the axis of the Trishuli River winds north to south or east to west. At Ch25 + 500, the Trishuli River meets the Rigdi Khola, the largest tributary of the Trishuli River within the study area (Fig. 1).

Geological setting

Geologically and tectonically, the Nepal Himalaya is divided into five major tectonic zones, namely, Terai, Sub-Himalaya (Siwaliks), Lesser Himalaya, Higher Himalaya, and Tibetan-Tethys Himalaya (Ganser 1964; Upreti 1999).The Tibetan-Tethys Himalayan zones are characterized by the fossiliferous sedimentary rocks, while the Higher Himalayan zones are dominated by the high-grade metamorphic rocks like gneiss, schist, quartzite, marble, and some granite. Limestone, marble, dolomite, sandstone, shale, slate, phyllite, quartzite, and schist are the dominant rock types in the Lesser Himalayan zone. The Sub-Himalayan zone is characterized by conglomerate, sandstone, mudstone, siltstone, and quaternary deposits.

The Mugling–Narayanghat road and its surrounding region belong to the Precambrian Lesser Himalayan rocks of Nawakot Complex (Stöcklin and Bhattarai 1978; Stöcklin 1980), Miocene Siwaliks, and Quaternary terrace deposits (Fig. 2). The Nawakot Complex is divided into the Lower Nawakot Group and Upper Nawakot Group, and along this road section, the rocks from both groups are observed. Kuncha Formation, Fagfog Quartzite, Dandagaun Phyllite, Nourpul Formation, and Dhading Dolomite are from the Lower Nawakot Group, while Benighat Slate Formation is the only formation exposed from the Upper Nawakot Group. The Siwalik Group in the study area consists of Lower Siwalik and Middle Siwalik. Quaternary deposits consist of river terraces of different ages. The main rock types are mudstone, sandstone, limestone, dolomite, slate, phyllite, quartzite, and amphibolite (Table 1).
Fig. 2

Geological map of the study area

Table 1

Description of geological units of the study area

 

Formation

Geological age

Rock type

 

Quaternary Deposit

Recent

Mainly river terraces

Siwaliks

Middle Siwaliks

Mesozoic

Coarse-grained, salt- and pepper-like massive sandstone

Lower Siwaliks

Mesozoic

Variegated mudstone, with some thick-bedded, light gray, fine-grained sandstones

 

MBT

   

Nawakot complex

Upper Nawakot group

Benighat Slate Formation

Palaeozoic

Dark bluish gray to nearly black, soft-weathering slates and phyllites; many are argillaceous, and subordinately siliceous or finely quartzitic

STF (Simaltal Thrust Fault)

Lower Nawakot group

Dhading Dolomite

Late Precambrian

Dolomite consisting of fine crystalline or dense and light blue–gray in color. It is thinly bedded and platy in the basal part and the beds are thick to massive with common occurrence of columnar stromatolite in some parts. Frequent intercalation of black slate

Nourpul Formation

Late Precambrian

Predominantly phyllitic, but contains a variable amount of quartzitic and calcareous intercalations and dolomites and dolomitic quartzites

Purebesi Quartzite

Late Precambrian

Purple quartzite

Amphibolits

 

Thin beds of greenish amphibolites

Dandagaun Phyllites Jhikhu Carbonate

Late Precambrian

Uniform argillaceous to finely quartzitic phyllites of dark blue–green color

Fagfog Quartzite

Late Precambrian

White quartzite made up of colloidal fine-grained chert to impure coarse orthoquartzite, with occasional reddish to pale orange tints, with some intercalation of phyllite. Ripple marks are present in quartzite

Kuncha Formation

Late Precambrian

Alternation of phyllites, phyllitic quartzites and phyllitic gritstones resembling graywackes

Main boundary thrust (MBT) is the main geological structure in the study area. It separates the Lesser Himalayan rocks from the overlying Siwaliks (Fig. 2). Beside the MBT, other faults in the study area are Jugadi Thrust (JT), Kamalpur Thrust (KT), Simaltal Thrust (ST), and Virkuna Thrust (VT) (Fig. 2) (Devkota et al. 2012). All these thrusts are trending from east to west. MBT crosses the highway near Ch14 of the highway. JT is located to the south of the MBT, and it crosses the highway at around Ch10 and follows the Jugadi Khola. To the north of the MBT, KT, ST, and VT lie and pass through Ch16 + 700, Ch23 + 00, and Ch27 + 600 of the highway, respectively (Fig. 2) (Devkota et al. 2012). Several small-scale normal faults are distributed throughout the study area (Fig. 2). The rocks of the study area are highly deformed and display numerous synclines, anticlines, and sometimes continuous folding. A major syncline (Jalbire Sinclonorium) passes through Jalbire (Fig. 2). Jalbire Syncline is an open, noncylindrical, and slightly plunging fold, with the general trend and plunge of the fold being 112° and 7°, respectively. In the study area, most of the instabilities are observed along the Das Khola, Khahre Khola, and Jugedi Khola valleys and around MBT and immediately north of it. Geomorphologically, the Mugling–Narayanghat road section is within the Mahabharat Range (Hagen 1969; Upreti 1999).

Input data preparation

For landslide susceptibility analysis, the main steps are data collection and construction of a spatial database from which the relevant factors are extracted, followed by assessment of landslide susceptibility using the relationship between landslide and landslide-related factors and validation of the results (Guzzetti et al. 1999; Ercanoglu and Gokceoglu 2004). It is important to construct a digitized database for making the landslide susceptible map using GIS. Data preparation involved the digitization or creation of a GIS database, which includes topographical data, geological data, land use data, etc. For landslide susceptibility assessment, the number of attributes used for rating varies widely. In this study, 11 factors were selected in the first stage: slope gradient, slope aspect, plan curvature, altitude, stream power index (SPI), topographic wetness index (TWI), lithology, land use, distance from faults, distance from rivers, and distance from highway. Topographic maps and aerial photographs provided by the Department of Survey, Government of Nepal were used as the basic data sources for generating these layers. A landslide distribution map was prepared using earlier reports, aerial photographs, and field surveys. These data sources were used to generate various thematic layers using ArcGIS 9.3 and ArcView 3.3 softwares. Brief descriptions of the procedure for the preparation of each data layer are given in the succeeding subsections.

Landslide inventory

The first step in landslide susceptibility assessments is to acquire information about the landslides that have occurred in the past. This stage is considered as the fundamental part of the landslide hazard studies (Guzzetti et al. 1999; Ercanoglu and Gokceoglu 2004). Since landslide occurrences in the past and present are keys to future spatial prediction (Guzzetti et al. 1999), a landslide inventory map is a prerequisite for such a study. A landslide inventory map provides the basic information for evaluating landslide hazards or risk on a regional scale. Accurate detection of the location of landslides is most important for probabilistic landslide susceptibility analysis. For landslide inventory mapping, both the desk study and field study were performed. First of all, the aerial photographs of the area were analyzed and the landslide inventory map was created, and this was rechecked during the field surveys. From this, more than 438 slides were identified and mapped (Fig. 1). From these, 295 (67 %) landslides were randomly selected as training data and the remaining 143 (33 %) were kept for validation purposes. Almost every type of landslides are observed in the study area; however, rotational types are the dominant ones. Rock falls, debris flow, and topples are also observed along the highway.

Geomorphological-related factors

Geomorphological parameters used in this study are slope gradient, slope aspect, plan curvature, altitude, TWI, and SPI. The topographic map was used to prepare the digital elevation model (DEM) of the study area with 20 × 20 m pixel size. Using this DEM and the ArcGIS 9.3 software, the previously mentioned geomorphological thematic data layers were generated.

Slope gradient

Slope gradient is one of the most important causes of slope instability (Ayalew et al. 2004; Guzzetti et al. 1999; Kolat et al. 2006; Ohlmacher and Davis 2003; Oyagi 1984). The moisture content and pore pressure could be influenced at local scales, whereas the regional hydraulic behavior could be controlled by slope angle patterns at larger scales (Mancini et al. 2010). The slope in the study area varies from 0° to 73.72°. This map was produced automatically in ArcGIS using the DEM with 20 × 20 m grid size and is divided into five categories (Fig. 3a).
Fig. 3

Various thematic maps used for the present study. a Slope (in degrees), b slope aspect, c plan curvature, d altitude (in meters), e SPI, f TWI, g land use, h distance from faults, i distance from rivers, and j distance from highway

Slope aspect

The slope aspect map was also produced from the DEM and it is defined as the direction of the maximum slope of the terrain surface. The slope aspect is related to the physiographic trends and the main precipitation direction (Ercanoglu and Gokceoglu 2002). In the present study, the aspect direction is divided into ten classes (Fig. 3b).

Plan curvature

Curvature analysis allows dividing the area into concave, convex, and flat surfaces and, consequently, may help to identify zones that exhibit proneness to landslide (Mancini et al. 2010). The curvature was derived from the DEM in ArcGIS. It is reclassified into three classes: concave, flat, and convex (Fig. 3c).

Altitude

The altitude map was produced using the DEM with 20 × 20 m grid size. In the present study area, the altitude ranges from 200 to 1,540 m and is reclassified into seven classes with 200 m interval (Fig. 3c).

Stream power index

SPI measures the erosion power of the stream and is also considered as a factor contributing toward stability within the study area. The SPI can be defined as (Moore and Grayson 1991):
$$ \mathrm{SPI}={A_{\mathrm{s}}}\tan \beta $$
(1)
where A s is the specific catchment area and β is the local slope gradient measured in degrees. In the present study, SPI is divided into four classes (Fig. 3e).

Topographic wetness index

Another important topographic factor within the runoff model is the TWI (Beven and Kirkby 1979) defined as:
$$ \mathrm{TWI}=\ln \left( {\frac{a}{{\tan \beta }}} \right) $$
(2)
where a is the cumulative upslope area draining through a point (per unit contour length) and tan β is the slope angle at the point. The \( \ln \left( {\frac{a}{{\tan \beta }}} \right) \) index reflects the tendency of water to accumulate at any point in the catchment (in terms of a) and the tendency of gravitational forces to move that water down slope (expressed in terms of tan β as an approximate hydraulic gradient) (Poudyal et al. 2010). In the present study, TWI is divided into three classes (Fig. 3f).

Lithology

Lithology plays an important role in landslide susceptibility studies because different geological units have different susceptibilities to active geomorphological processes of the Himalaya (Pradhan et al. 2006). Sandstone, mudstone, limestone, dolomite, slate, phyllite, amphibolites, quartzite, and quaternary deposits cover the study area. The lithological map was produced by the help of previous geological map and from detailed field investigation (Fig. 2).

Land use map

Land use also plays a significant role in the stability of the slope. The land covered by forest regulates continuous water flow and water infiltrates regularly, whereas the cultivated land affects the slope stability owing to saturation of covered soil (Devkota et al. 2012). The land use map was provided by the Survey Department of Nepal (Fig. 3g). It consists of nine different land use types; they are cutting, cultivation, forest, orchard, grass, bush, barren, sand, and river. Bush (39 %) covers the highest amount of area followed by cultivation land (33 %) and forest (21 %) (Fig. 3g).

Distance from faults

The spatial distribution of landslides in the study area shows a strong correlation with the tectonic fractures as faults. The landslides occurred mainly along the faults and decreased sharply with distance from it. A buffer map related to the fault with a 100-m buffer zone interval is shown in Fig. 3h. The fault lines were derived from the geological map in 1:100,000 scale.

Distance from rivers

Runoff plays an important role as a triggering factor for landslides. The distance from rivers is represented by the proximity of the rivers in the area. The river map obtained from the Department of Survey of Nepal was used in the ArcGIS 9.3 software to get the proximity to rivers (Fig. 3i).

Distance from highway

Landslides are very common along road cuts. This is mainly due to the fact that the natural condition of the slope is damaged during the process of road construction. Also, the road cut exposes the joints and fractures that make the slope unstable. Road cuts are usually sites of anthropological instability. Also, a given road segment may act as a barrier, a net source, a net sink, or a corridor for water flow, and depending on its location in the area, it usually serves as a source of landslides (Pradhan 2010a, b, c). The Mugling–Narayanghat Highway is buffered with a buffer distance of 100 m in ArcGIS 9.3 to generate the distance in the highway map (Fig. 3j).

Methods

Although there are several methods in the statistical approach of landslide susceptibility mapping, in the present study, we have adopted three approaches: FR, SI, and WoE for the landslide susceptibility assessment of Mugling–Narayanghat road corridor and its surrounding regions. Details of each approach are described in the succeeding subsections.

Frequency ratio model

Among several bivariate statistical methods for landslide susceptibility mapping, the FR model has been adopted for the present study (Pradhan and Lee 2009). The FR model is a simple and understandable probabilistic model, in which the FR is defined as the ratio of the area where landslides occurred in the total study area and is also the ratio of the probabilities of a landslide occurrence to a nonoccurrence for a given attribute (Bonham-Carter 1994; Pradhan and Lee 2009). It can be expressed by the given formula equation:
$$ \mathrm{LSI}=\sum {\mathrm{FR}} $$
(3)
where LSI is the landslide susceptibility index and FR is the frequency ratio. FR is expressed as:
$$ \mathrm{FR}=\frac{{\frac{{{N_{\mathrm{pix}}}\left( {S{X_i}} \right)}}{{\sum {_{i=1}^mS{X_i}} }}}}{{\frac{{{N_{\mathrm{pix}}}\left( {{X_j}} \right)}}{{\sum {_{j=1}^n{N_{\mathrm{pix}}}\left( {{X_j}} \right)} }}}} $$
(4)
where N pix(SX i ) is the number of pixels with landslides within class i of parameter variable X, N pix(X j ) is the number of pixels within parameter variable X j , m is the number of classes in the parameter variable X i , and n is the number of parameters in the study area. The FR of all the thematic layers used in the present study was calculated in Microsoft Excel and is given in Table 2. Using these FR values, the thematic maps were reclassified by the help of the spatial analyst tool in ArcGIS 9.3. These reclassified layers were then added using the raster calculator of ArcGIS to get the final susceptibility map.
Table 2

Spatial relationship between each factor and landslide by the FR, SI, and WoE models in Nepal

Factor

Class

(a)

(b)

FR (b/a)

SI (W +)

W

C

S 2(W +)

S 2(W )

S(C)

C/S(C)

Slope degree

0–15°

12.24

3.73

0.30

−1.19

0.09

−1.28

0.09

0.00

0.31

−4.17

15–25°

13.6

12.2

0.90

−0.11

0.02

−0.12

0.03

0.00

0.18

−0.70

25–35°

24.59

31.86

1.30

0.26

−0.10

0.36

0.01

0.00

0.12

2.89

35–45°

28.42

33.56

1.18

0.17

−0.07

0.24

0.01

0.01

0.12

1.95

>45°

13.13

18.64

1.42

0.35

−0.07

0.42

0.02

0.00

0.15

2.79

Slope aspect

Flat

2.96

0

0.00

0.00

0.03

0.00

0.00

0.00

0.00

0.00

North

6.78

4.07

0.60

−0.51

0.03

−0.54

0.08

0.00

0.29

−1.83

Northeast

10.82

11.19

1.03

0.03

0.00

0.04

0.03

0.00

0.18

0.20

East

8.68

10.51

1.21

0.19

−0.02

0.21

0.03

0.00

0.19

1.11

Southeast

10.75

19.32

1.80

0.59

−0.10

0.69

0.02

0.00

0.15

4.66

South

12.59

15.59

1.24

0.21

−0.03

0.25

0.02

0.00

0.16

1.55

Southwest

12.43

12.2

0.98

−0.02

0.00

−0.02

0.03

0.00

0.18

−0.12

West

10.29

9.83

0.96

−0.05

0.01

−0.05

0.03

0.00

0.20

−0.26

Northwest

10.52

10.51

1.00

0.00

0.00

0.00

0.03

0.00

0.19

−0.01

North

6.15

6.78

1.10

0.10

−0.01

0.10

0.05

0.00

0.23

0.45

Altitude (m)

<400

25.55

23.39

0.92

−0.09

0.03

−0.12

0.01

0.00

0.14

−0.85

400–600

24.06

33.9

1.41

0.34

−0.14

0.48

0.01

0.01

0.12

3.92

600–800

19.48

21.36

1.10

0.09

−0.02

0.12

0.02

0.00

0.14

0.81

800–1,000

14.16

13.56

0.96

−0.04

0.01

−0.05

0.03

0.00

0.17

−0.29

1,000–1,200

6.68

7.12

1.07

0.06

0.00

0.07

0.05

0.00

0.23

0.30

1,200–1,400

1.87

0.68

0.36

−1.01

0.01

−1.03

0.50

0.00

0.71

−1.44

>1,400

0.18

0

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Lithology

Quaternary

3.32

0

0.00

0.00

0.03

0.00

0.00

0.00

0.00

0.00

Terrace

9.58

13.56

1.42

0.35

−0.05

0.39

0.03

0.00

0.17

2.31

Middle Siwaliks

0.02

0

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Lower Siwaliks

1.01

2.37

2.34

0.85

−0.01

0.87

0.14

0.00

0.38

2.26

Benighat Slate

11.48

12.88

1.12

0.12

−0.02

0.13

0.03

0.00

0.17

0.76

Dhading Dolomite

10.06

12.2

1.21

0.19

−0.02

0.22

0.03

0.00

0.18

1.22

Nurpul Formation

41.65

47.12

1.13

0.12

−0.10

0.22

0.01

0.01

0.12

1.90

Purebesi Quartzite

4

3.05

0.76

−0.27

0.01

−0.28

0.11

0.00

0.34

−0.83

Amphibolite

0.25

0.34

1.35

0.30

0.00

0.30

1.00

0.00

1.00

0.30

Dandagau Formation

1.68

1.02

0.61

−0.50

0.01

−0.51

0.33

0.00

0.58

−0.87

Fagfog Quartzite

2.03

0.68

0.33

−1.09

0.01

−1.11

0.50

0.00

0.71

−1.56

Kuncha Formation

6.87

6.78

0.99

−0.01

0.00

−0.01

0.05

0.00

0.23

−0.06

Distance from rivers (m)

0–50

31.39

32.2

1.03

0.03

−0.01

0.04

0.01

0.01

0.12

0.30

50–100

38.76

45.08

1.16

0.15

−0.11

0.26

0.01

0.01

0.12

2.22

100–150

9.29

10.17

1.09

0.09

−0.01

0.10

0.03

0.00

0.19

0.52

150–200

5.05

4.41

0.87

−0.14

0.01

−0.14

0.08

0.00

0.28

−0.51

>200

7.48

8.14

1.09

0.08

−0.01

0.09

0.04

0.00

0.21

0.43

Distance from highway (m)

0–100

9.84

5.08

0.52

−0.66

0.05

−0.71

0.07

0.00

0.27

−2.69

100–200

9.56

14.58

1.53

0.42

−0.06

0.48

0.02

0.00

0.16

2.90

200–300

9.16

13.22

1.44

0.37

−0.05

0.41

0.03

0.00

0.17

2.40

300–400

16.75

17.97

1.07

0.07

−0.01

0.08

0.02

0.00

0.15

0.56

>400

46.66

49.15

1.05

0.05

−0.05

0.10

0.01

0.01

0.12

0.86

Distance from faults (m)

0–100

9.52

11.19

1.18

0.16

33.00

0.11

11.19

0.16

0.16

−0.02

100–200

8.75

10.51

1.20

0.18

31.00

0.11

10.51

0.18

0.18

−0.02

200–300

8.13

13.22

1.63

0.49

39.00

0.13

13.22

0.49

0.49

−0.06

300–400

7.43

10.17

1.37

0.31

30.00

0.10

10.17

0.31

0.31

−0.03

>400

58.14

54.92

0.94

−0.06

162.00

0.55

54.92

−0.06

−0.06

0.07

TWI

<5.5

68.75

87.8

1.28

0.24

−0.94

1.19

0.00

0.03

0.18

6.66

5.5–7.2

20.57

12.2

0.59

−0.52

0.10

−0.62

0.03

0.00

0.18

−3.50

>7.2

2.26

0

0.00

0.00

0.02

0.00

0.00

0.00

0.00

0.00

SPI

0–150

23.05

23.39

1.01

0.01

0.00

0.02

0.01

0.00

0.14

0.14

150–300

30.14

35.25

1.17

0.16

−0.08

0.23

0.01

0.01

0.12

1.91

300–450

26.06

31.53

1.21

0.19

−0.08

0.27

0.01

0.00

0.13

2.13

>450

12.72

9.83

0.77

−0.26

0.03

−0.29

0.03

0.00

0.20

−1.48

Land use

Barren

0.45

0

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

Bush

37.43

36.95

0.99

−0.01

0.01

−0.02

0.01

0.01

0.12

−0.17

Cultivation

30.34

23.73

0.78

−0.25

0.09

−0.34

0.01

0.00

0.14

−2.46

Cutting

0.36

1.36

3.75

1.32

−0.01

1.33

0.25

0.00

0.50

2.64

Forest

18.65

37.97

2.04

0.71

−0.27

0.98

0.01

0.01

0.12

8.19

Grass

1.17

0

0.00

0.00

0.01

0.00

0.00

0.00

0.00

0.00

Orchard

0.01

0

0.00

0.00

0.00

0.00

0.00

0.00

0.00

0.00

River

2.22

0

0.00

0.00

0.02

0.00

0.00

0.00

0.00

0.00

Sand

1.35

0

0.00

0.00

0.01

0.00

0.00

0.00

0.00

0.00

Plan curvature (100/m)

Concave

42.5

42.03

0.99

−0.01

0.01

−0.02

0.01

0.01

0.12

−0.16

Flat

4.17

0

0.00

0.00

0.03

0.33

0.00

0.58

−2.49

−0.16

Convex

45.3

57.29

1.26

0.23

−0.24

0.01

0.01

0.12

3.98

−0.16

Statistical index model

The second statistical approach used in this study is SI (W i ) model. It was introduced by van Westen et al. (1997) for landslide susceptibility analyses and has been adopted by various researchers. In the SI method, a weight value for a parameter class is defined as the natural logarithm of the landslide density in the class divided by the landslide density in the entire map. This method is based upon the following formula (van Westen et al. 1997):
$$ {W_{ij }}=\ln \left( {\frac{{{f_{ij }}}}{f}} \right)=\ln \left( {\frac{{A_{ij}^{*}}}{{{A_{ij }}}}} \right)\times \ln \left( {\frac{A}{{{A^{*}}}}} \right) $$
(5)
where W ij is the weight given to a certain class i of parameter j, f ij is the landslide density within the class i of parameter j, f is the landslide density within the entire map, \( A_{ij}^{*} \) is the area of landslides in a certain class i of parameter j, A ij is the area of a certain class i of parameter j, A * is the total area of landslides in the entire map, and A is the total area of the entire map. To perform the SI modeling, each thematic map is overlaid and crossed separately with the landslides location map (estimation group) using the ArcGIS 9.3 software, and the numbers of pixels forming the landslides that fall into the various classes of the maps of the different factors are calculated. The calculated numbers of pixels are divided by the total number of pixels for that selected parameter to calculate the density of the class. For this, all the cross table obtained in GIS are exported to Microsoft Excel, where all the data required to calculate the weight (W ij ) are obtained. The weight is positive when the landslide density is higher than the normal and is negative when it is less than normal. The final weight obtained from the analysis is analyzed using the weighted sum option in the spatial analyst tools of ArcGIS and the final landslide susceptibility map was constructed.

Weights-of-evidence

WoE is a quantitative “data-driven” method used to combine datasets. It uses the Bayesian probability model. In geology, it was originally developed for mineral potential assessment (Bonham-Carter et al. 1988, 1989; Agterberg 1992; Agterberg et al. 1993; Bonham-Carter 2002; Carranza and Hale 2002; Carranza and Castro 2006). Several authors have applied the method to mineral potential mapping using GIS (Carranza and Hale 2000). Cheng (2004) used the method to predict the location of flowing wells and Daneshfar and Benn (2002) used the WoE model to analyze spatial associations between faults and seismicity, whereas Carranza and Castro (2006) used this model for predicting lahar inundation zones. Recently, several researchers have used the WoE model in landslide susceptibility mapping (van Westen et al. 2003; Mathew et al. 2007; Neuhäuser and Terhorst 2007; Bui et al. 2008; Sharma and Kumar 2008; Dahal et al. 2008a, b; Regmi et al. 2010a, b; Pourghasemi et al. 2012a, Mohammady et al. 2012).

A detailed description of the mathematical formulation of the method is available in Bonham-Carter (1991, 1994, 2002) and the mathematical relationship for landslide susceptibility mapping is available in many research articles (Pradhan 2010a, b, c; Dahal et al. 2008b). They have all experimented with Bayes’ conditional probability theorem. Bayes’ probability theorem can be written as:
$$ P\left\{ {\frac{A}{B}} \right\}=\frac{{\left\{ {P\left( {\frac{B}{A}} \right)\times P(A)} \right\}}}{P(B) } $$
(6)
By overlaying landslide locations with each evidence (conditioning factors), the statistical relationship can be measured between them and assessed as to whether and how significant the evidence is responsible for the occurrence of past landslides (Neuhäuser and Terhorst 2007). On the other hand, the WoE model is fundamentally based on the calculation of positive and negative weights W + and W . The method calculates the weight for each landslide predictive factor (A) based on the presence or absence of the landslides (B) within the area (Bonham-Carter 1994) as follows:
$$ W_i^{+}=\ln \frac{{P\left\{ {B\left| A \right.} \right\}}}{{P\left\{ {B\left| A \right.} \right\}}} $$
(7)
$$ W_i^{-}=\ln \frac{{P\left\{ {\overline{B}\left| A \right.} \right\}}}{{P\left\{ {\overline{B}\left| \overline{A} \right.} \right\}}} $$
(8)
where P is the probability and ln is the natural log. Similarly, B is the presence of potential landslide predictive factor, \( \overline{B} \) is the absence of a potential landslide predictive factor, A is the presence of landslide, and \( \overline{A} \) is the absence of a landslide. A positive weight (W +) indicates that the predictable variable is present at the landslide locations and the magnitude of this weight is an indication of the positive correlation between the presence of the predictable variable and the landslides. A negative weight (W ) indicates the absence of the predictable variable and shows the level of negative correlation (Dahal et al. 2008a). In landslide susceptibility mapping, the weight contrast, C(C = W + − W ), measures and reflects the spatial association between the evidence feature and landslide occurrence. C is positive for a positive spatial association and negative for a negative spatial association (Dahal et al. 2008b).
The standard deviation of W is calculated as:
$$ S(C)=\sqrt{{{S^2}{W^{+}}+{S^2}{W^{-}}}} $$
(9)
where S(W +) is the variance of the positive weights and S(W ) is the variance of the negative weights. The variances of the weights can be calculated by the following expressions:
$$ {S^2}{W^{+}}=\frac{1}{{N\left\{ {B\cap A} \right\}}}+\frac{1}{{B\cap \overline{A}}} $$
(10)
$$ {S^2}{W^{-}}=\frac{1}{{\left\{ {\overline{B}\cap A} \right\}}}+\frac{1}{{\left\{ {\overline{B}\cap \overline{A}} \right\}}} $$
(11)

The studentized contrast is a measure of confidence and is defined as the ratio of the contrast divided by its standard deviation. The studentized contrast serves as an informal test that C is significantly different from zero or if the contrast is likely to be “real” (Bonham-Carter 1994).

Results and discussion

Application of frequency ratio model

To determine the level of correlation between the landslide locations and the conditioning factors, i.e., slope degree, slope aspect, altitude, lithology, land use, distance from rivers, distance from highway, distance from faults, TWI, SPI, land use, and plan curvature, the FR method was used. The final landslide susceptibility map obtained by the FR model is shown in Fig. 4. From Table 2, it is seen that landslides are higher in slopes >25°. The FR value is highest for slope class ≥45°, followed by the 25–35° slope class. As the slope angle increases, the shear stress in the soil or other unconsolidated material generally increases. Gentle slopes are expected to have a low frequency of landslides because of the generally lower shear stresses that are associated with low gradients. The steep slopes may not be susceptible to landsliding. The FR from the slope aspect analysis shows that the SE-facing slopes are most affected by landslides, followed by E-, NE-, and S-facing slopes. In the case of altitude, landslide density is highest at the elevation ranging from 400 to 600 m followed by from 600 to 800 m. The FR of the altitude (Table 2) shows that it is >1 at these altitude ranges. For the lithology, it can be seen that Lower Siwaliks, Dhading Dolomite, Dandagau Formation, Nurpul Formation, and Amphibolites have FR values >1 (Table 2). The influence of a drainage system upon the landslide susceptibility was also analyzed by identifying the drainage river line by buffering. The distance in between 0 and 150 m from the drainage river shows higher correlation with the landslides. The highway also has a great influence in landslide formation in the study area. The distance >100 m shows an FR value >1. This value decreases as the distance from the highway decreases. The relation between a landslide and its distance to a drainage line and road shows that, when distance from the drainage (river) line and road increases, the landslide occurrence probability decreases. For that reason, road construction and bank erosion are the most important factors in slope imbalance causing frequent occurrence of landslides.
Fig. 4

Landslide susceptibility map based on the FR model

The distance from the fault map shows that the landslides are highly dependent on these tectonic fractures as the FR value is >1 in between 0 and 400 m from the fault. The FR value decreases with increasing fault distance. The relation between TWI and landslide probabilities showed that the 0–5 classes have the highest FR value and that 300–450 SPI shows high FR value. In the case of land use, higher FR values were for cutting and forest-covered lands. The plan curvature value represents the morphology of the area. The convex curvature has the highest FR value.

Application of statistical index model

The resultant weights for each thematic map for the SI model are given in Table 2. These weights were then analyzed by using the weighted sum option in the spatial analyst tools of ArcGIS to get the final LSI map (Fig. 5). From Table 2 it is seen that, among the W i for various conditioning factors, slope classes >25° have positive weight values. Thus, it is clear that the slope becomes unstable as the slope becomes steeper. In the case of slope aspect, N, NE, E, SE, and S have positive W i values. Among all these, SE has the highest value, followed by S, E, N, and NE. The W i values of altitude are positive within the range of 400 to 800 m, and 1,000 to 1,200 m with the highest value at 400 to 600 m. It slowly decreases with both decreasing and increasing altitudes. Among the W i calculated for the lithological complexes, Lower Siwaliks has the highest W i value, followed by terrace deposits, Amphibolites, Dhading Dolomite, Nourpul Formation, and Benighat Slate Formation. The W i value for the distance from rivers (Drainages) shows that it decreases with the increase in distance from the drainage. From Table 2, it is seen that the distance between 0 and 150 m from the drainage have higher W i value. The W i value of distance from highway shows that it is positive for distance >100 m. It decreases as the distance increases. This shows that the highway constructions have a positive influence in landslide occurrence. The W i value of distance from faults shows that the distance in between 0 and 400 m have positive W i values, with the highest value occurring at the distance in between 200 and 300 m followed by 300 to 400 m. In the case of TWI and SPI, the 0–5 and 0–450 ranges, respectively, have positive W i value, showing good correlation with the landslides. In the case of land use, only the cutting and forest-covered lands show good correlation with the landslide formation as these land use system have positive W i values. Convex slope only consists of positive W i value.
Fig. 5

Landslide susceptibility map derived from the SI model

Application of weights-of-evidence model

The WoE modeling is mainly based on the W + and W . From these, the weight contrast C is calculated. C is positive for a positive spatial association and negative for a negative spatial association. The studentized value of C, the ratio of C to standard deviation or C/S(C), serves as a guide to the significance of the spatial association and acts as a measure of the relative certainty of the posterior probability (Bonham-Carter 1991). The weights and contrasts for each predictor pattern are summarized in Table 2. S 2 W + and S 2 W are the variances of W + and W . S(C) is the standard deviation of the contrast and C/S(C) is the studentized value of the contrast (Table 2). The detailed procedure of calculating C, S(C), S 2 W +, and S 2 W is given in the previous section. The relationship between the landslides and the landslide conditioning factors, contrast, and studentized C are presented in Table 2. The C/S(C) derived based on the WoE were assigned to the classes of each thematic layers to produce multiclass weighted maps for all evidence, which were overlaid and numerically added according to Eq. 12 in order to calculate a LSI map (Fig. 6):
$$ \mathrm{LS}{{\mathrm{M}}_{\mathrm{WoE}}}=\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Slope}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Aspect}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Curvature}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Altitude}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Lithology}}}+\mathrm{Wo}{{\mathrm{E}}_{{\mathrm{Land}\;\mathrm{use}}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Fault}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{River}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{Highway}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{TWI}}}+\mathrm{Wo}{{\mathrm{E}}_{\mathrm{SPI}}} $$
(12)
Fig. 6

Landslide susceptibility map derived from the WoE model

Based on the value of C/S(C), it can be seen that the slope gradient >25° has a maximum value. Thus, the area occupied by the slope gradient >25° shows maximum susceptibility with reference to landslides in the study area. In the case of slope aspect, NE- to S-trending slope have higher C/S(C) value, indicating that landslide susceptibility is highest in these slopes. In the case of altitude, the range in between 400 and 600 m has highest C/S(C) value, indicating high landslide susceptibility at this range of elevation. Among the different lithological classes terrace and Lower Siwaliks have highest C/S(C) value, followed by variegated mudstone, with some thick-bedded, light gray, fine-grained sandstones have higher value for C/S(C). These lithological units show maximum susceptibility with reference to landslides in the study area. The distance from rivers parameter also has shown positive influence towards slope destabilization as seen from the C/S(C) value. Slope saturation might be the reason for this phenomenon. The C/S(C) value of distance from highway shows that the distances in between 50 and 150 m have the highest value. It decreases as the distance increases. This is mainly due to the unplanned road cutting. In the case of distance from fault, only the fault distance >400 m shows higher C/S(C) value. In the case of TWI, the highest C/S(C) value is for classes 0–5, and for SPI, it is highest at the range of 150–450. Forest-covered land, followed by cutting land, has a higher value of C/S(C), showing maximum susceptibility to landslides. The convex slope has the highest value of C/S(C), indicating high landslide susceptibility in convex slope.

Validation of the landslide susceptibility maps

The overall performance of the analysis is generally judged on the number of correctly classified cells, and so a validation process is required. There exist several methods for validation of the landslide susceptibility maps. In the present study, landslide susceptibility maps were verified by comparing the susceptibility map with both the training data that were used for building the models and with the landslide locations that were not used during the model building process.

The area under the curve (AUC) represents the success rate and prediction rate percentage of the model. Variation from 0.9 to 1.0 is the ideal situation. AUC was calculated from 100 subdivisions of the LSI values of all the cells in the study area and the cumulative percentage of landslide occurrences in the classes. The AUC was obtained for both the training and the validation data (Fig. 7). The result showed that all the three models exhibited similar performance, with the FR model being the better one (success rate, 76.8 %; prediction rate, 75.4 %), followed by the WoE model (success rate, 75.6 %; prediction rate, 74.9 %) and the SI model (success rate, 75.5 %; prediction rate, 74.6 %) (Fig. 7a, b).
Fig. 7

a Success rate of the landslide susceptibility maps. b Prediction rate of the landslide susceptibility maps

Conclusions

Landslides are a serious threat to life and damage property; thus, landslide susceptibility mapping can be one of the preliminary steps in minimizing these damages. In the current research, three statistically based models, such as FR, SI, and WoE, were used for identifying the areas susceptible to landslides at the Mugling–Narayanghat road corridor and its surrounding area of Central Nepal Himalaya. The results of each model were given.

Finally, the performances of these models were compared. Susceptibility mapping was performed using various topographic, geological, structural, land use, and other datasets in GIS. In order to prepare the susceptibility maps, 11 thematic layers (slope gradient, slope aspect, plan curvature, altitude, SPI, TWI, lithology, land use, distance from faults, distance from rivers, and distance from highway) were used as the input data. The landslide inventory map of the study area was created by the analysis of aerial photographs, remote sensing images, from earlier published/unpublished reports, and by repeated field surveys. Among 438 landslides, 295 (67 %) were used as training data and the remaining 143 (33 %) were used for validation purposes. The AUC curves (success rate and prediction rate) were prepared for all the three models to test their accuracy. The validation results showed that all the three models show equal performance, with the FR model being the better one (success rate, 76.8 %; prediction rate, 75.4 %), followed by the WoE model (success rate, 75.6 %; prediction rate, 74.9 %) and the SI model (success rate, 75.5 %; prediction rate, 74.6 %). From this, the authors conclude that all the three models show almost similar results. As most of these landslide susceptibility maps show good prediction capacity, they can be used for planning future development works by avoiding the highly susceptible zones during project implementation.

Notes

Acknowledgments

The authors express their gratitude to MEXT (Ministry of Education, Culture, Sports, Science and Technology, Tokyo, Japan) for funding the present study. Dr. Michel Calvelo, Mr. Pravin Kayasta, and Mr. Ananta Man Shingh are sincerely acknowledged for their great help during the process of writing this paper. Also, the authors would like to thank two anonymous reviewers for their helpful comments on the previous version of the manuscript

References

  1. Adhikari TL (2009) Disaster mitigation works along Narayanghat Mugling Highway. International seminar on hazard management for sustainable development Google Scholar
  2. Agterberg FP (1992) Combining indicator patterns in weights of evidence modeling for resource evaluation. Nat Res Res 1(1):39–50CrossRefGoogle Scholar
  3. Agterberg FP, Bonham-Carter GF, Cheng Q, Wright DF (1993) Weights of evidence modeling and weighted logistic regression for mineral potential mapping. In: Davis JC, Herzfeld UC (eds) Computers in geology, 25 years of progress. Oxford University Press, Oxford, pp 13–32Google Scholar
  4. Akgün A, Kincal C, Pradhan B (2011) Application of remote sensing data and GIS for landslide risk assessment as an environmental threat to Izmir city (West Turkey). Environ Monit Assess. doi: 10.1007/s10661-011-2352-8
  5. Aleotti P, Chowdhury R (1999) Landslide hazard assessment: summary review and new perspectives. Bull Eng Geol Environ 58:21–44CrossRefGoogle Scholar
  6. Alexander DE (2008) A brief survey of GIS in mass movement studies, with reflections on theory and methods. Geomorphology 94:261–267CrossRefGoogle Scholar
  7. Althuwaynee OF, Pradhan B, Lee S (2012) Application of an evidential belief function model in landslide susceptibility mapping. Comput Geosci 44:120–135. doi: 10.1016/j.cageo.2012.3 CrossRefGoogle Scholar
  8. Anbalagan R (1992) Landslide hazard evaluation and zonation mapping in mountainous terrain. Eng Geol 32:269–277Google Scholar
  9. Ayalew L, Yamagishi H, Ugawa N (2004) Landslide susceptibility mapping using GIS based weighted linear combination, the case in Tsugawa area of Agano River, Niigata Prefecture, Japan. Landslides 1(1):73–81CrossRefGoogle Scholar
  10. Beven KJ, Kirkby MJ (1979) A physically based, variable contributing area model of basin hydrology. Hydro Sci Bull 24:43–69Google Scholar
  11. Binaghi E, Luzi L, Madella P, Pergalani F, Rampini A (1998) Slope instability zonation: a comparison between certainty factor and fuzzy Dempster–Shafer approaches. Nat Hazards 17:77–97CrossRefGoogle Scholar
  12. Bonham-Carter GF, Agterberg FP, Wright DF (1988) Integration of geological datasets for gold exploration in Nova Scotia. Photogr Eng Remote Sens 54:1585–1592Google Scholar
  13. Bonham-Carter GF, Agterberg FP, Wright DF (1989) Weights of evidence modeling: a new approach to mapping mineral potential. Statistical Applications in the Earth Science, Geological Survey of Canada Paper 89-9:171–183Google Scholar
  14. Bonham-Carter GF (1991) Integration of geoscientific data using GIS. In: Goodchild MF, Rhind DW, Maguire DJ (eds) Geographic information systems: principle and applications. Longman, London, pp 171–184Google Scholar
  15. Bonham-Carter GF (1994) Geographic information systems for geoscientists: modeling with GIS. In: Bonham-Carter F (ed) Computer methods in the geosciences. Pergamon, Oxford, p 398Google Scholar
  16. Bonham-Carter GF (2002) Geographic information systems for geoscientist: modeling with GIS. Pergamon, New York, pp 302–334Google Scholar
  17. Bui HB, Nguyen Q, Nguyen VT (2008) GIS-based weight of evidence modeling for landslide susceptibility mapping at Jaechon area, Korea. International Symposium on Geoinformatics for Spatial Infrastructure Development in Earth and Allied Sciences, 4 ppGoogle Scholar
  18. Bui DT, Pradhan B, Lofman O, Revhaug I, Dick OB (2011) Landslide susceptibility mapping at Hoa Binh province (Vietnam) using an adaptive neuro-fuzzy inference system and GIS. Comput Geosci. doi: 10.1016/j.cageo.2011.10.031
  19. Bui DT, Pradhan B, Lofman O, Revhaug I, Dick OB (2012a) Landslide susceptibility assessment in the Hoa Binh province of Vietnam using artificial neural network. Geomorphology. doi: 10.1016/ j.geomorph.2012.04.023
  20. Bui DT, Pradhan B, Lofman O, Revhaug I (2012b) Landslide susceptibility assessment in Vietnam using support vector machines, decision tree and naïve Bayes models. Math Probl Eng 2012:1–26. doi: 10.1155/2012/974638 Google Scholar
  21. Bui DT, Pradhan B, Lofman O, Revhaug I, Dick OB (2012c) Spatial prediction of landslide hazards in Vietnam: a comparative assessment of the efficacy of evidential belief functions and fuzzy logic models. Catena 96:28–40CrossRefGoogle Scholar
  22. Chen CH, Ke CC, Wang CL (2009) A back-propagation network for the assessment of susceptibility to rock slope failure in the eastern portion of the Southern Cross-Island Highway in Taiwan. Environ Geol 57:723–733CrossRefGoogle Scholar
  23. Cheng K, Wei C, Chang S (2004) Locating landslides using multitemporal satellite images. Adv Space Res 33:296–301Google Scholar
  24. Chung CJ, Fabbri AG (1998) Three Bayesian prediction models for landslide hazard. In: Buccianti R, Potenza R, Nardi G (eds) Proceedings of International Association for Mathematical Geology 1998 Annual Meeting (IAMG 98), Ischia, Italy, 1998, pp 204–211Google Scholar
  25. Chung CJ, Fabbri AG (2003) Validation of spatial prediction models for landslide hazard mapping. Nat Hazards 30:451–472CrossRefGoogle Scholar
  26. Chung CJ, Fabbri A (2005) Systematic procedures of landslide hazard mapping for risk assessment using spatial prediction models. In: Glade T, Anderson MG, Crozier MJ (eds) Landslide hazard and risk. Wiley, New York, pp 139–177Google Scholar
  27. Carranza EJM, Castro O (2006) Predicting lahar-inundation zones: case study in West Mount Pinatubo, Philippines. Nat Hazards 37:331–372CrossRefGoogle Scholar
  28. Carranza EJM, Hale M (2001) Geologically constrained fuzzy mapping of gold mineralization potential, Baguio District, Philippines. Nat Resour Res 10:125–136CrossRefGoogle Scholar
  29. Carranza EJM, Hale M (2002) Evidential belief functions for data-driven geologically constrained mapping of gold potential, Baguio district, Philippines. Ore Geol R 22:117–132CrossRefGoogle Scholar
  30. Carranza EJM, Hale M (2000) Geologically constrained probabilistic mapping of gold potential, Baguio district, Philippines. Nat Res Res 9:237–253CrossRefGoogle Scholar
  31. Dahal RK, Hasegawa S, Nonomura A, Yamanaka M, Dhakal S, Paudyal P (2008a) Predictive modeling of rainfall-induced landslide hazard in the Lesser Himalaya of Nepal based on weights-of evidence. Geomorphology 102(3–4):496–510CrossRefGoogle Scholar
  32. Dahal RK, Hasegawa S, Nonomura A, Yamanaka M, Masuda T, Nishino K (2008b) GIS-based weights-of-evidence modeling of rainfall-induced landslides in small catchments for landslide susceptibility mapping. Environ Geol 54(2):314–324CrossRefGoogle Scholar
  33. Dai FC, Lee CF, Ngai YY (2002) Landslide risk assessment and management: an overview. Eng Geol 64:65–87CrossRefGoogle Scholar
  34. Daneshfar B, Benn K (2002) Spatial relationships between natural seismicity and faults, southeastern Ontario and north-central New York State. Tectonophysics 353:31–44Google Scholar
  35. Dangol V (2000) Slope instability study and hazard mapping of Syangja district, Western Nepal. Unpublished Report, ICIMOD, Katmandu, NepalGoogle Scholar
  36. Dangol V, Upreti BN, Dhital MR, Wagner A, Bhattarai TN, Bhandari AN, Pant SR, Sharma MP (1993) Engineering geological study of a proposed road corridor in Eastern Nepal. Bull Dept Geol Tribhuvan University 3(1):91–107Google Scholar
  37. Deoja B, Dhital MR, Thapa B, Wagner A (1991) Mountain risk engineering handbook. ICIMOD, Kathmandu, 857 ppGoogle Scholar
  38. Devkota K, Regmi AD, Pourghasemi HR, Yishida K, Pradhan B, Ryu IC, Dhital MR, Althuwaynee OF (2012) 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. Nat Hazards. doi: 10.1007/s11069-012-0347-6
  39. Dhital MR, Upreti BN, Dangol V, Bhandari AN, Bhattarai TN (1991) Geological engineering methods applied in mountain road survey: an example from Baitadi–Darchula Road Project (Nepal). J Nepal Geol Soc 7:49–67Google Scholar
  40. Dhital MR (2005) Landslide investigation and mitigation in Himalayas: focus on Nepal. Proceedings of the International Symposium “Landslide Hazard in Orogenic Zone from the Himalaya to Island Arc in Asia,” 25–26 September, Kathmandu, Nepal, pp 1–15Google Scholar
  41. DWIDP (2009) The study on disaster risk management for Narayanghar–Mugling Highway. Unpublished reportGoogle Scholar
  42. Ercanoglu M, Gokceoglu C (2002) Assessment of landslide susceptibility for a landslide prone area (north of Yenice, NW Turkey) by fuzzy approach. Environ Geol 41:720–730CrossRefGoogle Scholar
  43. Ercanoglu M, Gokceoglu C (2004) Use of fuzzy relations to produce landslide susceptibility map of a landslide prone area (West Black Sea Region, Turkey). Eng Geol 75:229–250CrossRefGoogle Scholar
  44. Ganser A (1964) Geology of the Himalaya. Inter Science John Wiley, LondonGoogle Scholar
  45. Gokceoglu C, Sonmez H, Nefeslioglu HA, Duman TY, Can T (2005) The 17 March 2005 Kuzulu landslide (Sivas, Turkey) and landslide-susceptibility map of its near vicinity. Eng Geol 81:65–83CrossRefGoogle Scholar
  46. Gupta RP, Joshi BC (1990) Landslide hazard zonation using the GIS approach—a case study from the Ramganga Catchment, Himalayas. Eng Geol 28:119–131CrossRefGoogle Scholar
  47. Guzzetti F, Carrara A, Cardinali M, Reichenbach P (1999) Landslide hazard evaluation: a review of current techniques and their application in a multi-scale study, Central Italy. Geomorphology 31:181–216CrossRefGoogle Scholar
  48. Guzzetti F, Reichenbach P, Ardizzone F, Cardinali M, Galli M (2006) Estimating the quality of landslide susceptibility models. Geomorphology 81:166–184CrossRefGoogle Scholar
  49. Hagen T (1969) Report on the geological survey of Nepal preliminary reconnaissance: Zürich, Mémoires de la soc. Helvétique des sci. naturelles, p 185Google Scholar
  50. Kanungo DP, Arora MK, Sarkar S, Gupta RP (2006) A comparative study of conventional, ANN black box, fuzzy and combined neural and fuzzy weighting procedures for landslide susceptibility zonation in Darjeeling Himalayas. Eng Geol 85:347–366Google Scholar
  51. Kanungo DP, Arora MK, Gupta RP, Sarkar S (2008) Landslide risk assessment using danger pixel and fuzzy concepts in Darjeeling Himalayas. Landslides 5:407–416Google Scholar
  52. Kolat C, Doyuran V, Ayday C, Suzen LM (2006) Preparation of a geotechnical microzonation model using geographical information systems based on multicriteria decision analysis. Eng Geol 87 (3–4):241–255Google Scholar
  53. Lee S (2005) Application of logistic regression model and its validation for landslide susceptibility mapping using GIS and remote sensing data. Int J Remote Sens 26:1477–1491CrossRefGoogle Scholar
  54. Lee S (2007a) Comparison of landslide susceptibility maps generated through multiple logistic regression for three test areas in Korea. Earth Surf Proc & Landf 32:2133–2148CrossRefGoogle Scholar
  55. Lee S (2007b) Application and verification of fuzzy algebraic operators to landslide susceptibility mapping. Environ Geol 52:615–623CrossRefGoogle Scholar
  56. Lee S, Ryu JH, Lee MJ, Won JS (2006) The application of artificial neural networks to landslide susceptibility mapping at Janghung, Korea. Mathematical Geol 38:199–220CrossRefGoogle Scholar
  57. Lee S, Pradhan B (2006) Probabilistic landslide hazards and risk mapping on Penang Island, Malaysia. J Earth Syst Sci 115(6):661–667CrossRefGoogle Scholar
  58. Lee S, Pradhan B (2007) Landslide hazard mapping at Selangor, Malaysia using frequency ratio and logistic regression models. Landslides 4(1):33–41CrossRefGoogle Scholar
  59. Mancini F, Ceppi C, Ritrovato G (2010) GIS and statistical analysis for landslide susceptibility mapping in the Daunia area, Italy. Nat Hazards Earth Syst Sci 10:1851–1864Google Scholar
  60. Mathew J, Jha VK, Rawat GS (2007) Weights of evidence modeling for landslide hazard zonation mapping in part of Bhagirathi valley, Uttarakhand. Curr Sci 92(5):628–638Google Scholar
  61. Melchiorre C, Matteucci M, Azzoni A, Zanchi A (2008) Artificial neural networks and cluster analysis in landslide susceptibility zonation. Geomorphology 94:379–400CrossRefGoogle Scholar
  62. Mohammady M, Pourghasemi HR, Pradhan B (2012) Landslide susceptibility mapping at Golestan Province, Iran: a comparison between frequency ratio, Dempster-Shafer, and weights-of-evidence models. J Asian Earth Sci 61:221–236. doi: 10.1016/j.jseaes.2012.10.005
  63. Moore ID, Grayson RB (1991) Terrain-based catchment partitioning and runoff prediction using vector elevation data. Water Resour Res 27(6):1171–1191Google Scholar
  64. Muthu K, Petrou M, Tarantino C, Blonda P (2008) Landslide possibility mapping using fuzzy approaches. IEEE Trans Geosci Remote Sens 46(4):1253–126CrossRefGoogle Scholar
  65. Nagarajan R, Roy A, Vinod Kumar R, Mukherjee A, Khire MV (2000) Landslide hazard susceptibility mapping based on terrain and climatic factors for tropical monsoon regions. Bull Eng Geol Env 58:275–287CrossRefGoogle Scholar
  66. Nefeslioglu H, Sezer E, Gokceoglu C, Bozkir A, Duman T (2010a) Assessment of landslide susceptibility by decision trees in the metropolitan area of Istanbul, Turkey. Mathematical Problems in Engineering 2010:Article ID 901095. doi: 10.1155/2010/901095
  67. Nefeslioglu HA, Sezer E, Gokceoglu C, Bozkir AS, Duman TY (2010b) Assessment of landslide susceptibility by decision trees in the metropolitan area of Istanbul, Turkey. Mathematical Problems in Engineering 2012:1–15. doi: 10.1155/2010/901095
  68. Neuhäuser B, Terhorst B (2007) Landslide susceptibility assessment using “weights-of-evidence” applied to a study area at the Jurassic escarpment (SW-Germany). Geomorphology 86:12–24CrossRefGoogle Scholar
  69. Oh HJ, Pradhan B (2011) Application of a neuro-fuzzy model to landslide-susceptibility mapping for shallow landslides in a tropical hilly area. Comput Geosci 7(9):1264–1276. doi: 10.1016/j.cageo.2010.10.012 CrossRefGoogle Scholar
  70. Ohlmacher GC, Davis JC (2003) Using multiple logistic regression and GIS technology to predict landslide hazard in northeast Kansas, USA. Eng Geol 69(3–4):331–343Google Scholar
  71. Oyagi N (1984) Landslides in weathered rocks and residuals soils in Japan an surrounding areas: state-of-the-art report. 4th International Symposium on Landslides, Toronto, pp 1–131Google Scholar
  72. Pantha BR, Yatabe R, Bhandari NP (2008) GIS-based landslide susceptibility zonation for roadside slope repair and maintenance in the Himalayan region. Episodes 31(4):384–291Google Scholar
  73. Poudyal CP, Chang C, Oh HJ, Lee S (2010) Landslide susceptibility maps comparing frequency ratio and artificial neural networks: a case study from the Nepal Himalaya. Environ Earth Sci 61:1049–1064CrossRefGoogle Scholar
  74. Pourghasemi HR, Pradhan B, Gokceoglu C, Mohammadi M, Moradi HR (2012a) Application of weights-of-evidence and certainty factor models and their comparison in landslide susceptibility mapping at Haraz watershed Iran. Arab J Geosci. doi: 10.1007/s12517-012-0532-7
  75. Pourghasemi HR, Pradhan B, Gokceoglu C, Deylami Moezzi K (2012b) A comparative assessment of prediction capabilities of Dempster–Shafer and weights-of-evidence models in landslide susceptibility mapping using GIS. Geomatics Nat Hazards Risk. doi: 10.1080/19475705.2012.662915
  76. Pourghasemi HR, Pradhan B, Gokceoglu C (2012c) Application of fuzzy logic and analytical hierarchy process (AHP) to landslide susceptibility mapping at Haraz watershed, Iran. Nat Hazards 63(2):965–996. doi: 10.1007/s11069-012-0217-2
  77. Pourghasemi HR, Mohammady M, Pradhan B (2012d) Landslide susceptibility mapping using index of entropy and conditional probability models in GIS: Safarood Basin, Iran. Catena 97:71–84. doi: 10.1016/j.catena.2012.05.005 CrossRefGoogle Scholar
  78. Pourghasemi HR, Gokceoglu C, Pradhan B, Deylami Moezzi K (2012e) Landslide susceptibility mapping using a spatial multi criteria evaluation model at Haraz Watershed, Iran. In: Pradhan B, Buchroithner M (eds) Terrigenous mass movements. Springer, Berlin, pp 23–49. doi: 10.1007/978-3-642-25495-6-2
  79. Pourghasemi HR, Goli Jirandeh A, Pradhan B, Xu C, Gokceoglu C (2012f) Landslide susceptibility mapping using support vector machine and GIS. J Earth Syst Sci 61(15):221–236Google Scholar
  80. Pourghasemi HR, Pradhan B, Gokceoglu C (2012g) Remote sensing data derived parameters and its use in landslide susceptibility assessment using Shannon’s entropy and GIS. Appl Mech Mater 225:486–491Google Scholar
  81. Pradhan B (2010a) Remote sensing and GIS-based landslide hazard analysis and cross validation using multivariate logistic regression model on three test areas in Malaysia. Adv Space Res 45:1244–1256CrossRefGoogle Scholar
  82. Pradhan B (2010b) Application of an advanced fuzzy logic model for landslide susceptibility analysis. Int J Comput Intell Syst 3:370–381Google Scholar
  83. Pradhan B (2010c) Landslide susceptibility mapping of a catchment area using frequency ratio, fuzzy logic and multivariate logistic regression approaches. J Indian Soc Remote Sens 38(2):301–320CrossRefGoogle Scholar
  84. Pradhan B (2012) A comparative study on the predictive ability of the decision tree, support vector machine and neuro-fuzzy models in landslide susceptibility mapping using GIS. Comput Geosci 51:350–365. doi: 10.1016/j.cageo.2012.08.023
  85. Pradhan B (2011a) Manifestation of an advanced fuzzy logic model coupled with geoinformation techniques for landslide susceptibility analysis. Environ Ecol Stat 18(3):471–493. doi: 10.1007/s10651-010-0147-7 CrossRefGoogle Scholar
  86. Pradhan B (2011b) Use of GIS-based fuzzy logic relations and its cross application to produce landslide susceptibility maps in three test areas in Malaysia. Environ Earth Sci 63(2):329–349CrossRefGoogle Scholar
  87. Pradhan B, Singh RP, Buchroithner MF (2006) Estimation of stress and its use in evaluation of landslide prone regions using remote sensing data. Adv Space Res 37:698–709CrossRefGoogle Scholar
  88. Pradhan B, Lee S (2009) Landslide risk analysis using artificial neural network model focusing on different training sites. I J Physical Sci 3(11):1–15Google Scholar
  89. Pradhan B, Buchroithner MF (2010) Comparison and validation of landslide susceptibility maps using an artificial neural network model for three test areas in Malaysia. Environ Eng Geosci 16(2):107–126Google Scholar
  90. Pradhan B, Pirasteh S (2010) Comparison between prediction capabilities of neural network and fuzzy logic techniques for landslide susceptibility mapping. Disaster Adv 3(2):26–34Google Scholar
  91. Pradhan B, Lee S (2010a) Delineation of landslide hazard areas using frequency ratio, logistic regression and artificial neural network model at Penang Island, Malaysia. Environ Earth Sci 60:1037–1054CrossRefGoogle Scholar
  92. Pradhan B, Lee S (2010b) Regional landslide susceptibility analysis using back propagation neural network model at Cameron Highland, Malaysia. Landslides 7:13–30CrossRefGoogle Scholar
  93. Pradhan B, Lee S, Buchroithner MF (2010a) Remote sensing and GIS-based landslide susceptibility analysis and its cross-validation in three test areas using a frequency ratio model. Photogrammetrie, Fernerkundung, Geoinformation (PFG) 1(1):17–32Google Scholar
  94. Pradhan B, Lee S, Buchroithner MF (2010b) A GIS-based back-propagation neural network model and its cross application and validation for landslide susceptibility analyses. Comput Environ & Urban Syst 34:216–235CrossRefGoogle Scholar
  95. Pradhan B, Chaudhari A, Adinarayana J, Buchroithner MF (2012) Soil erosion assessment and its correlation with landslide events using remote sensing data and GIS: a case study at Penang Island, Malaysia. Environ Monit Assess 184(2):715–727. doi: 10.1007/s10661-011-1996-8 CrossRefGoogle Scholar
  96. Regmi AD, Yoshida K, Dhital MR, Devkota K (2012a) Effect of rock weathering, clay mineralogy, and geological structures in the formation of large landslide, a case study from Dumre Besei landslide, Lesser Himalaya Nepal. Landslides. doi: 10.1007/s10346-011-0311-7
  97. Regmi AD, Yoshida K, Nagata H, Pradhan AMS, Pradhan B, Pourghasemi HR (2012b) The relationship between geology and rock weathering on the rock instability along Mugling–Narayanghat road corridor, Central Nepal Himalaya. Nat Hazards. doi: 10.1007/s11069-012-0497-6
  98. Regmi NR, Giardino JR, Vitek JD (2010a) Assessing susceptibility to landslides: using models to understand observed changes in slopes. Geomorphology 112:25–38CrossRefGoogle Scholar
  99. Regmi NR, Giardino JR, Vitek JD (2010b) Modeling susceptibility to landslides using the weight of evidence approach: Western Colorado, USA. Geomorphology 115:172–187CrossRefGoogle Scholar
  100. Saito H, Nakayama D, Matsuyama H (2009) Comparison of landslide susceptibility based on a decision-tree model and actual landslide occurrence: the Akaishi Mountains, Japan. Geomorphology 109:108–121CrossRefGoogle Scholar
  101. Saha AK, Gupta RP, Arora MK (2002) GIS-based landslide hazard zonation in the Bhagirathi (Ganga) valley, Himalayas. Int J Remote Sens 23:357–369CrossRefGoogle Scholar
  102. Saha AK, Gupta RP, Sarkar I, Arora KM, Csaplovics E (2005) An approach for GIS-based statistical landslide susceptibility zonation with a case study in the Himalayas. Landslides 2(1):61–69CrossRefGoogle Scholar
  103. Sezer EA, Pradhan B, Gokceoglu C (2011) Manifestation of an adaptive neuro-fuzzy model on landslide susceptibility mapping: Klang valley, Malaysia. Expert Syst Appl 38(7):8208–8219CrossRefGoogle Scholar
  104. Sharma M, Kumar R (2008) GIS-based landslide hazard zonation: a case study from the Parwanoo area, Lesser and Outer Himalaya, H.P., India. B Eng Geol Environ 67:129–137CrossRefGoogle Scholar
  105. Stöcklin J (1980) Geology of Nepal and its regional frame. J Geol Soc London 137:1–34CrossRefGoogle Scholar
  106. Stöcklin J, Bhattarai KD (1978) Geology of the Kathmandu area and central Mahabharat range, Nepal Himalaya. Unpublished Report of the Department of Mines and Geology/UNDP, 86 pGoogle Scholar
  107. Upreti BN (1999) An overview of the stratigraphy and tectonics of the Nepal Himalaya. J Asian Earth Sci 17:577–606CrossRefGoogle Scholar
  108. Vahidnia MH, Alesheikh AA, Alimohammadi A, Hosseinali F (2010) A GIS-based neuro-fuzzy procedure for integrating knowledge and data in landslide susceptibility mapping. Comput Geosci 36(29):1101–1114Google Scholar
  109. van Westen CJ, Rengers N, Terlien MTJ, Soeters R (1997) Prediction of the occurrence of slope instability phenomena through GIS based hazard zonation. Geol Rundsch 86:404–414CrossRefGoogle Scholar
  110. van Westen CJ, Seijmonsbergen AC, Mantovani F (1999) Comparing landslide hazard maps. Nat Hazards 20(2/3):137–158CrossRefGoogle Scholar
  111. van Westen CJ, Rengers N, Soeters R (2003) Use of geomorphological information in indirect landslide susceptibility assessment. Nat Hazards 30:399–419CrossRefGoogle Scholar
  112. Wan S (2009) A spatial decision support system for extracting the core factors and thresholds for landslide susceptibility map. Eng Geol 108:237–251CrossRefGoogle Scholar
  113. Wan S, Lei TC, Huang PC, Chou TY (2008) Knowledge rules of debris flow event: a case study for investigation ChenYu Lan River Taiwan. Eng Geol 98:102–114CrossRefGoogle Scholar
  114. Wan S, Lei TC (2009) A knowledge-based decision support system to analyze the debris-flow problems at Chen-Yu-Lan River, Taiwan. Knowl Based Syst 2(8):580–588Google Scholar
  115. Wan S, Lei TC, Chou TY (2009) Anovel data mining technique of analysis and classification for landslide problems. Nat Hazards 52:211–230Google Scholar
  116. Xiao CC, Tian Y, Shi WZ, Guo QH, Wu L (2010) A new method of pseudo absence data generation in landslide susceptibility mapping with a case study of Shenzhen. Sci China Tech Sci 53(1):75–84. doi: 10.1007/s11431-010-3219-x CrossRefGoogle Scholar
  117. Yao X, Tham LG, Dai FC (2008) Landslide susceptibility mapping based on support vector machine: a case study on natural slopes of Hong Kong, China. Geomorphology 101(4):572–582Google Scholar
  118. Yeon YK, Han JG, Ryu KH (2010) Landslide susceptibility mapping in Injae, Korea, using decision tree. Eng Geol 116:274–283CrossRefGoogle Scholar
  119. Yilmaz I (2009a) A case study from Koyulhisar (Sivas-Turkey) for landslide susceptibility mapping by artificial neural networks. B Eng Geol Environ 68:297–306CrossRefGoogle Scholar
  120. Yilmaz I (2009b) Landslide susceptibility mapping using frequency ratio, logistic regression, artificial neural networks and their comparison: a case study from Kat landslides (Tokat-Turkey). Comput Geosci 35:1125–1138CrossRefGoogle Scholar
  121. Yilmaz I (2010a) Comparison of landslide susceptibility mapping methodologies for Koyulhisar, Turkey: conditional probability, logistic regression, artificial neural networks, and support vector machine. Environ Earth Sci 61:821–836CrossRefGoogle Scholar
  122. Yilmaz I (2010b) The effect of the sampling strategies on the landslide susceptibility mapping by conditional probability and artificial neural networks. Environ Earth Sci 60:505–519CrossRefGoogle Scholar
  123. Zhu C, Wang X (2009) Landslide susceptibility mapping: a comparison of information and weights-of evidence methods in Three Gorges Area. Int Conf Environ Sci Inf Appl Technol 2009:342–346. doi: 10.1109/ESIAT.2009.187 Google Scholar
  124. Zare M, Pourghasemi HR, Vafakhah M, Pradhan B (2012) Landslide susceptibility mapping at Vaz watershed (Iran) using an artificial neural network model: a comparison between multi-layer perceptron (MLP) and radial basic function (RBF) algorithms. Arab J Geosci. doi: 10.1007/s12517-012-0610-x

Copyright information

© Saudi Society for Geosciences 2013

Authors and Affiliations

  • Amar Deep Regmi
    • 1
  • Krishna Chandra Devkota
    • 2
  • Kohki Yoshida
    • 1
  • Biswajeet Pradhan
    • 3
  • Hamid Reza Pourghasemi
    • 4
  • Takashi Kumamoto
    • 5
  • Aykut Akgun
    • 6
  1. 1.Department of Geology, Faculty of ScienceShinshu UniversityMatsumotoJapan
  2. 2.Department of GeologyKyungpook National UniversityDaeguSouth Korea
  3. 3.Geospatial Information Science Research Centre (GISRC), Department of Civil Engineering, Faculty of EngineeringUniversiti Putra MalaysiaSerdangMalaysia
  4. 4.College of Natural Resources and Marine SciencesTarbiat Modares University (TMU)MazandaranIran
  5. 5.Department of Earth Sciences, Faculty ScienceOkayama UniversityOkayamaJapan
  6. 6.Geological Engineering DepartmentBlack Sea Technical UniversityTrabzonTurkey

Personalised recommendations