Application of GIS based data driven evidential belief function model to predict groundwater potential zonation

Journal of Hydrology 513 (2014) 283–300

Contents lists available at ScienceDirect

Journal of Hydrology journal homepage: www.elsevier.com/locate/jhydrol

Application of GIS based data driven evidential belief function model to predict groundwater potential zonation Haleh Nampak a, Biswajeet Pradhan a,⇑, Mohammad Abd Manap b a Faculty of Engineering, Department of Civil Engineering, Geospatial Information Science Research Centre (GISRC), University Putra Malaysia, Serdang, Selangor DarulEhsan 43400, Malaysia b Minerals and Geoscience Department (JMG), 19-22th Floor, BangunanTabung Haji, JalanTunRazak, Kuala Lumpur 50658, Malaysia

a r t i c l e

i n f o

Article history: Received 29 September 2013 Received in revised form 17 February 2014 Accepted 19 February 2014 Available online 20 March 2014 This manuscript was handled by Geoff Syme, Editor-in-Chief, with the assistance of Craig T. Simmons, Associate Editor Keywords: Groundwater potential Evidential belief function (EBF) Logistic regression (LR) GIS Remote sensing Malaysia

s u m m a r y The objective of this paper is to exploit potential application of an evidential belief function (EBF) model for spatial prediction of groundwater productivity at Langat basin area, Malaysia using geographic information system (GIS) technique. About 125 groundwater yield data were collected from well locations. Subsequently, the groundwater yield was divided into high (P11 m3/h) and low yields (<11 m3/h) respectively, based on the groundwater classification standard recommended by Department of Mineral and Geosciences (JMG), Malaysia. Out of all of the borehole data, only 60 wells possessed higher yield at P 11 m3/h. Further, these wells were randomly divided into a testing dataset 70% (42 wells) for training the model and the remaining 30% (18 wells) was used for validation purpose. To perform cross validation, the frequency ratio (FR) approach was applied into remaining groundwater wells with low yield to show the spatial correlation between the low potential zones of groundwater productivity. A total of twelve groundwater conditioning factors that affect the storage of groundwater occurrences were derived from various data sources such as satellite based imagery, topographic maps and associated database. Those twelve groundwater conditioning factors are elevation, slope, curvature, stream power index (SPI), topographic wetness index (TWI), drainage density, lithology, lineament density, land use, normalized difference vegetation index (NDVI), soil and rainfall. Subsequently, the Dempster–Shafer theory of evidence model was applied to prepare the groundwater potential map. Finally, the result of groundwater potential map derived from belief map was validated using testing data. Furthermore, to compare the performance of the EBF result, logistic regression model was applied. The success-rate and prediction-rate curves were computed to estimate the efficiency of the employed EBF model compared to LR method. The validation results demonstrated that the success-rate for EBF and LR methods were 83% and 82% respectively. The area under the curve for prediction-rate of EBF and LR methods were calculated 78% and 72% respectively. The outputs achieved from the current research proved the efficiency of EBF in groundwater potential mapping. Ó 2014 Elsevier B.V. All rights reserved.

1. Introduction Groundwater is one of the most important natural resources worldwide serving as a major source of water to communities, industries and agricultural purposes (Ayazi et al., 2010; Manap et al., 2012, 2013; Neshat et al., 2013; Pradhan, 2009). Groundwater is defined as water in saturated zone (Fitts, 2002) which fills the pore spaces among mineral grains or cracks and fractured rocks in rock mass. Groundwater is usually formed by rain or snow melts

⇑ Corresponding author. Tel.: +60 3 89466383; fax: +60 3 89468470. E-mail addresses: (B. Pradhan).

[email protected],

http://dx.doi.org/10.1016/j.jhydrol.2014.02.053 0022-1694/Ó 2014 Elsevier B.V. All rights reserved.

[email protected]

which seeps down through the soil into the underlying rocks (Banks et al., 2002; Saraf and Choudhury 1998). The traditional approach of groundwater exploration through drilling, geological, hydro-geological, and geophysical methods are costly and time consuming (Sander et al., 1996; Singh and Prakash, 2002). A common method used to prepare groundwater potential maps is mainly based on ground surveys (Ganapuram et al., 2009). Recently, with the popular use of geographic information systems (GISs) and remote sensing (RS) based technologies, groundwater potential mapping has become an easy procedure (Singh and Prakash, 2002). GIS is a powerful tool to handle huge amount of spatial data and can be used in the decision making process in a number of fields such as geology and environmental management. The information about surface features related to

284

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

groundwater such as landforms, land use, lineaments can be extracted through RS data. Those data can be easily entered to GIS to integrate with other associated tabular data, followed by spatial analysis and visual interpretation (Jha et al., 2007). In Malaysia, groundwater has been considered as a hot issue especially during prolonged drought periods. The Selangor state faced a long period of drought in 1998 due to El Nino effects. Groundwater, in other states of Malaysia such as Kelantan, Perlis, Terengganu, Pahang, Sarawak and Sabah has been utilized as a main source of water supply (Suratman, 2004). Moreover, it is being exploited by private sectors for commercial production of mineral water. The failure to recognize the vast potential zone is the main reason for underutilization of groundwater resources in Malaysia. More recently, a lot of studies have been applied using index based models for assessing groundwater potential mapping (Dar et al., 2010; Madrucci et al., 2008; Nag et al., 2012; Prasad et al., 2008). In some studies probabilistic models such as multi-criteria decision analysis (Chenini et al., 2010; Gupta and Srivastava, 2010; Murthy and Mamo, 2009), weights-of-evidence (Corsini et al., 2009; Lee et al., 2012), frequency ratio (FR) (Oh et al., 2011), and analytical hierarchy process (AHP) (Chowdhury et al., 2009; Pradhan, 2009) have been used for groundwater potential mapping. In recent years, some soft computing techniques such as fuzzy logic (Shahid et al., 2002; Ghayoumian et al., 2007), numerical modelling and decision tree (DT) (Chenini and Mammou, 2010) approaches have been applied in groundwater potential mapping. Magesh et al. (2012) carried out weighted overlay analysis using a multi-influencing factors and assigned weights to various groundwater conditioning factors. In this paper, an EBF model was applied for groundwater potential mapping (Shafer, 1976; Dempster, 2008). The EBF approach has been popularly used in mineral potential mapping (Moon, 1990). Carranza and Hale (2003) proposed a data-driven approach based on the Dempster’s rule of combination using GIS for mineral potential mapping (Carranza and Castro, 2006). Similarly multivariate based logistic regression model (LR) has been applied in groundwater potential mapping (Ozdemir, 2011). LR model is useful to describe the significance and correlation of groundwater occurence to each conditioning factor. The main aim of the present study is to evaluate the efficiency of the EBF model for groundwater potential mapping. In order to compare the robustness of the proposed EBF model, a well-known LR model was applied to identify the significant groundwater conditioning factors and subsequently the EBF model was re-run to check its efficiency. Through this analysis, the relationships between wells and each conditioning factor will be quantitatively defined. The main difference between this research and the approaches described in the aforementioned publications is that an EBF model is applied and the result is validated for groundwater potential mapping in the Langat basin, Malaysia. The application of EBF in groundwater potential mapping provides originality to this study. 2. Study area Langat River catchment is located in the southeast part of Selangor state. It is considered as the most urbanized river basin in Malaysia providing two thirds of the water in the Selangor state. However, with the large-scale physical and economic development in the area, water scarcity and water quality deterioration is emerging in recent years. Bringemeier (2006) reported that the number of water-intensive enterprises (e.g. steel works, pulp and paper industry, power plants) play a vital role in water shortage and groundwater quality reduction in this region. This trend will increase the water crisis problem within the next 20 years.

The study area lies between 101°190 2000 E to 102°10 1000 E latitude and 2°400 1500 N to 3°160 1500 longitude covering an area about 2100 km2 (Fig. 1). The basin includes several districts such as Kuala Langat, Sepang, Hulu Langat and part of Seremban district. Topographically, the area is divided into three distinct regions (Manap et al., 2012). The Langat River has several tributaries with the principal ones being the Semenyih River, the Lui River, and the Beranang River. The main and other tributaries flow westward to the Malacca strait and create flat alluvial zone where soils are mostly peat with clay and silt soil. High average and constant annual temperature coupling with high precipitation and high humidity affect the hydrology and geomorphology of the study area. In addition, the weather conditions are influenced by the southwest monsoon that blows across the Strait of Malacca (Juahir et al., 2010) and the study area experiences the wet season in April to November and a relatively drier period from January to March. According to Malaysian Meteorological Services Department, the mean annual temperature of Hulu Langat area is 32 °C and the annual rainfall about 2316.5– 4223.4 mm. The bedrock geology of the study area mainly consists of granite, sedimentary, alluvium and volcanic rocks (Fig. 2). 3. Data used 3.1. Groundwater occurrence characteristics The groundwater data such as topography, number of wells, yield and depth were obtained from Department of Mineral and Geosciences (JMG), Malaysia. Groundwater yield is based on actual pumping test of groundwater well e.g. m3/h. Moreover, groundwater potential is based on prediction of the best potential for groundwater extraction in the study area. There are 125 individual groundwater borehole wells in the study area. In the year 2007, JMG had set up their own standard of groundwater potential classification classes which are based on yield value of well production. The high productivity value was based on yield value P 11 m3/h. The groundwater productivity data from 60 wells were selected and randomly divided into a training dataset 70% (42 groundwater wells) and a validation dataset 30% (18 groundwater wells). Additionally, same numbers of points (42) were selected as non-well occurrences pixels (i.e. the absence of a well over the area) to allocate the value of 0 for applying LR model. It is expected to improve the accuracy of the outcomes through the consideration of non-well occurrences pixels in the logistic regression analysis. The rest of the well occurrences (18 points) were applied for the purpose of validation. The remaining groundwater wells with less than 11 m3/h productivity were used for FR modelling for a crossvalidation and to compare the modelling results. Fig. 3 shows the groundwater well locations of the study area. 3.2. Groundwater conditioning factors Generally, the occurrence and movement of groundwater in a given area is governed by various conditioning factors. These factors are topography, lithology, geological structure, fracture density, aperture and connectivity of fractures, secondary porosity, groundwater table distribution, groundwater recharge, slope, drainage pattern, landform, land use and land cover, climatic condition (Mukherjee, 1996; Oh et al., 2011; Ozdemir, 2011). To create a groundwater potential map, the spatial database was considered to be a set of related spatial conditioning factors that influence groundwater occurrence. In this study, a total of twelve conditioning factors are used. Those conditioning factors are: elevation, slope, curvature, stream power index (SPI), topographic wetness index (TWI), rainfall, normalized difference vegetation index (NDVI),

285

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Fig. 1. Location map of the study area.

drainage density, lineament density, geology, land use and soil (Table 1). 3.2.1. Topographic factors A digital elevation model (DEM) with 20 m resolution was used (from the 1:25,000 scale topographic map) to derive various topographic factors such as elevation (Fig. 4a), slope angle (Fig. 4b) and slope curvature (Fig. 4c) using ArcGIS. 3.2.2. Water related factors Various factors such as drainage density, TWI and SPI were obtained from the DEM as a measure of surface water, sub-surface water and groundwater. The TWI has been widely applied to explain the impact of topography on the location and size of saturated source zones of runoff generation. Eq. (1) proposed by Moore et al. (1991) was used for TWI computation:

TWI ¼ lnðAS= tan bÞ

ð1Þ 2

where AS is the specific catchment’s area (m /m) and b is slope gradient (in degrees). The TWI map is illustrated in Fig. 4d. Fig. 4e presents the SPI which is a measure of the erosive power of water flow based on the assumption that discharge is proportional to specific catchment area Eq. (2) (Moore et al., 1991).

SPI ¼ ðAS  tan bÞ

ð2Þ

The geology is one of the most significant indicators of hydrogeological features (Charon, 1974). The relationship between infiltration and runoff is controlled largely by permeability, which is in turn a function of the rock type and fracturing of the underlying rock or surface bedrock (Fig. 4f). High river density values show high surface runoff of an area (Prasad et al., 2008) and then such zones are regarded as favorable for arresting excessive runoff (Krishnamurthy et al., 1996).

286

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Fig. 2. Lithological map showing the lineaments in the study area.

3.2.3. Geological factors The lineament map was provided by the Department of Mineral & Geosciences, Malaysia and the lineament density maps was generated using ArcGIS spatial analyst extension (Fig. 4g). The bedrock in the mountainous area consists of Permian igneous rock and Pre-Devonian schist and phyllite (Hutchison and Tan, 2009). The bedrock of hilly part includes Kajang formation and Kenny hill formation which consists of phyllite, shale and quartzite (Yin, 1976). Unconsolidated gravel, sand, silt and clay formed as the Simpang formation in lowlands (Fig. 4h).

built up area, mining, rubber, oil palm, plantation, grassland, forest and swamps. The NDVI map was produced from RS imagery showing the surface vegetation coverage and density in an image. The NDVI value was computed using Eq. (3) (Pradhan et al., 2010). Fig. 4j shows the NDVI map.

3.2.4. Land use/Land cover The land use map (Fig. 4i) was obtained from the Department of Agriculture, Malaysia at a scale of 1:25,000. For analysis, the land use map was constructed with ten classes which mainly include

3.2.5. Soil The soil map was obtained from the Department of Agriculture, Malaysia. The soil cover varied between highland and lowland zones (Fig. 4k) and it contains coarse to fine sandy loam. As the

NDVI ¼ ðNIR  VISÞ=ðNIR þ VISÞ

ð3Þ

where VIS and NIR stand for the spectral reflectance measurements acquired in the visible (red) and near-infrared regions, respectively.

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

287

Fig. 3. Groundwater well locations in the study area. Table 1 The spatial database construction. Classification

Subclassification

Data type

Scale

Groundwater Base map

Tube well Topographic map Geological map Lineament map Land use map Soil map NDVI Rainfall map

Point Point, line and polygon coverage Polygon Polyline Grid Grid Grid Grid

1:25,000 1:25,000 1:63,360 1:63,360 20  20 20  20 20  20 20  20

major and other tributaries flow westward to the Malacca strait by creating flat alluvial zone where soils are mostly peat with clay and silt soil. 3.2.6. Rainfall The precipitation amount determines the amount of water that would be percolating into the groundwater system as the major source of recharge. For that reason, an annual rainfall map of the study area was prepared using the historical rainfall data of past 29 years (1981–2010) measured at the meteorology stations located in the surrounding study area (Fig. 4l). All the aforementioned groundwater conditioning factors were converted to a grid comprising of 20  20 m grid cells, with an area of 3285 rows by 4001 columns (total number of cells is 6,117,068).

(4) description and visual interpretation of the results. Fig. 5 illustrates the general methodological flow chart used in this study. 4.1. Evidential belief function (EBF) model To apply the EBF model, at first the thematic layers (groundwater conditioning factors) should be transformed into evidential data layers. After that they can be integrated to produce a predictive groundwater potential index map (GWPI) using the quantitative knowledge of the spatial relationship between the wells and the groundwater conditioning factors. The EBF model contains Bel (degree of belief), Dis (degree of disbelief), Unc (degree of uncertainty) and Pls (degree of plausibility) in range of [0, 1] (Carranza and Hale, 2003; Althuwaynee et al., 2012; Pradhan et al., 2014). The primary parts of the theory are shown by Bel and Pls as lower and upper probability respectively and basic probability assignment function (bpa or m) describes a mapping of the power set to (0–1). Eq. (4) shows the Dempster–Shafer theory of evidence which is synthesized from Carranza and Hale (2003), Park (2011), Althuwaynee et al. (2012).

m : PðHÞ ¼ f0; 1g mðøÞ ¼ 0 mðBÞ ¼ 1 mðBÞ ¼ 1 :

4. Methodology The groundwater potential zonation mapping consists of four main steps: (1) data collection and spatial database construction for the groundwater related conditioning factors, (2) assessment of groundwater potential using the relationships between wells and groundwater conditioning factors, (3) validation of the results,

ð4Þ

X

mðHÞ ¼ 1

HPðHÞ

Belief is a lower probability which defines as the sum of all beliefs committed identically to every subset of B by m and plausibility shows the degree that the evidence remains plausible (Park, 2011). Based on bpa or mass function, belief and plausibility function can be expressed into following Eq. (5).

288

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Fig. 4. Groundwater conditioning factors; (a) elevation, (b) slope, (c) curvature, (d) TWI, (e) SPI, (f) river density, (g) lineament density, (h) lithology, (i) land use, (j) NDVI, (k) soil, and (l) rainfall.

BelðBÞ ¼

X mðHÞ

ð5Þ

BelðBÞ 6 PlsðBÞ

HB

PlsðBÞ ¼

X

mðHÞ

H\B–ø

These two functions have the following properties:

PlsðBÞ ¼ 1  Bel  Is negation of B and it is called disbelief As shown in Eq. (6), B  function and it is classical complement of B.

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

289

Fig. 4 (continued)

PlsðBÞ ¼

X H\B–ø

mðHÞ ¼ BelðBÞ ¼

X  mðHÞ ¼ BelðBÞ

ð6Þ

HB

Eq. (7) represents the difference between belief and plausibility indicating the uncertainty (ignorance) as well (Park, 2011).

PlsðBÞ  BelðBÞ ¼ Uncðignorance or doubtÞ If it is assumed that Unc = 0, therefore Pls = Bel

ð7Þ

The schematic relationships of three mass functions are presented in Fig. 6. 4.2. Logistic regression (LR) model LR involves a multivariate regression between a dependent variable and various independent variables (Hosmer and Lemeshow,

290

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Fig. 4 (continued)

2000). The aim of LR is to identify the suitable model to define the relationship between a dependent variable and conditioning factors to produce the coefficient for each variable (Umar et al., 2014). The ratio of each conditioning factor can be estimated by using the coefficients derived from logistic regression. LR is effective compared to linear regression because the dependent variable in logistic regression can be a mixture of continuous and categorical or binary variables. LR is useful to anticipate

the presence or absence of a feature or product relying on the values of predictive variables (Lee, 2005). LR model corresponds to the generalized linear model that can be computed as follows;

P ¼ ez =ð1 þ ez Þ

ð8Þ

where P is the probability of an occurrence. Z is a value from 1 to +1, defined by the following equation;

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

291

Fig. 5. Methodological flowchart of the study.

Z ¼ a þ B1 X 1 þ B2 X 2 þ    þ Bn X n

ð9Þ

where and Z is a linear combination function of the conditioning factors indicating a linear relationship. a is the cutoff of model, n is the number of conditioning factors, and B1, B2, . . ., Bn are coefficients, which deliberate the contribution of conditioning factors X1, X2, . . ., Xn (Ayalew and Yamagishi, 2005; Akgun, 2012). A positive coefficient implies the positive correlation between dependant and

conditioning factor, and negative coefficient represents the opposite effect. In the LR model, the dependent variable can be expressed as;

Z ¼ loge



 P ¼ log itðPÞ 1P

ð10Þ

where P is the probability that the well occurrence as dependent variable (binary) illustrating the presence or absence of

292

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

In this part, L Numbers of multiple spatial data layer should be considered as evidence (Eij), where (ij) represents (i) as amount of layers and (j) class attribute individually to obtain certain accurate results. The integrated EBF values of the groundwater conditioning factors were implemented one after another by using Eqs. (13)–(18). Table 2 illustrates the estimated EBF for the 12 groundwater conditioning factors. Eqs. (13) and (14) shows how Bel results could be derived;

k ¼ ðTpÞEij ¼ ½NðL \ EijÞ=NðLÞ=½NðEijÞ  NðL \ EijÞ=ðNðAÞ  NðLÞÞ ¼ N=D

ð13Þ

Fig. 6. Schematic relationships of evidential belief functions; (Carranza et al., 2005).

Bel ¼ kðTpÞEij= groundwater by values of 0 and 1, function Z indicates log (p/1  p) to base e, called as logit (P) and p/(1  p) represents odds or likelihood ratio. Probabilities differ between 0 and 1. While the value of Z increases, the probability of P strictly increases. As a probability becomes closer to 1, the numerator of the odds becomes larger relative to the denominator, and the odds become an increasingly large number. On the contrary, if a probability becomes closer to 0, the numerator of the odds becomes smaller relative to the denominator (Ayalew and Yamagishi, 2005; Kavzoglu et al., 2013). 4.3. Frequency ratio (FR) model

X

kðTpÞEij

ð14Þ

where N (L \ Eij) is the number of groundwater occurrence pixels in the domain

8 > < NðLÞ or Total number of groundwater occurrence > :P NðL \ EijÞ N (Eij) is the number of pixel in the domain

8 > < NðAÞ or Total number of pixels in the domain > :P NðEijÞ

Frequency ratio (FR) model as bivariate statistical technique can be applied as a simple geospatial assessment tool to compute the probabilistic relationship between dependent and independent variables (Oh et al., 2011). Here, in this research FR model was applied to illustrate the quantitative relationship between distribution of groundwater occurrences with low yield and conditioning factors as a cross validation approach. The calculation and output processes are very easy and can be readily realized as follows;

N is the proportion of groundwater occurrence, D is proportion of non-groundwater occurrence area. Correspondingly, the Dis value was computed by Eqs. (15) and (16).

FR ¼ ðA=BÞ=ðC=DÞ ¼ E=F

where K is the proportion of groundwater occurrence that does not occur and H is proportion of non-groundwater occurrence areas in other attributes outside the class. Eqs. (13) and (15) were applied on all the groundwater conditioning factor classes, and then Eqs. (14) and (16) were applied to produce Bel and Dis results. The Unc and Pls values were obtained using Eqs. (17) and (18).

ð11Þ

where A is the area of a class for each groundwater conditioning factor; B is the total area of the each factor; C is the number of groundwater occurrences with low yield in the class area of the factor; D is the number of total groundwater occurrences with low yield in the study area; E is the percentage for area with regard to a class for the factor; F is the percentage for the entire domain; and FR is the ratio of the area where groundwater occurred to the total area, so that a value of 1 is an average value. If the value is greater than 1, it means a higher correlation, and a value lower than 1 means lower correlation (Pradhan and Lee, 2010) 5. Results and discussion 5.1. Application of EBF to groundwater potential mapping Once three mass functions for all input data layers were prepared, Dempster’s rule of combination was applied to obtain four combined mass functions including the belief, disbelief, ignorance, and plausibility functions. Eq. (12) shows the Dempster–Shafer theory direct mass function.

; g  ¼ fTp; Tpg M : 2 ¼ fø; Tp; T p  express the following: where Tp and T p  Tp = class pixels involved by groundwater wells.  = class pixels not involved by groundwater wells.  Tp

ð12Þ

Þ ¼ ½ðNðLÞ  NðL \ EijÞÞ=NðLÞ=½ðNðAÞ  NðLÞ  NðEijÞ kðT p þ NðL \ EijÞ þ NðL \ EijÞÞ=ðNðAÞ  NðLÞÞ ¼ K=H ÞEij Dis ¼ kðT p

.X

ÞEij kðT p

ð15Þ ð16Þ

Unc ¼ 1  Dis  Bel

ð17Þ

pls ¼ 1  Dis

ð18Þ

The values of belief and plausibility range between 0 and 1. According to Park (2011), an important constraint about EBF is that if there is no value for Belief in a certain class, then it indicates that there is no groundwater occurrence in the same class. The estimated EBFs results of three mass functions of belief, disbelief and uncertainty are illustrated in Table 2. The values were derived through the wells with yield value P 11 m3/h. A comparatively high value of bel implies a higher probability of groundwater potential, while a low value of bel indicates a lower probability of groundwater potential. The degree of belief shows a higher value of groundwater potential with lower elevation, mild slope, and flat curvature. As a result, a higher elevation, steeper slope, and convex–concave curvatures could produce a higher rainfall–runoff rate and lower infiltration, thus possibly produce a lower groundwater potential. The previous study by Manap et al. (2012) confirmed that the areas with flat surface are more suitable zones for groundwater exploration.

293

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300 Table 2 The estimated EBF values for conditioning parameters and logistic regression coefficient for significant parameters. Factor

Class

No. of class pixels

No.of yield P11 m3/h pixels

% of yield P11 m3/h pixels

Bel

Dis

Unc

Elevation (m)a

0–20 20–39 39–40 40–59 59–60 60–80 80–141 141–246 246–400 400–1440 0–1 1–3 3–6 6–10 10–14 14–18 18–21 21–25 25–29 29–82 Concave Flat Convex 8.86 to 2.36 2.36–1.65 1.65–2.95 2.95–5.78 5.78–8.62 8.62–10.04 10.04–10.98 10.98–11.81 11.81–12.52 12.52–21.38 13.2 to 8.2 8.2 to 5.1 5.1 to 3.6 3.6 to 2.7 2.7 to 1.9 1.9 to 1.3 1.3 to 0.8 0.8 to 0.06 0.06–0.9 0.9–13.25 0–0.000114 0.000114–0.000305 0.000305–0.000162 0.000467–0.00062 0.00062–0.000792 0.000792–0.001002 0.001002–0.00125 0.00125–0.001545 0.001545–0.001927 0.001927–0.002442 0–0.000052 0.000052–0.000164 0.000164–0.000281 0.000282–0.000392 0.000392–0.000504 0.000504–0.000628 0.000628–0.000785 0.000785–0.000985 0.000985–0.001236 0.001236–0.001675 0.6 to 0.17 0.17 to 0.002 0.002–0.13 0.13–0.21 0.21–0.26 0.26–0.29 0.29–0.32 0.32–0.34 0.34–37

2,477,881 172,438 497,618 319,257 497,704 441,197 430,492 428,265 429,521 422,695 3,601,727 375,683 310,625 334,617 304,669 311,385 235,649 279,597 196,952 166,164 1,053,679 3,965,553 1,097,836 601,340 621,009 630,354 638,769 610,574 583,255 591,805 597,268 577,112 665,582 581,272 632,041 620,965 623,492 634,166 676,363 686,841 602,202 542,601 517,125 2,252,376 1,041,203 1,029,373 771,116 491,426 231,927 147,461 89,479 49,114 13,607 4,163,391 449,805 367,530 330,289 325,684 188,473 138,547 83,966 51,932 17,464 612,309 609,351 635,412 614,554 599,660 580,580 608,795 683,032 580,953

10,400 400 2800 1600 800 800 0 0 0 0 14,000 1200 0 1200 0 0 0 0 400 0 1600 14,000 1200 800 800 0 1200 1200 2800 2800 2800 2800 1600 1200 1200 3200 4400 800 2800 1200 1200 400 400 7600 1600 2000 1600 2400 800 0 400 400 0 14,400 800 0 1200 0 400 0 0 0 0 4400 3200 1600 2000 1600 1600 800 800 800

62 2 17 10 5 5 0 0 0 0 83 7 0 7 0 0 0 0 2 0 10 83 7 5 5 0 7 7 17 17 17 17 10 7 7 19 26 5 17 7 7 2 2 45 10 12 10 14 5 0 2 2 0 86 5 0 7 0 2 0 0 0 0 26 19 10 12 10 10 5 5 5

0.204 0.113 0.273 0.244 0.078 0.088 0.000 0.000 0.000 0.000 0.306 0.252 0.000 0.282 0.000 0.000 0.000 0.000 0.160 0.000 0.247 0.575 0.178 0.047 0.046 0.000 0.067 0.070 0.172 0.169 0.168 0.174 0.086 0.077 0.071 0.192 0.264 0.047 0.154 0.065 0.074 0.027 0.029 0.113 0.051 0.065 0.069 0.163 0.115 0.000 0.150 0.274 0.000 0.315 0.162 0.000 0.331 0.000 0.193 0.000 0.000 0.000 0.000 0.262 0.191 0.091 0.118 0.097 0.100 0.048 0.042 0.050

0.065 0.102 0.092 0.097 0.105 0.104 0.109 0.109 0.109 0.109 0.042 0.102 0.109 0.102 0.109 0.109 0.108 0.108 0.104 0.106 0.405 0.175 0.420 0.106 0.106 0.111 0.104 0.103 0.092 0.092 0.092 0.092 0.102 0.103 0.104 0.090 0.082 0.106 0.094 0.105 0.103 0.107 0.107 0.087 0.109 0.106 0.104 0.093 0.099 0.103 0.099 0.099 0.100 0.046 0.107 0.110 0.102 0.110 0.105 0.106 0.105 0.105 0.104 0.082 0.090 0.101 0.098 0.100 0.100 0.106 0.107 0.105

0.731 0.786 0.634 0.660 0.817 0.808 0.891 0.891 0.891 0.891 0.652 0.646 0.891 0.616 0.891 0.891 0.892 0.892 0.736 0.894 0.348 0.249 0.403 0.847 0.848 0.889 0.829 0.827 0.736 0.738 0.740 0.734 0.813 0.821 0.826 0.718 0.654 0.847 0.752 0.830 0.823 0.865 0.865 0.800 0.839 0.829 0.827 0.743 0.785 0.897 0.751 0.628 0.900 0.639 0.732 0.890 0.567 0.890 0.703 0.894 0.895 0.895 0.896 0.656 0.719 0.808 0.784 0.803 0.800 0.847 0.850 0.845

Slope angle (degree)a

Curvature

Topographic wetness index (TWI)a

Stream power index (SPI)

Drainage density (km/km2)

Lineament density (km/km2)

NDVI

Logistic regression coefficient 9.374

0.27

0.6931

(continued on next page)

294

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Table 2 (continued) Factor

Land use/land covera

Soila

Lithology

Rainfall (mm)

a

a

Class

No. of class pixels

0.37–0.53 Urban area Other crop Rubber Oil palm Grass Forest Swamp Clear land Water bodies Mines area

602,481 1,027,250 497,219 709,056 1,714,702 165,448 1,350,150 277,377 155,956 64,987 92,083

Clay Sandy loam sandy clay Gravel clay-silty Clay–clay Fine sandy clay Fine sandy clay loam to sandy clay–clay Coarse sandy clay–clay Fine sandy clay loam Coarse sandy clay Sandy clay Sand Schist Acid t intermediate Schist, phyllite, slate and limestone Limestone/marble Phyllite, schist and slate Acid intrusives (undifferentiated) Peat, humic clay and silt Phyllite, slate, shale and sandstone Clay, silt, sand and gravel Clay and silt (marine) 2070–2162 2162–2239 2239–2296 2296–2349 2349–2390 2390–2428 2428–2465 2465–2501 2501–2551 2551–2686

No.of yield P11 m3/h pixels

% of yield P11 m3/h pixels

Bel

Dis

Unc

Logistic regression coefficient

0 8000 1600 1600 4000 0 0 800 0 800 0

0 48 10 10 24 0 0 5 0 5 0

0.000 0.253 0.104 0.073 0.075 0.000 0.000 0.093 0.000 0.402 0.000

0.111 0.063 0.098 0.102 0.106 0.103 0.129 0.100 0.102 0.096 0.101

0.889 0.684 0.798 0.825 0.819 0.897 0.871 0.807 0.898 0.502 0.899

7.693 12.99 11.28 2.067 0.000 30.72 2.928 0.000 0.000 0.000

2,168,614 343,039

8000 400

48 2 7

0.075 0.024 0.360

0.081 0.103 0.094

0.843 0.873 0.546

2.131 34.65 14.5

68,824

1200 14

0.135

0.091

0.774

3.282

362,355 271,021 843,144 301,460 1,404,502 201,510 107,187

2400 800 800 0 400 2400 400

5 5 0 2 14 2

0.060 0.019 0.000 0.006 0.245 0.076

0.100 0.111 0.105 0.127 0.089 0.099

0.840 0.870 0.895 0.867 0.667 0.825

55.32 13.26 0.000 6.000 14.8 0.000

132,655 7386 17,047 26,656 1,173,771 2,480,347 619,293 447,393 45,247 1,162,530 599,637 610,656 607,363 620,282 626,851 626,529 618,099 599,190 589,201 619,260

0 0 0 0 7200 1200 3200 1200 800 3200 0 2400 1600 2800 2400 1600 2000 1600 2000 400

0 0 0 0 43 7 19 7 5 19 0 14 10 17 14 10 12 10 12 2

0 0 0 0 0.175 0.014 0.147 0.076 0.510 0.078 0.000 0.143 0.096 0.165 0.140 0.093 0.118 0.097 0.124 0.024

0.101 0.099 0.099 0.099 0.070 0.154 0.089 0.099 0.094 0.098 0.111 0.095 0.100 0.093 0.095 0.101 0.098 0.100 0.097 0.109

0.899 0.901 0.901 0.901 0.756 0.832 0.764 0.825 0.396 0.823 0.889 0.761 0.804 0.742 0.765 0.806 0.784 0.802 0.779 0.868

1.327

Statistically significant parameter.

According to Beven and Kirkby (1979), TWI relates upslope areas as a measure of water flowing towards a certain point (i.e. to the local slope), which is a measure of the subsurface lateral transmissivity. A higher TWI value represents a lower slope and a larger slope area. Therefore, there is a positive correlation between groundwater occurrence and TWI which indicates a higher groundwater potential over an increasing value of TWI. While, there is a negative correlation between SPI and groundwater productivity. The classes between 3.6 and 2.7 had a higher groundwater potential. In the case of river density the results represented the positive relationship with degree of belief between the denser drainage, and the greater groundwater potential. The degree of Bel and Dis values indicated the highest value between 0.001545 and 0.001927 km/km2. Accordingly, topography plays a significant role in hydro-geological systems. For lineament density, it can be seen that as the lineament density increased, the groundwater occurrence generally increased. For lineament density between 0.000281 and 0.000392 km/km2, the bel value, indicated a high probability. However, for lineament density value between 0.0000504 and 0.0000628 km/km2, the bel value was lower which was due to the less number of lineaments present in the study area.

In the case of lithology, the bel value was higher in clay, silt, sand and gravel areas which suggest a higher probability of groundwater occurrence than other lithological units. The probability was lower in acid intrusive (undifferentiated) areas. The igneous rocks are assumed as poor groundwater potential due to difficulty in terms of groundwater movement (Thakur and Raghuwanshi, 2008). For the land use/land cover factor, the water bodies had the highest values indicating the highest bel probability of groundwater potential. It was followed by urban areas, other crops, swamp, oil palm and rubber. In the soil type factor, the highest values were gravel, clay, silty, and sandy clay while the lowest values were observed for the remaining soil type groups. In the case of vegetation index, for NDVI values above 0 the bel values were lower than NDVI values less than 0. This result implied that groundwater potential decreased with the increase of the vegetation index value. As a result it was shown that high rainfall was favorable for high groundwater potential. The integrated results are shown in Fig. 7. The belief map (Fig. 7a) was compared to the disbelief map (Fig. 7b) which showed that belief values were high for areas where disbelief values are low and vice versa. It indicated that high groundwater potential

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

295

Fig. 7. Integrated results of EBF model; (a) belief, (b) disbelief, (c) uncertainty, and (d) plausibility.

was for the areas where there were high degrees of belief and low degree of disbelief for the occurrence. The uncertainty map (Fig. 7c) showed lack of information to provide a real prove for groundwater occurrences. The high uncertainty values were in the areas where belief values were low. The plausibility map (Fig. 7d) shows high values for areas where both belief and uncertainty values are high. Finally, the groundwater potential index (GWPI) was prepared based on the belief function as shown in Fig. 8.

5.2. Application of logistic regression model The multivariate statistical estimation method inquires the relative strength and significance of the variable. Also, LR analysis measured the coefficient of each conditioning factors. Among twelve groundwater conditioning factors, six of them including; elevation, landuse, soil, rainfall, topographic wetness index and slope were chosen because they were statistically significant (Fig. 9) based on the LR model.

296

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Fig. 8. Groundwater productivity potential index map.

For groundwater potential mapping, after calculating the LR coefficients of the six groundwater conditioning factors (Table 2), values were computed in a raster calculator of the ArcGIS software as follows;

Z ¼ ð9:374 elevationÞ þ landuseb þ soilb þ ð1:327 rainfallÞ þ ð0:6931 TWIÞ þ ð0:27 slopeÞ þ 49:07

ð19Þ

where landuseb and soilb are logistic multiple regression coefficient as listed in Table 2. The probability index ranges from 0 to 1. Fig. 10a represents the probability map generated from LR model. This index indicates the predicted probabilities of groundwater occurrence for each pixel in the presence of a given set of conditioning factors. The probability map were grouped into five classes of very high (10%), high (10%), moderate (20%), low (20%) and very low (40%) using quantile

method of classification (Pradhan and Lee, 2010; Pradhan, 2013; Tehrany et al., 2014). The probability map shows that almost 45% and 52% of total area are located in relative high and no well classes respectively. Subsequently, the non-significant groundwater conditioning factors were removed from the EBF and the model was re-run and a new EBF final map was generated. The final map (Fig. 10b) derived from EBF model using six significant parameters, contains 38.5% of the total area, which is allocated to be of high and very high GWPI. Moderate, low and no well GWPI zones constitutes 19%, 13% and 22% of the total area, respectively. 5.3. Validation of the groundwater potential map 5.3.1. Success-rate and prediction-rate Validation is considered to be the most important process of modelling and without validation; the models will have no

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

297

Fig. 9. Statistically significant groundwater conditioning factors.

Fig. 10. Groundwater productivity potential index map from (a) LR method; and (b) EBF from LR method.

scientific significance (Chang-Jo and Fabbri, 2003). By overlaying the groundwater potential maps with the groundwater well locations in the training datasets, cumulative percentage of the groundwater occurrence (starting from the highest to the lowest of GWPI values) were calculated (Fig. 11). The success-rate curves were then obtained by plotting on the x axis (cumulative percentage of potential map) and on the y axis (the cumulative percentage of groundwater occurrence). Evaluation of prediction and success rate is a required outcome for every scheme (Tehrany et al., 2013). Fig. 11 shows the success-rate curve of EBF model using all twelve conditioning factors. The validation of the model indicated that the area under the curve (AUC) was 0.830 which corresponds to 83% of success accuracy. Since the success-rate method used the training groundwater well locations that have already been used during the EBF model building, thus the success-rate is not a

suitable method for assessment of the prediction capability of the model (Pradhan, 2013). However, the success-rate method may help to determine how well the resulting groundwater potential map has classified the areas of existing wells. On the other hand, the prediction-rate described how well the model and predictor variable anticipates the groundwater occurrence (Pradhan, 2013). For quantitative comparison, the areas under the prediction-rate curves (AUC) were calculated. Likewise, the prediction validation was carried out by using the groundwater occurrence dataset that was not used in the training phase (i.e. 18 groundwater well samples). The AUC for the prediction curve of the groundwater potential map was 0.779 (77.9%) which is reasonable for regional groundwater potential mapping.Moreover, Fig. 11 shows the validation of two results obtained using significant conditioning factors; the LR method generated the value of 0.820 for

298

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Fig. 11. The success and prediction rate curve for groundwater potential map; (a) success rate and (b) prediction rate.

Fig. 12. The frequency ratio histogram for low potential wells.

area under the curve which indicates 82% success accuracy. The AUC for prediction of the groundwater potential index was 0.720 implying a perdition accuracy of 72%. In the case of EBF modelling based on using six significant conditioning factors, the area under the curve for success and prediction accuracy were calculated 0.750 and 0.680 respectively. As it can be seen, the success rate accuracy of both LR and EBF are almost similar, with higher accuracy of EBF in prediction using all groundwater conditioning factors. But EBF modelling through significant parameters is not as good as LR result. These results indicate that the EBF through significant parameters is relatively poor estimator compared to the LR model. According to the study of Ozdemir (2011), LR showed poor estimator for spring potential mapping. The performance of the employed EBF model is in agreement with the result obtained by other researchers applied in various environmental and natural hazard studies (Althuwaynee et al., 2014; Bui et al., 2012, 2013; Lee et al., 2013; Mohammady et al., 2012). 5.3.2. Cross-validation using low yield well and FR model Additionally, the groundwater productivity potential index was cross-validated with the aid of low yield wells and using FR approach. The FR approach is based on the observed relationships between the remaining 65 low yield wells <11 m3/h and groundwater conditioning factors. In the histogram of GWPI derived from the FR (Fig. 12) the groundwater locations with low

productivity coincided with the sites falling in the no-well and low classes. This correlation results showed the negative relationship between groundwater well locations and potential area. Relative frequencies of areas affected by different groundwater potential zones were calculated from the ratio. Ideally speaking, the FR value should decrease from a no-well zone to the very high-potential zone. It is necessary to note that the histogram explained how well the EBF model and predictor variables predicted the groundwater potential zones. 6. Conclusion Groundwater is one of the most important natural resources which play an increasingly significant role for water supplies throughout the world. Therefore, detection and prediction of spatial distribution of potential locations for groundwater exploration have become an important topic for private, government, and research institutions worldwide. In this paper, a data driven EBF model was successfully applied to delineate the potential groundwater zones in the Langat basin in Malaysia. The validation results indicated that the success-rate for the EBF model (bel map using 12 conditioning parameters) was 83% with prediction-rate of 78%. In addition, a FR model was applied to cross-verify the results derived from EBF model for low yield wells. In this method, instead of using high yield wells, low yield wells were also used to show where the low groundwater zones are located.

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

To map groundwater potential zones, the primary step was the preparation of the groundwater conditioning factors which affect the groundwater potential. The groundwater conditioning factors were then integrated in a spatial database using EBF model, which indicated the relationship between groundwater yield values and the conditioning factors. Hence, the quantitative relationships between known groundwater occurrence and hydrogeological layers are useful in transforming hydrogeological map data into evidential data layers which can be integrated to produce a predictive groundwater potential map. The main advantage of Dempster–Shafer theory is that, the application of the EBF allows not only the predictive mapping of favorable zones for groundwater occurrence but also allows modelling of the degrees of uncertainty in the prediction. To check the robustness of the proposed EBF model, a well-known LR model was applied using the twelve groundwater conditioning parameters. Subsequently, out of twelve conditioning factors, six of them were selected as significant parameters and coefficient values were assigned to each of them. They were calculated and a final map was produced showing the probability from 0 to 1. In the next step, the EBF model was re-run using those six significant conditioning factors to check the efficiency of the model. In summary, the results of this study proved that EBF model can be successfully applied in groundwater potential mapping. The result obtained in this study might be useful for related agencies in Malaysia for comprehensive evaluation of groundwater exploration development and environmental management for future planning. The proposed method provided rapid, accurate and cost effective results. Furthermore, the analysis may be transferable to other districts with similar topographic and hydro-geological characteristics. Acknowledgments This research was supported by RUGS Grant/05-02-12-2195 RU at the University Putra Malaysia (Vote No. 9376500). The Authors would like to thank Department of Survey and Mapping Malaysia (JUPEM), Minerals and Geosciences Department Malaysia (JMG) and Department of Agriculture for providing various data sets for the research. Thanks to Mahyat Shafapour Tehrany, Mustafa Neamah Jebur and Omar Althuwaynee for their valuable contribution while preparing the manuscript. Thanks to anonymous reviewers for their valuable inputs which were useful to improve the quality of the manuscript. References Akgun, A., 2012. A comparison of landslide susceptibility maps produced by logistic regression, multi-criteria decision, and likelihood ratio methods: a case study at _ Izmir, Turkey. Landslides 9 (1), 93–106. Althuwaynee, O.F., Pradhan, B., Lee, S., 2012. Application of an evidential belief function model in landslide susceptibility mapping. Comput. Geosci. 44, 120– 135. Althuwaynee, O.F., Pradhan, B., Park, H.J., Lee, J.H., 2014. A novel ensemble bivariate statistical evidential belief function with knowledge-based analytical hierarchy process and multivariate statistical logistic regression for landslide susceptibility mapping. Catena 114, 21–36. Ayalew, L., Yamagishi, H., 2005. The application of GIS-based logistic regression for landslide susceptibility mapping in the Kakuda-Yahiko Mountains, Central Japan. Geomorphology 65, 15–31. Ayazi, M.H., Pirasteh, S., Arvin, A.K.P., Pradhan, B., Nikouravan, B., Mansor, S., 2010. Disasters and risk reduction in groundwater: Zagros Mountain Southwest Iran using geoinformatics techniques. Disaster Adv. 3 (1), 51–57. Banks, D., Robins, N.S., Robins, N., 2002. An Introduction to Groundwater in Crystalline Bedrock. Norges Geologiske Undersokelse. Beven, K.J., Kirkby, M.J., 1979. A physically based, variable contributing area model of basin hydrology/Un modèle à base physique de zone d’appel variable de l’hydrologie du bassin versant. Hydrol. Sci. J. 24 (1), 43–69. Bringemeier, 2006. Groundwater exploration adjacent to the Kuala Lumpur International Airport/Malaysia—challenges and chances of exploring a fractured rock aquifer, . Bui, D.T., Pradhan, B., Lofman, O., Revhaug, I., Dick, O.B., 2012. Spatial prediction of landslide hazards in Hoa Binh province (Vietnam): a comparative assessment of

299

the efficacy of evidential belief functions and fuzzy logic models. Catena 96, 28–40. Bui, D.T., Pradhan, B., Revhaug, I., Nguyen, D.B., Pham, H.V., Bui, Q.N., 2013. A novel hybrid evidential belief function-based fuzzy logic model in spatial prediction of rainfall-induced shallow landslides in the Lang Son city area (Vietnam). Geomatics, Nat. Hazard Risk, 1–30. http://dx.doi.org/10.1080/ 19475705.2013.843206 (article online first available). Carranza, E.J.M., Castro, O.T., 2006. Predicting lahar-inundation zones: case study in West Mount Pinatubo, Philippines. Nat. Hazards 37 (3), 331–372. Carranza, E.J.M., Hale, M., 2003. Evidential belief functions for data-driven geologically constrained mapping of gold potential, Baguio district, Philippines. Ore. Geol. Rev. 22 (1), 117–132. Carranza, E.J.M., Woldai, T., Chikambwe, E.M., 2005. Application of data-driven evidential belief functions to prospectivity mapping for aquamarine-bearing pegmatites, Lundazi district, Zambia. Nat. Resour. Res. 14 (1), 47–63. Chang-Jo, F., Fabbri, A.G., 2003. Validation of spatial prediction models for landslide hazard mapping. Nat. Hazards 30 (3), 451–472. Charon, J.E., 1974. Hydrogeological applications of ERTS satellite imagery. In: Proc UN/FAO Regional Seminar on Remote Sensing of Earth Resources and Environment, Cairo. Commonwealth Science Council, pp. 439–456. Chenini, I., Mammou, A.B., 2010. Groundwater recharge study in arid region: an approach using GIS techniques and numerical modelling. Comput. Geosci. 36 (6), 801–817. Chenini, I., Mammou, A.B., Moufida, E.M., 2010. Groundwater recharge zone mapping using GIS-based multi-criteria analysis: a case study in Central Tunisia (Maknassy Basin). Water Resour. Manage 24 (5), 921–939. Chowdhury, A., Jha, M.K., Chowdary, V.M., Mal, B.C., 2009. Integrated remote sensing and GIS-based approach for assessing groundwater potential in West Medinipur district, West Bengal, India. Int. J. Remote Sens. 30 (1), 231–250. Corsini, A., Cervi, F., Ronchetti, F., 2009. Weight of evidence and artificial neural networks for potential groundwater spring mapping: an application to the Mt. Modino area (Northern Apennines, Italy). Geomorphology 111 (1), 79–87. Dar, I.A., Sankar, K., Dar, M.A., 2010. Remote sensing technology and geographic information system modeling: an integrated approach towards the mapping of groundwater potential zones in Hardrock terrain, Mamundiyar basin. J. Hydrol. 394, 285–295. Dempster, A., 2008. Upper and lower probabilities induced by a multivalued mapping. In: Yager, R., Liu, L., Dempster, A.P., Shafer, G. (Eds.), Classic Works of the Dempster-Shafer Theory of Belief Functions. Springer, Berlin, Heidelberg, pp. 57–72 (Chapter 3). Fitts, C.R., 2002. Groundwater Science. Academic Press. Ganapuram, S., Kumar, G.T., Krishna, I.V., Kahya, E., Demirel, M.C., 2009. Mapping of groundwater potential zones in the Musi basin using remote sensing data and GIS. Adv. Eng. Software 40 (7), 506–518. Ghayoumian, J., MohseniSaravi, M., Feiznia, S., Nouri, B., Malekian, A., 2007. Application of GIS techniques to determine areas most suitable for artificial groundwater recharge in a coastal aquifer in southern Iran. J. Asian Earth Sci. 30 (2), 364–374. Gupta, M., Srivastava, P.K., 2010. Integrating GIS and remote sensing for identification of groundwater potential zones in the hilly terrain of Pavagarh, Gujarat, India. Water Int. 35 (2), 233–245. Hosmer, D.W., Lemeshow, S., 2000. Applied Logistic Regression, second ed. Jhon Wiley & Sons Inc., NewYork. Hutchison, C.S., Tan, D.N.K., 2009. Geology of Peninsular Malaysia, University of Malaya. Jha, M.K., Chowdhury, A., Chowdary, V.M., Peiffer, S., 2007. Groundwater management and development by integrated remote sensing and geographic information systems: prospects and constraints. Water Resour. Manage 21 (2), 427–467. Juahir, H., Zain, S.M., Aris, A.Z., Yusof, M.K., Samah, M.A.A., Mokhtar, M., 2010. Hydrological trend analysis due to land use changes at Langat River Basin. Environ. Asia 3 (2010), 20–31. Kavzoglu, T., Sahin, E.K., Colkesen, I., 2013. Landslide susceptibility mapping using GIS-based multi-criteria decision analysis, support vector machines, and logistic regression. Landslides, 1–15. Krishnamurthy, J., Kumar, N.V., Jayaraman, V., Manivel, M., 1996. An approach to demarcate groundwater potential zones through remote sensing and a geographic information system. Int. J. Remote Sens. 17 (10), 1867–1884. 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, 1477–1491. Lee, S., Kim, Y.S., Oh, H.J., 2012. Application of a weights-of-evidence method and GIS to regional groundwater productivity potential mapping. J. Environ. Manage. 96 (1), 91–105. Lee, S., Hwang, J., Park, I., 2013. Application of data-driven evidential belief functions to landslide susceptibility mapping in Jinbu, Korea. Catena 100, 15–30. Madrucci, V., Taioli, F., Cesar de Araujo, C., 2008. Groundwater favourability map using GIS multi criteria data analysis on crystalline terrain, Sao Paulo State, Brazil. J. Hydrol. 357, 153–173. Magesh, N.S., Chandrasekar, N., Soundranayagam, J.P., 2012. Delineation of groundwater potential zones in Theni district, Tamil Nadu, using remote sensing. GIS and MIF techniques. Geosci. Frontiers 3 (2), 189–196. Manap, M.A., Nampak, H., Pradhan, B., Lee, S., Sulaiman, W.N.A., Ramli, M.F., 2012. Application of probabilistic-based frequency ratio model in groundwater potential mapping using remote sensing data and GIS. Arab. J. Geosci., 1–14. http://dx.doi.org/10.1007/s12517-012-0795-z.

300

H. Nampak et al. / Journal of Hydrology 513 (2014) 283–300

Manap, M.A., Sulaiman, W.N.A., Ramli, M.F., Pradhan, B., Surip, N., 2013. A knowledge-driven GIS modeling technique for groundwater potential mapping at the Upper Langat Basin, Malaysia. Arab. J. Geosci. 6 (5), 1621– 1637. http://dx.doi.org/10.1007/s12517-011-0469-2. Mohammady, M., Pourghasemi, H.R., 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. Moon, W.M., 1990. Integration of geophysical and geological data using evidential belief function. IEEE Trans. Geosci. Remote Sens. 28 (4), 711–720. Moore, I.D., Grayson, R.B., Ladson, A.R., 1991. Digital terrain modelling: a review of hydrological, geomorphological, and biological applications. Hydrol. Proc. 5 (1), 3–30. Mukherjee, S., 1996. Targetting saline aquifer by remote sensing and geophysical methods in a part of Hamirpur-Kanpur, India. Hydrol. J. 19, 1867–1884. Murthy, K.S.R., Mamo, A.G., 2009. Multi-criteria decision evaluation in groundwater zones identification in Moyale-Teltele sub-basin, South Ethiopia. Int. J. Remote Sens. 30 (11), 2729–2740. Nag, A., Ghosh, S., Biswas, S., Sarkar, D., Sarkar, P.P., 2012. An image steganography technique using X-box mapping. In: Advances in Engineering, Science and Management (ICAESM), 2012 International Conference on pp. 709–713. IEEE. Neshat, A., Pradhan, B., Pirasteh, S., Shafri, H.Z.M., 2013. Estimating groundwater vulnerability to pollution using a modified DRASTIC model in the Kerman agricultural area, Iran. Environ. Earth Sci. 1–13. http://dx.doi.org/10.1007/ s12665-013-2690-7. Oh, H.J., Kim, Y.S., Choi, J.K., Park, E., Lee, S., 2011. GIS mapping of regional probabilistic groundwater potential in the area of Pohang City, Korea. J. Hydrol. 399 (3), 158–172. Ozdemir, A., 2011. Using a binary logistic regression method and GIS for evaluating and mapping the groundwater spring potential in the Sultan Mountains (Aksehir, Turkey). J. Hydrol. 405 (1), 123–136. Park, N.W., 2011. Application of Dempster–Shafer theory of evidence to GIS-based landslide susceptibility analysis. Environ. Earth Sci. 62 (2), 367–376. Pradhan, B., 2009. Groundwater potential zonation for basaltic watersheds using satellite remote sensing data and GIS techniques. Cent. Eur. J. Geosci. 1 (1), 120– 129. Pradhan, B., 2013. 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. http://dx.doi.org/10.1016/ j.cageo.2012.08.023. Pradhan, B., Abokharima, M.H., Jebur, M.N., Tehrany, M.S., 2014. Land subsidence susceptibility mapping at Kinta Valley (Malaysia) using the evidential belief

function model in GIS. Nat. Hazards, 1–24. http://dx.doi.org/10.1007/s11069014-1128-1. Pradhan, B., Oh, H.J., Buchroithner, M., 2010. Weights-of-evidence model applied to landslide susceptibility mapping in a tropical hilly area. Geomat. Nat. Haz. Risk 1 (3), 199–223. Pradhan, B., Lee, S., 2010. Regional landslide susceptibility analysis using backpropagation neural network model at Cameron Highland, Malaysia. Landslides 7 (1), 13–30. Prasad, R.K., Mondal, N.C., Banerjee, P., Nandakumar, M.V., Singh, V.S., 2008. Deciphering potential groundwater zone in hard rock through the application of GIS. Environ. Geol. 55 (3), 467–475. Sander, P., Chesley, M.M., Minor, T.B., 1996. Groundwater assessment using remote sensing and GIS in a rural groundwater project in Ghana: lessons learned. Hydrogeol. J. 4 (3), 40–49. Saraf, A.K., Choudhury, P.R., 1998. Integrated remote sensing and GIS for groundwater exploration and identification of artificial recharge sites. Int. J. Remote Sens. 19 (10), 1825–1841. Shafer, G., 1976. A Mathematical Theory of Evidence, vol. 1. Princeton University Press, Princeton. Shahid, S., Nath, S.K., Kamal, A.S.M.M., 2002. GIS integration of remote sensing and topographic data using fuzzy logic for ground water assessment in Midnapur District, India. Geocarto Int. 17 (3), 69–74. Singh, A.K., Prakash, S.R., 2002. An integrated approach of remote sensing, geophysics and GIS to evaluation of groundwater potentiality of Ojhala subwatershed, Mirjapur district, UP, India. In: Asian Conference on GIS, GPS, Aerial Photography and Remote Sensing, Bangkok-Thailand. Suratman, S., 2004. IWRM: Managing the Groundwater Component. Paper presented at Malaysia Water Forum, Kuala Lumpur, Malaysia. Tehrany, M.S., Pradhan, B., Jebur, M.N., 2013. Spatial prediction of flood susceptible areas using rule based decision tree (DT) and a novel ensemble bivariate and multivariate statistical models in GIS. J. Hydrol. 504, 69–79. Tehrany, M.S., Pradhan, B., Jebur, M.N., 2014. Flood susceptibility mapping using a novel ensemble weights-of-evidence and support vector machine models in GIS. J. Hydrol. 512, 332–343. http://dx.doi.org/10.1016/j.jhydrol.2014.03.008. Thakur, G.S., Raghuwanshi, R.S., 2008. Perspect and assessment of groundwater resources using remote sensing techniques in and around Choral river basin, Indore and Khargone districts, MP. J. Indian Soc. Remote Sens. 36 (2), 217–225. Umar, Z., Pradhan, B., Ahmad, A., Jebur, M.N., Tehrany, M.S., 2014. Earthquake induced landslide susceptibility mapping using an integrated ensemble frequency ratio and logistic regression models in West Sumatera Province, Indonesia. Catena 118, 124–135. http://dx.doi.org/10.1016/j.catena.2014. 02.005. Yin, E.H., 1976. Geologic map of Selangor, Sheet 94 (Kuala Lumpur). Scale 1, 63360.