Keywords

6.1 Introduction

Flood is the most common among all catastrophic environmental hazards; it causes massive damages to the natural and human resources. In respect to the human society, flood is the most devastating natural hazard because of its irreversible damage capacity to infrastructure, life loss, loss of agricultural, social disruption, etc. (Taylor et al., 2011; Dawod et al., 2011; Swain et al., 2020). The low-lying regions of the world are mostly facing the terrible form of floods due to water logging for incessant rainfall, exhaust of water beyond the carrying capacity of the stream, settlement in the front of swirling water, etc. (Ghosh & Kar, 2018; Natarajan et al., 2021). Globally, on an average, 170 million people are adversely affected by flood every year (Kowalzig, 2008; Das, 2019). In between 1998 and 2017, more than two billion people are affected due to the devastation caused by flood events across the world (World Health Organization, 2018), and 656 million USD economic loss with more than 1.4 lakh human life loss have occurred during this time span (Wallemacq & Rowena, 2018). A total number of 4731 people have lost their lives across the world only in floods occurred in 2016, where average annual death (2006–2015) in flood is 5709 people (Guha-Sapir et al., 2016; Chowdhuri et al., 2020). According to world disaster report (2015), Asia recorded highest number of affected and dead people due to flood around the world in between 2005 and 2014 (World Disasters Report 2015). Most of the developing countries of Asia bear the devastating results of flood, such as China, India, Pakistan, Bangladesh, Sri Lanka, Indonesia, etc. (Wallemacq & Rowena, 2018). In India, every year since 1953 to 2016, 7.19 million hectare area and 31.88 million people on an average are affected by flood. During this time period due to flood India has suffered a loss of 347,581.201 crore rupees and the increasing trend of loss with years is going on (Flood Damage Statistics, 2018). In 2017, a total number of 868 people lost their lives by flood during mid of august (Hindustan Times, 2017a). In the next year (August 2018) a massive flood hit Kerala, where 1.24 million people dislocated and 474 people died (Sphere India, 2018). This terrible scenario of adverse effects of flood does not change very much for West Bengal. The flood damage statistics of 2018 revealed that around 244.23 million people of West Bengal are affected by floods occurs in between 1953 and 2016, and the state has suffered with a loss of 43,954.792 crore rupees (Flood Damage Statistics, 2018).

Floods in India are mainly occurred because of natural (heavy monsoonal rainfall, cyclone, cloud blast, etc.) and manmade (large dam, siltation of river, sand mining, unsystematic management of drainage, etc.) factors (Şen, 2018; Chowdhuri et al., 2020; Sarkar & Mondal, 2020). In India, rural areas are facing huge agricultural loss, occasional loss of livestock and human life, etc. and urban areas are suffering with losses of domestic assets and infrastructures due to inundation of flood plain of the river, costal area, etc. (Şen, 2018). This amount of losses and damages from floods are increasing every year in relation with increasing trends of flood frequency (Flood Damage Statistics, 2018). In the previous two decades, the frequency of floods has increased by 40% worldwide (Hirabayashi et al., 2013; Khosravi et al., 2019). Therefore, impeding an increase in loss and damages by flood are essential for the sake of human society and environment (Jongman, 2018; Genovese & Thaler, 2020). The analysis of flood events is very significant to reduce the potential damages of floods (Huang et al., 2008; Samanta et al., 2018). Pre-flood measure, flood forecasting, post-flood measures are the impost steps of flood analysis (Wanders et al., 2014; Chowdhuri et al., 2020). The first step of flood mitigation is the identification of flood vulnerable zone, which helps to forecasting the flood events (Sanyal & Lu, 2003). Therefore, flood susceptibility mapping is the important part of flood hazard mitigation system, and it is very useful for early warning system, mitigation of future floods events, reducing causalities, etc. (Tehrany et al., 2015; Ali et al., 2019; Sarkar & Mondal, 2020). The susceptibility of flood is determined by several geographical and flood influencing factors, such as geomorphology, geology, rainfall, slope, elevation, land use land cover, drainage density, flow accumulation, etc. (Kazakis et al., 2015; Das, 2018; Hong et al., 2018; Das, 2019). All these factors are analyzed to assess the flood susceptibility zones. Various researchers from different field have adopted different kinds of mathematical and statistical techniques to analyze these factors, such as Analytical Hierarchy Process (AHP) (Das, 2018; Ali et al., 2019; Yousuf Gazi et al., 2019: Jabbar et al., 2019), Frequency Ratio (FR) (Khosravi et al., 2016; Lee et al., 2018; Ali et al., 2019; Liuzzo et al., 2019; Rahman et al., 2019; Sarkar & Mondal, 2020), Logistic Regression (LR) (Pradhan, 2010; Tehrany & Jones, 2017; Lee et al., 2018; Liuzzo et al., 2019; Rahman et al., 2019), Evidential Belief Function (EBF) (Shafapour Tehrany et al., 2019b; Chowdhuri et al., 2020), Artificial Neural Network (ANN) (Jahangir et al., 2019; Rahman et al., 2019), Analytic Network Process (ANP) (Dano et al., 2019), Shannon’s Entropy (SE) (Haghizadeh et al., 2017; Liuzzo et al., 2019), Statistical Index (SI) (Shafapour Tehrany et al., 2019a), Weight of Evidence (WoE) (Khosravi et al., 2016; Rahmati et al., 2016), and Support Vector Machine (SVM) (Tehrany et al., 2015, 2019b), etc. All the models have some advantages and limitations; thus, selection of best model for flood susceptibility is a tough work. The most favorable and preferred multi-criteria decision-making technique is the AHP techniques for flood susceptibility mapping, and a wide number of researchers have already adopted AHP with adequate accuracy (Lawal et al., 2012; Matori et al., 2014; Elkhrachy, 2015; Ghosh & Kar, 2018; Yousuf Gazi et al., 2019; Hammami et al., 2019; Hoque et al., 2019; Jabbar et al., 2019; Phrakonkham et al., 2019; Rahman et al., 2019; Vojtek & Vojteková, 2019; Mishra & Sinha, 2020; Subbarayan & Sivaranjani, 2020; Chakraborty & Mukhopadhyay, 2019). In this technique, the impact of flood provoking factors is assessed by the application of pair-wise comparison matrix and ranked all the factors to delineate priorities zones (Vojtek & Vojteková, 2019; Swain et al., 2020). Enumerating, weight computation, interpolating and integration methods are incorporated in this technique (Chen et al., 2011). In the recent time, the integration of different factor to delineate the susceptible zones of flood is done through the application of Geographical Information System (GIS) and Remote Sensing (RS).

In the recent past, various researchers from different disciplines successfully integrate multi-criteria decision-making tools and GIS for the delineation of flood vulnerable zones, such as Das (2019) worked on flood susceptibility mapping and hydro-geomorphic response to flood. Integrated application of AHP and GIS is used to fulfill the objective; Saha and Agrawal (2020) adopted AHP and GIS technique to assess the flood risk of Prayagraj district, India. In their study, flow accumulation factor is considered as the high influencing factor to food; Sarkar and Mondal (2020) conducted a study on Kulik river basin to delineate the flood vulnerability zone. The FR model has been successfully used to perform the quantification of determination potentiality of causative factors of flood and integrated using GIS platform; Swain et al. (2020) have tried to find out the flood susceptibility zone of a flood vulnerable zone of Bihar using AHP and GIS technique. In their study, a total number of 21 flood causative factors are considered; other studies are Ghosh and Kar (2018); Chowdhuri et al. (2020); Deepak et al. (2020); Natarajan et al. (2021); Malik and Pal (2021); Arya and Singh (2021); etc. So, widely accepted integration approach of AHP and GIS has been used in the present study to find the susceptible zone of flood in Silabati river basin.

In perspective of flood, Silabati river is one of the most vulnerable rivers in south Bengal (Government of West Bengal, 2019). Almost every year overflow of Silabati river causes flood, specifically in the areas of Banka, Khirpai, and Ghatal of lower Silabati river basin (IWAI, 2016; Das et al., 2020). Due to inundation of adjacent areas of Silabati, thousands of people have dislocated and lots of damages have occurred (Hindustan Times, 2017b). In between 1978 and 2016, a total number of 10 high magnitude floods hit the flood plain of Silabati basin (Dandapat & Panda, 2018). During the inundation period, the loss of resources in lower reach of Silabati is maximum (Das et al., 2020). A number of flood hazards have occurred in the recent past years (2011, 2013, 2014, 2015, 2016, and 2017) in Silabati River caused huge damages (NRSC, 2018). The narrow channel of lower Silabati River indulges the overflow of water and causing of embankment breach (Das et al., 2020). This event causing high inundation of area and people suffered with huge agricultural losses, which terrifies the people of several villages (News18, 2019). Additionally, every year some causeways are submerged under the water of Silabati River (Anandabazar Patrika Online, 2020). In this context, flood susceptibility zone delineation and level of vulnerability assessment are very much essential for flood management in Silabati river basin. Considering the flood prone characteristics and adverse effect of flood in Silabati River, the present study was conducted. Thus, the present study tried to find out the flood susceptibility zones of Silabati river basin using AHP and geoinformatics. It will help to provide better forecasting about inundation intensity to the floodplain dwellers and make a positive step in flood management of Silabati river basin.

6.2 Geographical Accounts of Silabati River Basin

The Silabati River is also known as “Silai” which is originated from plateau fringe region of Chota Nagpur plateau and meets with Dwarakeswar River near Ghatal of Paschim Medinipur district, West Bengal. Every year flooding characteristics cannot permit to imagine about the source of Silabati river; it is originated from a pond near Napara village (23°14′09.94″ N and 86°38′42.14″ E) of Puncha community development (C.D.) block of Purulia district. After origin, the river run about 185 km southeasterly and several tributaries, such as Joypanda, Kubai, Tamal, Betal, Parang, Champa, Tarajoli, Mulajor, Purandar and Amoor, etc. enters into the main river channel during it course. The entire Silabati basin covers an area of 4195 km2 and located Purulia, Bankura and West Medinipur districts of West Bengal, India (Fig. 6.1). The geographical extension of the river basin is 22°23′N to 23°15′N and 86°38′E to 87°46′E. Geomorphologically, the basin is located in between Bengal plateau and stable-self part of Bengal basin. Topography of Silabati river basin gradually varies from upper reach (undulating surface of Chota Nagpur plateau) to lower reach (alluvium plain of Ganga delta) (Geological Survey of India, 2001). The isolated hillocks (average 150 m height) are found in the upper part, while isolated highlands are observed at the lower part. The high concentration of lateritic upland of Pleistocene and river flow along this upland formed the famous geomorphosite of West Bengal, i.e., bad land of Gangani (Mandal & Chakrabarty, 2021). The study area experiences sub-humid tropical climatic conditions, and receives annual 100–150 cm rainfall annually (Mahala, 2020). Most of the rainfall is occurred in Monsoon period. Almost in every year, lower reach of the basin encounters flood owing to physiographic characteristics and excessive rainfall over a short time span. It is quite evident from the structures of river bank protection near Talda, Khamardanga, Mathurakata, Gangra, Jhumka and Gugia villages (Bankura district) that the flood occurring nature of the river is very much severe. The important towns within the catchment area are Taldangra and Simlapal of Bankura district; Garbeta, Salboni, and Chandrakona of Paschim Medinipur district.

Fig. 6.1
figure 1

Location map of Silabati river basin. Different symbols used within the basin indicate age of various geological formations. (Source: GSI and Google Earth Image)

Full size image

6.3 Material and Methods

6.3.1 Data Sources

In order to assess the flood risk and vulnerability mapping of Silabati river basin, a total number of 11 flood influencing factors, such as elevation, slope, flow accumulation, distance from river, drainage density, geomorphology, lithology, surface runoff, land use land cover, topographic wetness index, and curvature are taken into consideration. Data set of these selected causative factors is obtained from different types of sources, like Digital Elevation Model (DEM) of Shuttle Radar Topography Mission (SRTM) with spatial resolution of 1 Arc sec (acquired on 23rd September, 2014) downloaded from the United States Geological Survey (USGS) website (https://earthexplorer.usgs.gov); Landsat 8 Operational Land Image (OLI) of 30 m spatial resolution (acquired on 16th May, 2018) obtained from https://earthexplorer.usgs.gov; Topographical Maps (73I/12, 73I/16, 73J/13, 73M/4, 73M/8, 73M/12, 73N/1, 73N/2, 73N/5, 73N/6, 73N/9 and 73N/10) of Survey of India (SOI) on a scale of 1:50,000; District Resource Maps of Geological Survey of India (GSI) on a 1:250,000 scale (Purulia, Bankura and West Medinipur districts); geomorphological map derived from GSI on a scale of 1:250,000; Socio-economic and cultural infrastructures from Google Earth Image; and yearly surface runoff data acquired from https://bhuvan.nrsc.gov.in maintained by National Remote Sensing Centre (NRSC) of Government of India.

6.3.2 Preparation of Geospatial Layers

The flood susceptibility map of Silabati river basin is prepared using the mentioned factors (Fig. 6.2). All these factors are analyzed in geospatial platform, where geospatial layers of all these factors are prepared. The basin map of the study area is delineated from SRTM-DEM using the hydrology tools in ArcGIS software. Before the operation, SRTM-DEM has been projected with Universal Transverse Mercator (UTM), World Geodetic Survey 1984 (WGS-84) 45 North Zone. After the delineation, basin map was rectified with topographical maps. The elevation, slope, flow accumulation, topographic wetness index, curvature and drainage layer of Silabati river basin are also extracted from SRTM-DEM using spatial analyst tools in ArcGIS software. The drainage density map is prepared by using Inverse Distance Weighted (IDW) technique in ArcGIS. In the same way, buffer tools are used to prepare distance from river layer. The land use land cover layer has been prepared from Landsat 8 (OLI) satellite imagery using maximum likelihood algorithm of supervise classification in Erdas Imagine software. Geomorphology and lithology layer of the basin is prepared from geomorphological and geological map of GSI, respectively. The surface runoff layer of Silabati river basin is prepared from runoff data of NRSC and using the interpolation method (IDW) in ArcGIS software. The geomorphological and geological map has been geo-rectified with the UTM, WGS-84, 45N zone and prepared the required maps in ArcGIS 10.3.1 platform. Finally, all thematic layers are classified into five distinct categories following the Jenks natural break method .

Fig. 6.2
figure 2

Flowchart of the methodology adopted in the study

Full size image

6.3.3 Analytical Hierarchy Process (AHP)

All the 11 flood causative factors are interrelated. In the present study, AHP technique is used to quantify the influencing capacity of each factor in flood occurrence through the assigning of weight to each factor (Fig. 6.3). The AHP technique is introduced by Saaty in 1980 (Saaty, 1980). It is a semi-quantitative multi-criteria decision-making approach, in which decisions are made through pair-wise comparison between different factors without inconsistency (Das, 2018; Chowdhuri et al., 2020; Ghosh et al., 2020a; Karmakar et al., 2021). In the field of environmental science several problems have severe complexity and it can be reduced by the application of AHP technique (Kannan, 2010; Çelik, 2019; Arya & Singh, 2021). Another advantage of the technique is the quantification of influencing factor by assigning weight for factor based on the relative importance of factor (Maity & Mandal, 2019; Waris et al., 2019; Ghosh et al., 2020a). In this method, weight assignment is done on the basis of expert judgment of decision makers, and assessment of consistency of judgment value by consistency index (CI) is the strongest part of the method (Das, 2019; Murmu et al., 2019; Karmakar et al., 2021). In this work, the AHP is consisted of two major parts, such as evaluation of causative factors and assign weight and integration of all these factors to compute flood susceptibility index. The evaluation of factors has been done in four steps. In the beginning, weight of each factor is assigned from the literature review, field visits, and expert judgments. Using the nine-point scale of Saaty weight of each factor is assigned based on the relative importance of a factor on the occurrence of flood (Table 6.1). In the next step, comparison of all factors with each other is done by a matrix which is known as pair-wise comparison matrix (Eq. 6.1).

Fig. 6.3
figure 3

Flowchart of analytical hierarchy process

Full size image
Table 6.1 Nine-point scale of Saaty and Random consistency index
Full size table
$$ W=\left|\begin{array}{cccc}{v}_{11}& {v}_{12}& \cdots & {v}_{1n}\\ {}{v}_{21}& v& \cdots & {v}_{2n}\\ {}\cdots & \cdots & \cdots & \cdots \\ {}{v}_{n1}& {v}_{n2}& \cdots & {v}_{nn}\end{array}\right| $$
(6.1)

Here, W is the pair-wise comparison matrix , v is the assign weight of ith factor. In this matrix, assignment of weight depends on the importance of the factor compare to other factor. Thus, the greater influence factor has been assigned to absolute number (Table 6.1) as weight per requirement and less importance factor assigned reciprocal of the absolute number in relation of the greater influence factor. The pair-wise comparison matrix is shown in Table 6.2. After that, estimated Eigen value of each factor is calculated from the matrix table (Table 6.2) by multiplying of all assigned weight of all factors in a row and then the Nth root of the product result is computed by using the Eq. (6.2).

Table 6.2 Pair-wise comparison matrix of causative factors
Full size table
$$ EE=\sqrt[N]{V_a\times {V}_b\times {V}_c\times \dots {V}_N} $$
(6.2)

Here, EE represent the estimated Eigen value of ith factor; Va, Vb Vc Vn is the values of the row value of ith factor and N is the total number of factors. The principle Eigen value of each factor is computed using the Eq. (6.3), where Ev denotes the principle Eigen value.

$$ Ev=\frac{\sqrt[N]{V_a\times {V}_b\times {V}_c\times \dots {V}_N}}{\sum \limits_{i=1}^n EE} $$
(6.3)

After the principle Eigen value (λmax) has been calculated , the consistency index (CI) is calculated using the Eq. (6.4). The CI value is divided by the RI value (Eq. 6.5) to compute the consistency ratio (CR) and verify the judgment coherence. The RI value is varying with number of parameter taken into consideration. In the consistency assessment, CR value must be less than 0.1 for acceptance of the judgment value of factor; if not, the pair-wise comparison matrix has to revise with new judgment values and this process will continue until the CR value comes down to less than 0.1 (Saaty, 1990).

$$ \mathrm{CI}=\frac{\lambda \max -n}{n-1} $$
(6.4)
$$ \mathrm{CR}=\frac{\mathrm{CI}}{\mathrm{RI}} $$
(6.5)

The CR value of the present study is 0.054, which revealed that a consistency is present in the judgment values. Thereafter, weight is assigned to all sub-classes of each factor and the principle Eigen value is computed by considering same procedure of weight assigning .

6.3.4 Flood Susceptibility Index (FSI)

The flood susceptibility index (FSI) is computed using all the selected flood causative factors. The weighted linear combination method is adopted to calculate the FSI. In this method, weight of each factor is multiplied with the rank of sub-classes. The rank is assigned accordingly to sub-classes based on the weightage value of each sub-class of a factor (Table 6.3). Now, the Eq. (6.6) is used to calculate the FSI, where Evi is represented the principle Eigen value of ith factor and rj is the rank of jth sub-classes of respective ith factor.

Table 6.3 Allocation of weight to the selected parameters and their sub-classes
Full size table
$$ \mathrm{FSI}=\sum \limits_{i=1}^n\left({Ev}_i\times {r}_j\right) $$
(6.6)

Finally, FSI value is classified into five distinct classes, very low, low, moderate, high, and very high. Based on this method, all the geospatial layers are integrated in ArcGIS software using the map algebra tool.

6.3.4.1 Accuracy Assessment

Accuracy assessment or validation of model is an integral part of any decision-making model. Without accuracy assessment, any simulation model is incomplete (Chung & Fabbri, 2003; Ghosh et al., 2020b; Sarkar & Mondal, 2020). There are several techniques used by different researchers across the world to validate the simulation models. However, the application of ground level data for accuracy assessment is the most accurate and convenient technique to validate such a model. In case of flood susceptibility assessment, use of historical data of inundated areas due to flood is also evolved as a significant data set for validation of the flood susceptibility map (Pradhan, 2010; Lee et al., 2018; Ali et al., 2019; Das, 2019). In the present study, a total number of 197 known flood sites are demarcated during the flooding season of 2017 using handheld GPS (Garmin eTrex-20) and inundation area identified from the flood map of 2017 provided by NRSC (collected from https://bhuvan.nrsc.gov.in). In this study, the Area Under Curve (AUC) is considered to evaluate the performance and efficiency of the AHP model. To construct the AUC, resultant flood susceptibility map is classified into 100 classes, and cumulative percentage of flood occurrence in different susceptibility classes is also computed. Based on calculation, the AUC is plotted to validate the model (Fig. 6.17).

6.4 Results and Discussion

6.4.1 Elevation

In general, low laying areas are more prone to flood than elevated areas (Nakajima & Umeyama, 2015). Thus, elevation is the prime factor for determining of flood susceptible areas (Das, 2018, 2019; Janizadeh et al., 2019). Flat low land area tends to have higher vulnerability to inundation than the area located at higher elevation (Das, 2018; Ali et al., 2019; Khosravi et al., 2016; Liuzzo et al., 2019; Rahman et al., 2019; Sarkar & Mondal, 2020; Vojtek & Vojteková, 2019). The prepared elevation map of Silabati river basin shows that the elevation varies from 6 to 227 m (Fig. 6.4). The highest elevation is found in the north-western part of the basin, and continuously decreases towards the south-east direction. The category of elevation of 37–66 m dominates the river basin with a share of 29.10% area followed by 6–37 m (27.38%) and 66–96 m (25.49%).

Fig. 6.4
figure 4

Elevation map of Silabati river basin. (Source: SRTM DEM)

Full size image

6.4.2 Slope

Slope refers to the amount of inclination of surface in respect to horizontal plane. This factor plays a crucial role in controlling surface runoff, infiltration process, and sub-surface flow (Ali et al., 2019; Das, 2018; Hammami et al., 2019). An area with gentle slope makes itself more vulnerable during flood as flat terrains are more susceptible to water stagnation compared to the area under steep slope (Hammami et al., 2019; Periyasamy et al., 2018). It is observed from the slope map that the slope ranges from 0° to 28.67° in the study area (Fig. 6.5). An area of 47.63% of the river basin experiences slope in between 0.67° and 1.91° followed by 1.91° to 3.14° (24.17%), and only an area of 1.19% belongs to the slope category of 4.83–28.63°.

Fig. 6.5
figure 5

Distribution of slope in Silabati river basin. (Source: SRTM DEM)

Full size image

6.4.3 Flow Accumulation

Flow accumulation is the concentration of flow in a pixel draining out from neighboring pixels (Das, 2018, 2019; Vojtek & Vojteková, 2019). Generally, flow accumulation is observed to be higher and lower in lower reach and upper reach, respectively, in a basin, since stream order and flow accumulation are positively correlated (Das, 2018). An increase in flow accumulation enhances the propensity of flood risk (Vojtek & Vojteková, 2019). In this study, highest flow accumulation is seen in the lower reach of the basin due to flat terrain and very gentle slope (Fig. 6.6).

Fig. 6.6
figure 6

Flow accumulation map of Silabati river basin. (Source: SRTM DEM)

Full size image

6.4.4 Distance from River

Distance of an area from an active channel is very significant in the field of flood risk mapping (Das, 2018). Areas near the active channels are more vulnerable to flood (Ali et al., 2019; Yousuf Gazi et al., 2019). Based on the drainage network map, a buffer analysis of active channels is carried out with an interval of 0.5 km using proximity analysis tool in ArcGIS software (Das, 2018). A total amount of 27.40% of river basin area lies within 1 km from the active river channel followed by 23.64% area (within 1–2 km), while 48.96% basin area lies greater than 2 km buffer (Fig. 6.7).

Fig. 6.7
figure 7

Distance of basin area from active river channel in Silabati river basin

Full size image

6.4.5 Drainage Density

Drainage density describes the closeness of spacing of stream channels (Ghosh et al., 2020a). It is defined as the total length of streams per unit of area (Kale & Gupta, 2001; Pallard et al., 2009). The area with higher value of drainage density exhibits greater number of flow accumulation paths, and ultimately becomes more susceptible to flood (Chapi et al., 2017; Elkhrachy, 2015; Purnawali et al., 2017; Vojtek & Vojteková, 2019). In Silabati river basin, greater drainage density is observed in lower middle portion and it decreases to all directions of the basin (Fig. 6.8). The major areas lowest drainage density is noticed in the uppermost area of Silabati river basin .

Fig. 6.8
figure 8

Distribution of drainage density in Silabati river basin. (Source: SRTM DEM and Topographical map)

Full size image

6.4.6 Geomorphology

Geomorphology is the study of earth’s landforms. Low-lying flood plains are more susceptible to flood compared to pediment–pediplain complex and denudational hills (Das, 2018; Şen, 2018). The geomorphological map of the study area is extracted in ArcGIS environment using the geomorphological map published by GSI. The map reveals that an area of 82.11% of the total basin is formed by pediment–pediplain complex, while 6.12% and 0.14% of areas are occupied by older flood plain and active flood plain, respectively (Fig. 6.9). Therefore, it is clear that only the lower basin area and areas along the main river channel are characterized with plain lands.

Fig. 6.9
figure 9

Distribution of different geomorphological units in Silabati river basin. (Source: GSI)

Full size image

6.4.7 Lithology

Lithological map is used in flood susceptibility assessment due to diverse sensitivity of lithological units (Tien Bui et al., 2019). An area of hard rock lithology, many a time, is characterized by low drainage density and stream frequency (Kale & Gupta, 2001); hence the area faces less probability of being flooded (Tien Bui et al., 2019). A lithological map of the area is obtained from the District Resource Maps of GSI and processed in ArcGIS software. North-western part of the basin is covered with hard rock lithology, which is comprised of Chota Nagpur gneissic complex, Singbhum groups of rocks and metamorphosed basic rocks. Newer alluvium is found along the Silabati river and south-eastern part of the basin, while the older alluvium occupies around 70% of the total study area (Fig. 6.10).

Fig. 6.10
figure 10

Lithological map of Silabati river basin. (Source: GSI and https://bhuvan.nrsc.gov.in)

Full size image

6.4.8 Surface Runoff

During precipitation , the water drains over the land surface following the local gradient and gravity. This process of water flow is called surface runoff (Uzor-Totty & Lawal, 2019). The interaction between precipitation and surface runoff depends on time and space. Surface runoff is influenced by both climatic factors (precipitation type, rainfall amount, intensity, duration, distribution, soil moisture resulting from earlier precipitation, direction of storm movement, evaporation, relative humidity, and seasons) and physical factors (vegetation, soil, elevation, slope, drainage area, basin shape, drainage network) (Uzor-Totty & Lawal, 2019). Moreover, anthropogenic activities, such as urbanization and other constructional activities, and its resultant impervious surfaces reduce the infiltration rate, increase surface runoff, and shorten runoff time into streams. Finally, all these increase the chance of higher magnitude and frequency of flood in nearby streams (Uzor-Totty & Lawal, 2019). To prepare a surface runoff map, maximum surface runoff of 4 years (2016–2019) is obtained from NRSC and interpolated using the Inverse Distance Weighted (IDW) method in ArcGIS software. At the end, the resultant map has been classified into five zones based on Jenks natural break method. The higher amount of surface runoff is recorded in the lower middle and lower reach of the basin, mainly in interfluves zones (Fig. 6.11).

Fig. 6.11
figure 11

Surface runoff map of Silabati river basin. (Source: https://bhuvan.nrsc.gov.in)

Full size image

6.4.9 Topographic Wetness Index

Topographic wetness index is commonly used to evaluate the influence of topography in accumulation of flow or generation of runoff at any point of the basin area (Ali et al., 2019; Das, 2018, 2019; Moore et al., 1991; Sarkar & Mondal, 2020; Sørensen et al., 2006). It is expressed as TWI =  ln (a/ tan B); where “TWI” refers to topographic wetness index, “a” is the specific basin area, and “B” is the local slope (Das, 2018, 2019; Sørensen et al., 2006; Tien Bui et al., 2019). The area with higher TWI value indicates high potentiality of flood event (Das, 2018; Tien Bui et al., 2019). It is evident that the higher TWI value is found in active flood plain region of Silabati river basin because of lower elevation of this area (Fig. 6.12).

Fig. 6.12
figure 12

Topographic wetness index map of Silabati river basin. (Source: SRTM DEM)

Full size image

6.4.9.1 Land Use Land Cover

Land use land cover has a significant role in determining surface runoff, which is directly related to flood event in catchment area (Phrakonkham et al., 2019). An area covered with vegetation reduces the intensity of surface runoff, and enhances the proliferation of infiltration process; whereas build-up area strongly impedes water percolation into the ground and hastens the surface flow (Das, 2018; Hammami et al., 2019; Roslee et al., 2018; Samanta et al., 2018; Sarkar & Mondal, 2020). To perceive the nature of land use in the study area, a supervised classification is done adopting maximum likelihood method in Erdas Imagine software (Das, 2019; Ghosh et al., 2020a; Ghosh et al., 2020b; Sarkar & Mondal, 2020). The land use pattern of Silabati river basin is dominated by agricultural land (76.79%), while forest cover is the second highest land cover occupying 20% of the total basin area. Here, Sal (Shorea Robusta), Shishu (Dalbergia), Palas (Butea Monosperma), Arjuna (Terminalia Arjuna), Sonajhuri (Acacia Auriculiformis) and Eucalyptus (Eucalyptus Globules), etc., plants are commonly found. Only 2.11% of the basin is covered by build-up area, while the water body is recorded in a tiny amount , i.e., 0.42% (Fig. 6.13).

Fig. 6.13
figure 13

Distribution of different types of land use land cover in Silabati river basin. (Source: Landsat 8 OLI Satellite Image)

Full size image

6.4.9.2 Curvature

Curvature determines surface flow and infiltration process, and ultimately influences the incidence of flood (Das, 2018, 2019). It can be three types: (1) concave (positive curvature), (2) flat (zero curvature), and (3) convex (negative curvature). There are different opinions about the role of curvature in controlling the surface flow. Young and Mutchler (1969) advocated that a convex slope can produce much more runoff than the concave slope (Chapi et al., 2017; Young & Mutchler, 1969). Hudson and Kesel (2000) demonstrated that the area with the curvature value in between 1.0 and 2.0 is more vulnerable to flood (Das, 2019; Hudson & Kasel, 2000). The curvature category of −0.11 to 0.11 possesses an area of 45.72% followed by 0.11 to 0.33 (23.93%) and −0.33 to −0.11 (17.34%). It is the −2.37 to −0.33 category that shares only 2.89% of the basin area (Fig. 6.14).

Fig. 6.14
figure 14

Curvature map of Silabati river basin . (Source: SRTM DEM)

Full size image

6.4.10 Assessment of Flood Susceptibility

The final susceptibility map is generated based on integration of factor weight and sub-class rank using raster calculator in ArcGIS software. The derived map is categorized into five distinctive classes by Jenks natural break method. The classes are very high, high, moderate, low, and very low susceptibility of flood, and these classes cover 14.04%, 20.67%, 21.76%, 20.69%, and 22.84% of the total basin area, respectively (Fig. 6.15). The lower segment of the river basin is observed to be in vulnerable condition, since it belongs to the very high to high flood susceptible category. It is because the area has low elevation (6–30 m) and low slope (<20). This portion is distinguished by flat alluvial plain and well developed flood plain. The north-western part of the basin comes under very low to low flood susceptibility zone as the elevation is observed to be high (>100 m), and slope is seen more than 20. In this section, the river Silabati, flowing through the undulating topography of Chota Nagpur plateau, is characterized by narrow restricted valley, lower width-depth ratio, lacking of flood plain development, rough bed configuration and fewer number of bars; whereas the middle reach of the basin with moderate elevation shows moderate vulnerability to flood. It is observed that flood susceptibility varies from higher in south-east to lower in north-west direction (Fig. 6.15). Thus, the susceptibility of flood decreases as the elevation increases towards north-west direction.

Fig. 6.15
figure 15

Flood susceptibility map of Silabati river basin. The red dots in the map indicate flood points verified during 2017 flood

Full size image

The degree of flood vulnerability is greatly influenced by social and economic factors of an area (Działek et al., 2019). Very high and high flood susceptibility zones of the study area are mainly located at Keshpur, Ghatal, Chandrakona-I, Chandrakona-II and Daspur-I Community Development (C.D.) blocks of West Medinipur district (Fig. 6.15). Moreover, Garbeta-I, Garbeta-II, Garbeta-III, Salbani, and Medinipur C.D. blocks of West Medinipur district are also affected by flood. The population density and flood vulnerability are positively correlated to each other (Hoque et al., 2019; Sarkar & Mondal, 2020). These C.D. blocks have a population density ranging from 340 persons/km2 to 1212 persons/km2. High population density is found in Daspur-I (1212 persons/km2), Ghatal (1016 persons/km2), Chandrakona-II (819 persons/km2), Chandrakona-I (702 persons/km2), and Keshpur (702 persons/km2). The Decadal (2001–2011) population growth rate varies from 13.20 to 21.37% in these blocks (Census of India, 2011). Therefore, a large number of people can be affected significantly by the floods. The areas with large number of households are more at-risk during occurrence of flood (Sarkar & Mondal, 2020). The household frequency ranges in between 71 and 261 households/km2 in the studied C.D. blocks (Census of India, 2011). It is reported that the higher number of household frequency is recorded in Daspur-I (262 households/km2), Ghatal (215 households/km2), Chandrakona-II (175 households/km2), Chandrakona-I (157 households/km2), and Keshpur (142 households/km2) (Census of India, 2011). Another important aspect needs to be mentioned here that, the flood vulnerability of an area gets momentum owing to its nature of house type (Hoque et al., 2019; Sarkar & Mondal, 2020). It is noticed that the percentage of permanent, semi-permanent, and temporary houses in these C.D. blocks varies from 6.39% to 37.95%, 40.76% to 78.83%, and 8.29% to 35.38%, respectively. The large number of temporary houses is located at Medinipur (35.38%) and Keshpur (18.26%) C.D. blocks; there are significant numbers of semi-permanent houses in Chandrakona-I (78.83%), Chandrakona-II (77.92%), Keshpur (75.17%), Garbeta-II (72.05%), and Garbeta-III (71.46%) (Census of India, 2011). During flood event, children and females become more vulnerable because of their incapacitation to move at once during the emergency evacuation situations (Hoque et al., 2019). The percentage of females and children among the studied C.D. blocks ranges from 48.44% to 49.53% and 11.08% to 13.52%, respectively (Census of India, 2011). There is a negative relation between literacy rate and flood vulnerability (Hoque et al., 2019), and the average literacy rate of these C.D. blocks is calculated to be 67.28% (Census of India, 2011). The entire study area is agriculturally dependent, and flood incident brings the crop production to an end pushing thousands of men engaged in cultivation works in distress conditions (Hoque et al., 2019). It is to be noted that the agriculturally dependent populations of these C.D. blocks vary in between 59.90 and 83.67%, where the higher number of agriculturally dependent population is found in the C.D. blocks of Garbeta-II (83.67%), Chandrakona-II (83.55%), Keshpur (79.88%), Chandrakona-I (76.80%), Garbeta-I (76.37%), and Salbani (75.31%) (Census of India, 2011). There are several permanent and temporary bridges over Silabati river, and roads stretch for a considerable length within very high to moderate flood susceptibility zones (Figs. 6.16 and 6.17). Hence, it is clear that a large number of people along with many socio-economic infrastructures can significantly be affected by the flood events of Silabati river basin.

Fig. 6.16
figure 16

Distribution of different socio-economic and cultural structures present within Silabati river basin overlie flood susceptibility map. (Source: Google Earth Image)

Full size image
Fig. 6.17
figure 17

(a) Flood water spilling over the road near Bhelaidiha, Bankura on 25th July, 2017; (b) Flood situation near Simlapal, Bankura on 25th July, 2017 that submerged adjacent road; (c and d) Bank protection structures near Jhumka village, Bankura constructed by Irrigation & Waterways Department, Kangsabati Circle, Bankura to reduce vulnerability of river bank failure during high discharge events or floods

Full size image

The geospatial modeling provided an effective way of flood management in the study area. Various spatial components of flood are identified and the extent of potential flooded area is quantified as well. It will facilitate the implementation of evacuation strategy, rehabilitation plan, and damage assessment during critical flood situation. It may also be effective in the development of policy, guidelines, and recommendation of land use planning.

6.4.11 Validation of the Flood Susceptibility Map

In order to assess the validity of the flood susceptibility map, the AUC is computed, and the output value, i.e., 76.41% depicts that the accuracy level of the flood map prepared adopting AHP technique is well acceptable (Fig. 6.18). The value of AUC less than 50% is considered as inappropriate for flood vulnerability mapping, while the calculated value of AUC exceeding 75% is well accepted for the predicted model (Akgun et al., 2008; Egan, 1975; Ozdemir & Altural, 2013; Pedzisai, 2010; Saha, 2017; Sarkar & Mondal, 2020).

Fig. 6.18
figure 18

Area under curve for validation of flood susceptibility map of Silabati river basin

Full size image

6.5 Conclusion

To analyze and perceive the flood vulnerability of the Silabati river, three factors, such as elevation, slope, and flow accumulation turn out to be most dominating parameters out of adopted 11 factors considered in the present study. It is found that an area of 14.04 and 20.67% of the total basin comes under very high and high flood susceptibility zones, respectively, and most of this area lies in West Medinipur district. Amongst all C.D. blocks of the West Medinipur district, only ten are more or less affected by the flood events, but five C.D. blocks, namely: Keshpur, Ghatal, Chandrakona-I, Chandrakona-II, and Daspur-I are severely influenced by flood due to flat topography, very low elevation, and higher amount of flow accumulation compared to other C.D. blocks. As a result, a large number of populations have been witnessing and suffering from floods almost in every year for decades. Their misery is the consequences of losing shelters, crops, and cattle. To minimize the consequences of floods in Silabati river basin, implementation of different management strategies in higher flood susceptible areas is very much required. This research work may help the policy makers and implementing authorities to gather basic information related to the flood, including its vastness and areas under risk. It is suggested that C.D block level assessment of flood susceptibility can give accurate flood potentiality at local scale. Moreover, a flood simulation study can be helpful to quantify depth, velocity, and duration of flood and to access the surface water and groundwater interaction during flood.