Content uploaded by Sukanta Malakar
Author content
All content in this area was uploaded by Sukanta Malakar on Sep 24, 2024
Content may be subject to copyright.
Seismic risk assessment using integrated MCDM method in West
Bengal, India
Monalisa Malakar
a
, Sukanta Malakar
b,c,*
, Mohd Sayeed Ul Hasan
c,d
, Abhishek K. Rai
c
,
Vijay K. Kannaujiya
c
a
Department of Civil Engineering, University Institute of Technology, Burdwan, India
b
Department of Geography, Adamas University, Kolkata, India
c
Centre for Ocean, River, Atmosphere and Land Sciences, Indian Institute of Technology Kharagpur, India
d
Department of Civil Engineering, Aliah University, Kolkata, India
ARTICLE INFO
Keywords:
AHP-Entropy
Vulnerability
Hazard
Coping capacity
Seismic risk
West Bengal
ABSTRACT
Earthquakes are one of the natural hazards that threaten human lives and properties. Consequently, seismic risk
assessment plays a signicant role in disaster mitigation. This study estimates seismic risk in West Bengal, India,
by integrating the two multi-criteria decision-making (MCDM) models: analytical hierarchy process (AHP) and
Entropy. Integrated AHP-Entropy is used to determine vulnerability, seismic hazard, and coping capacity. The
seismic risk was then assessed by integrating the thematic information of vulnerability, seismic hazard, and
coping capacity. The results show that about 19% of the total area and 70% of the total population in West
Bengal may be at very high seismic risk. The result is validated through a receiver operating characteristic curve,
displaying satisfactory performance in seismic risk estimation. The ndings of this study may help governmental
agencies identify seismic-risk zones and establish seismic hazard plans in advance against any potential threat in
the study region.
1. Introduction
Earthquakes are natural disasters that can result in severe economic,
social, and physical devastation that can persist for years. Estimates
suggest that 1.87 million people died globally due to earthquakes in the
twentieth century, with an average of 2052 fatalities per event between
1990 and 2010 (Wisner et al., 2008). Also, from 1994 to 2013, earth-
quakes were the most frequent natural hazard after landslides and
ooding, resulting in $787 billion in nancial losses (Centre for
Research on the Epidemiology of Disasters, 2015). The data also shows
that out of the total number of deaths that occurred due to different
natural disasters, earthquakes and related events, including tsunamis,
may have caused 55% of them worldwide over the past two decades
(Alexander, 2017). As per the projections made by Dunbar et al. (2003),
the potential Indian economic loss resulting from an earthquake could
reach up to 650 million US dollars within the next 50 years, with a 10%
probability of occurrence. However, proper urban planning and imple-
mentation of disaster mitigation strategies may safeguard people and
property from potential events. However, in developing countries, rapid
unplanned urbanization, unregulated migration, and inadequate land
use management pose signicant threats to the effective implementation
of disaster mitigation strategies in the context of seismic hazards. Thus,
comprehensive research on earthquakes and associated hazards is
required to identify seismically risk zones for disaster mitigation.
Seismic hazards are primarily estimated using conventional seis-
mological techniques that include deterministic and probabilistic
seismic hazard models. Mohanty and Walling (2008) estimated the
seismic hazard of Kolkata metropolitan city based on the maximum
expected magnitude, which was calculated by a quasi-probabilistic
method. A logic tree-based synoptic probabilistic seismic hazard
model (PSHM) for Kolkata has been developed by Nath et al. (2014).
Maiti et al. (2017) developed PSHM for West Bengal based on improved
seismogenic source characterization. However, recent advances in
geospatial and multi-criteria decision-making (MCDM) techniques have
facilitated the ability to construct models for assessing seismic hazard,
vulnerability, and risk by considering multiple inuential criteria over a
large scale. The Analytical Hierarchy Process (AHP) with Geographic
Information Systems (GIS) has been used widely to understand the
* Corresponding author. Department of Geography, Adamas University, Kolkata, India.
E-mail address: malakarsukanta031@gmail.com (S. Malakar).
Contents lists available at ScienceDirect
Evolving Earth
journal homepage: www.sciencedirect.com/journal/evolving-earth
https://doi.org/10.1016/j.eve.2024.100036
Received 4 July 2024; Received in revised form 18 September 2024; Accepted 18 September 2024
Evolving Earth 2 (2024) 100036
Available online 19 September 2024
2950-1172/© 2024 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC license ( http://creativecommons.org/licenses/by-
nc/4.0/ ).
seismic vulnerability in different study areas (Karaman and Erden, 2014;
Panahi et al., 2014; Alizadeh et al., 2018; Flores et al., 2021; Rai et al.,
2023). Nath et al. (2015) and Sinha et al. (2016) used AHP to assess the
seismic vulnerability and risk in the Kolkata and Delhi regions. Pal et al.
(2008) used the fuzzy-AHP technique to analyze seismic hazards in the
Sikkim Himalayas. The Ordered Weighted Average method has been
applied by Martins et al. (2012) to calculate the social vulnerability to
seismic risk in Vila Franca do Campo, Portugal. Jena et al. (2020)
assessed the seismic vulnerability of the Banda Aceh region using AHP
integrated with Criteriums Optimizacija I Kompromisno Resenje
(VIKOR). Nyimbili et al. (2018) estimated seismic hazards by integrating
AHP and the Technique for Order Preference by Similarity to the Ideal
Solution (TOPSIS). Alam and Haque (2021) used AHP and Weighted
Linear Combination integrated methodology to assess the seismic
vulnerability of Mymensingh, Bangladesh. The Himalayan region’s
seismic vulnerability has been assessed using the integration of AHP and
Grey Relational Analysis (GRA), AHP and VIKOR, fuzzy-AHP and
fuzzy-TOPSIS methods (Malakar and Rai, 2022a, 2022b, Malakar and
Rai, 2022a, 2022b).
Several studies have been conducted in West Bengal through the
geospatial technique for groundwater, ood, landslide susceptibility and
other environmental applications (Ramachandra et al., 2014; Hasan and
Rai, 2020; Ray et al., 2023; Molla et al., 2023; Kannaujiya et al., 2024).
However, the seismic risk assessment in the study area has not been
found through geospatial techniques. This paper presents a compre-
hensive analysis of the seismic risk in West Bengal, which will assist in
developing mitigation strategies. However, acquiring datasets, espe-
cially for a larger area, is complicated and challenging, so studies focus
on smaller areas except a few. It is also observed from the literature that
several studies have used integrated MCDM models to estimate seismic
hazard and vulnerability, and the widely used one is AHP. AHP provides
parameter weights based on experts’ opinions and literature published.
However, while calculating weights, experts may ignore data informa-
tion, leading to result uncertainty (Bhattacharya et al., 2010; Rodcha
et al., 2019). As discussed by many researchers, this issue may be
overcome with integrated MCDM models (Jena et al., 2020).
This work will estimate the seismic risk in the West Bengal research
area using publicly accessible datasets and the AHP-Entropy integrated
method and validate the result. The study helps us identify the most
seismic risk zones, which can help reduce earthquake damage and
casualties.
2. Study area
With 88,752 km
2
and over 91 million people, West Bengal is located
in the Surma Valley in the Himalayan foothills and western foreland of
the Assam-Arakan Orogenic Belt (Fig. 1). The northern section was
formed by Precambrian rocks from the foothills of the Himalayas and
elevated Shillong plateau (Maiti et al., 2017). Precambrian Indian shield
rocks can be found in the western study area. Bureau of Indian Standards
(2002) established that the majority of the study region lies within
seismic zone II–V and peak ground acceleration (PGA) ranging from 0.1
to 0.36 g. Eocene Hinge Zone (EHZ), Pingla Fault, Rajmahal Fault,
Garhmoyna-Khandaghosh Fault, Malda-Kishanganj Fault, Sainthia
Bahmani Fault, Debagram Bogra Fault, and Jangipur-Gaibandha Fault
are important tectonic features in Bengal Basin. The 25 km wide and 4.5
km deep EHZ is a major tectonic structure that trends NE-SW. The 1964
Sagar Island earthquake was triggered by the NNE-trending fault reac-
tivation across the EHZ, which featured primarily a thrust focal mech-
anism. Among the major earthquakes that have struck the region are the
1897 Shillong earthquake, the 1934 Bihar-Nepal earthquake, and the
1950 Assam earthquake. Signicant casualties were suffered by the
Fig. 1. Location of the study area. Earthquake locations (M >4) with population density, lithology, and districts are shown.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
2
1934 Bihar-Nepal VI-VII intensity earthquake (GSI, 1939). Thus, seismic
risk zones must be identied to reduce future earthquake casualties in
the study area.
3. Methods and materials
3.1. Data sources
The datasets used for this analysis were compiled from sources that
were freely accessible to the public. The Registrar General & Census
Commissioner of India provided the relevant social information,
including population and literacy rate. Information about active faults
was acquired from Bhukosh, Geological Survey of India (https://bhuko
sh.gsi.gov.in/Bhukosh/Public). The earthquake catalogue has been ac-
quired from the National Center for Seismology, India (https://seismo.
gov.in/). This study also used the United States Geological Survey
time-averaged shear-wave velocity up to 30 m depth (Vs30) (https://e
arthquake.usgs.gov/data/vs30/) and Oak Ridge National Laboratory
Distributed Active Archive Center sediment thickness (Pelletier et al.,
2016). Secondary parameters, including communication networks, were
acquired from DIVA-GIS, whereas other vulnerability parameters were
gathered from OpenStreetMap. OpenTopography provided the SRTM
elevation data (SRTM, 2013), and Universitat Hamburg provided li-
thology (Gleeson et al., 2011), which was also used. Caprio et al. (2015)
resulted in a global peak ground acceleration (PGA) relationship, which
was applied to estimate peak ground acceleration. The algorithms used
to estimate and classify layers included Euclidean distance, inverse
distance weighting, Kernel density, and quantile classication. The in-
uence of the individual parameters used in this study has been
explained in detail by many researchers, including Jena et al. (2021),
Malakar et al. (2023), and Zhu et al. (2023). The datasets were last
assessed on July 27, 2023. The detailed methodology is explained in the
following subsection.
3.2. AHP-entropy integration
The Analytic Hierarchy Process (AHP) resulted from Saaty (1980) as
a widely used MCDM model. Through pairwise comparison matrices,
AHP uses a hierarchical framework to calculate the criteria’ priority in
the decision-making problem. These matrices predominantly rely on
expert opinion and published literature. At rst, this method uses the
criterion scores to create the pairwise comparison matrix. Expert
knowledge and information accumulated from the published literature
determine the criteria score ranging between 1 and 9 (Saaty, 1980).
After that, the matrix is normalized, and the criterion weights are
calculated. The obtained output was checked for consistency using the
consistency ratio (CR). The CR is formulated as follows:
CI =λmax −n
n−1and CR =CI
RI (1)
Here, RI is a randomness indicator, n is the number of criteria utilized,
and λmax indicates the principal eigenvalue. Saaty (1980) predicts the
value of RI for matrices with dimensions between 1 and 15. The level of
consistency is acceptable for determining the priority wʹ
iif CR <0.1.
Entropy, on the other hand, uses probability theory to assess the
degree of disorder or uncertainty in a system (Shannon, 1948). Entropy
states that a higher-weight index value is more valuable than a lower
one.
At first,the matrix was normalized as pij =xij
m
i=1
xij
(2)
where xij is the performance measure of the jth attribute in the ith
alternative, and pij is the normalized value. Then, the entropy value is
estimated as:
Ej= − 1
ln(m)
m
i=1
lnpij*pij (3)
The entropy weight wʹʹ
iis computed as:
wʹʹ
i=1−Ej
n
j=11−Ej
(4)
The degree of diversication factor 1−Ejdescribes each criterion’s
intrinsic information divergence. The entropy weight indicates the cri-
terion’s decision-making importance.
The analytical hierarchy process is a subjective model, whereas the
entropy is an objective MCDM model. The AHP is used to evaluate expert
systems quantitatively through a hierarchical structure, and the entropy
weight method is introduced to reduce the negative effect of individual
subjective evaluation bias on the accuracy of comprehensive evaluation
(Dyer and Forman, 1992). The AHP calculations are based on available
literature and expert opinion, and experts usually ignore data informa-
tion (Bhattacharya et al., 2010; Rodcha et al., 2019); in contrast, the
entropy calculations are entirely based on the data, and the weights are
sometimes different from reality (Wang et al., 2009; Cui et al., 2018),
resulting in uncertainty in the result. So, to estimate the robust weight of
the criteria, we have integrated the subjective AHP weight (wʹ
iand the
objective entropy weight (wʹʹ
i) as suggested by Chuansheng et al. (2012),
to establish a new weight by:
ω
i=
α
wʹ
i+ (1−
α
)wʹʹ
i(5)
The
α
can range from 0 to 1. Here, we used 0.6 as the value of
α
.
(Chuansheng et al., 2012). The integrated AHP-entropy through GIS has
been applied to estimate vulnerability, seismic hazard, and coping ca-
pacity for the study region (Fig. 2). Tables 1–3show the comparison
matrix and predicted overall weights.
3.3. Seismic risk estimation
The seismic risk was calculated by the spatial data of the region’s
seismic hazard, vulnerability, and coping capacity (Jena et al., 2021;
Malakar et al., 2023). World Health Organization (2009) has given the
equation for estimating the risk, which is:
Risk =(Vulnerability*Hazard)
Coping Capacity (6)
4. Results
This study proposes a method for assessing seismic risk that in-
tegrates the AHP and entropy MCDM models. The rationale for inte-
grating two MCDM models was covered in previous sections. The
seismic risk was estimated using 24 criteria that may inuence the re-
gion’s seismicity, and the World Health Organization (2009) recom-
mended formulation was used.
4.1. Vulnerability
The vulnerability parameter weights were calculated using AHP-
Entropy integration (Table 1). Our result shows that the northern and
eastern parts of West Bengal are relatively highly vulnerable (Fig. 3).
Several major metropolitan cities in the study area are also found to be
in very high vulnerability zones. Numerically, it is found that 19.41% of
the study area may lie under a very low vulnerable zone, 20.06% low,
19.76% moderate, 21.42% high, and 19.35% very high vulnerable zone.
12.04% of the population may live in very low vulnerable regions,
7.11% low, 11.01% moderate, and approximately 70% may reside in
high to very high vulnerable zones (Table 4). Consequently, more than
two-thirds of the studied area is in a highly vulnerable zone, and
M. Malakar et al.
Evolving Earth 2 (2024) 100036
3
authorities should prioritize the population in the vulnerable regions.
4.2. Seismic hazard
The seismic hazard was estimated using historical seismicity, active
faults, PGA, Vs30, sediment thickness, lithology, and elevation. The
parameter weights are presented in Table 2. The results show that the
northern region of the study area exhibits a comparatively high level of
seismic hazard. A major part of the eastern and western portions of the
studied region also lies in the high seismic hazardous zone (Fig. 4). The
result also indicates that 19.95% of the area may lie under a very low
hazard zone, 21.41% low, 21.24% moderate, 21.40% high, and 15.99%
very high. However, 6.81% of the population may live in the very low
hazard zone, 43.49% low, 22.33% moderate, 12.70% high, and 14.66%
very high (Table 4). These show that approximately 28% of the popu-
lation may live in highly seismic hazard zones, a cause for concern for
organizations engaged in disaster mitigation.
4.3. Coping capacity
During a seismic event, coping capacity is essential, as it is usually
assumed that an educated population can effectively deal with hazards.
The coping capacity is a continuous process that includes proper
training, awareness, and resource management. This study determined
the area’s coping capacity by analyzing the hospital accessibility,
educated people, service centers, and communication networks. The
AHP-Entropy method is applied to compute the parameter weights uti-
lized to assess the coping capacity for the study region (Table 3). Our
result shows that the southern and western parts of West Bengal have a
relatively low coping capacity (Fig. 5). The eastern part of West Bengal
includes certain prominent cities that lie under high coping capacity.
Notably, 0.77% of the study area population may reside under very low
coping capacity, 2.88% low, 4.39% moderate, 13.94% high, and 78.01%
very high. Government and private agencies should focus on hospital
development, which can be easily accessible, especially for the rural
population, with an extensive network of communication and educate
and train people to handle potentially hazardous circumstances, which
is essential for development and disaster prevention.
4.4. Seismic risk
The seismic risk map is assessed using seismic hazard, vulnerability,
and coping capacity spatial data (Fig. 6). The map is divided into ve
classes: red for high-risk and green for low-risk. According to our nd-
ings, a number of cities in the study area are in very high-risk zones. The
northern and eastern parts of West Bengal, including Kolkata, are rela-
tively at high seismic risk. The study found that 18.59% of the region
may have very low seismic risk, 20.91% low, 21.81% moderate, 19.92%
Fig. 2. A framework of the adopted seismic risk estimation methodology for the study region.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
4
high, and 18.77% very high. The result also shows that 2.52% of the
total population may live in very low seismic risk zones, 5.01% in low
seismic risk zones, 8.06% in moderate seismic risk zones, 14.31% in high
seismic risk zones, and 70.10% may be in very high seismic risk zones
(Table 4).
5. Discussions
High population density, poor land use management, and a non-
homogenized distribution of infrastructures may all contribute to the
high level of vulnerability in metropolitan cities. Rural areas are also
affected by earthquakes. However, with the movement of the rural
population to urban areas with a high population density (that is, ur-
banization), the exposure rate to earthquakes changes (Bai et al., 2017),
making the region more vulnerable. Kolkata, the capital city of West
Bengal, has a population density of more than 24,000 persons per square
kilometer (Census, 2011); this is 20 times the study area’s average
population density. The Darjeeling Himalayas have 923.57 people per
square kilometer (Apollo, 2017). However, Darjeeling has several pop-
ular tourist attractions and gets packed during peak season, making the
region highly vulnerable. The growing population accelerated un-
planned and rapid urbanization and poor land use management,
resulting in a shortage of resources. In addition, building densities are
Table 1
Priority of vulnerability-mapping parameters.
12345678910
11223344567
2 1/2 1 2 2 3 3 4 5 5 6
3 1/2 1/2 1 2 2 3 3 4 5 5
4 1/3 1/2 1/2 1 2 2 3 3 4 5
5 1/3 1/3 1/2 1/2 1 2 2 3 4 5
6 1/4 1/3 1/3 1/2 1/2 1 2 2 3 4
7 1/4 1/4 1/3 1/3 1/2 1/2 1 2 2 3
8 1/5 1/5 1/4 1/3 1/3 1/2 1/2 1 2 2
9 1/6 1/5 1/5 1/4 1/4 1/3 1/2 1/2 1 2
10 1/7 1/6 1/5 1/5 1/5 1/4 1/3 1/2 1/2 1
Sl.
No.
Selected Layers AHP Weight
(Subjective)
Entropy Weight
(Objective)
Overall
Weight
1 Population Density 24.50% 06.70% 17.40%
2 Building Density 19.20% 13.40% 16.90%
3 Transportation
Terminals
14.90% 13.50% 14.30%
4 Popular Places 11.40% 14.30% 12.50%
5 Visiting Places 09.20% 13.10% 10.80%
6 Historical Places 06.80% 11.50% 08.70%
7 Museums 05.10% 09.10% 06.70%
8 Religious Places 03.80% 07.80% 05.40%
9 Park Density 02.90% 06.10% 04.20%
10 Dams 02.20% 04.40% 03.10%
Number of comparisons =45.
Consistency Ratio CR =02.60%.
Principal eigenvalue =10.34.
Table 2
Priority of seismic hazard-mapping parameters.
1 2 3 4 5 6 7 8
1 1 2 2 3 3 4 5 5
2 1/2 1 2 2 3 3 4 5
3 1/2 1/2 1 2 2 3 3 4
4 1/3 1/2 1/2 1 2 2 3 4
5 1/3 1/3 1/2 1/2 1 2 2 3
6 1/4 1/3 1/3 1/2 1/2 1 2 3
7 1/5 1/4 1/3 1/3 1/2 1/2 1 2
8 1/5 0.2 1/4 1/4 1/3 1/3 1/2 1
Sl.
No.
Selected Layers AHP Weight
(Subjective)
Entropy Weight
(Objective)
Overall
Weight
1 Distance from
Fault
28.00% 08.70% 20.30%
2 Distance from
Epicentres
20.90% 17.60% 19.60%
3 Seismic Density 15.70% 16.50% 16.00%
4 Vs30 11.90% 17.40% 14.10%
5 Peak Ground
Acceleration
08.70% 13.70% 10.70%
6 Sediment
Thickness
06.80% 12.20% 08.90%
7 Lithology 04.80% 09.20% 06.60%
8 Elevation 03.30% 04.90% 03.90%
Number of comparisons =28.
Consistency Ratio CR =02.50%.
Principal eigenvalue =08.24.
Table 3
Priority of coping capacity-mapping parameters.
1 2 3 4 5 6
1 1 2 2 3 4 5
2 1/2 1 2 3 3 4
3 1/2 1/2 1 2 3 3
4 1/3 1/3 1/2 1 2 3
5 1/4 1/3 1/3 1/2 1 2
6 1/5 1/4 1/3 1/3 1/2 1
Sl.
No.
Selected
Layers
AHP Weight
(Subjective)
Entropy Weight
(Objective)
Overall
Weight
1 Hospital 33.80% 12.30% 25.20%
2 Educated
People
24.90% 24.80% 24.90%
3 Road
Networks
17.40% 20.60% 18.70%
4 Rail
Networks
11.30% 20.40% 14.90%
5 Service
Centers
7.50% 13.70% 10.00%
6 Waterways 5.10% 08.30% 06.40%
Number of comparisons =15.
Consistency Ratio CR =02.50%.
Principal eigenvalue =06.15.
Fig. 3. Vulnerability map of the study region.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
5
higher in metropolitan areas than in rural regions. Kolkata has relatively
high building density, historical infrastructure, and the most crowded
transit facilities, including the Howrah, Sealdah, and Kolkata railway
stations, Netaji Subhas Chandra Bose International Airport. All these
factors may contribute to high vulnerability, as observed in Fig. 3. The
low-vulnerability areas in the study region may have low populations,
favorable socio-economic conditions, and low building density.
Besides population density and urbanization, tectonic prominence
may pose a signicant risk against seismic hazards. The Eocene Hinge
Zone, Debagram Bogra Fault, and 1994, 1996, and 2013 earthquakes
make the eastern part of the study area seismically hazardous. The
Pingla Fault dominates the west, and earthquakes occurred in 1993,
2008, 2019, and 2020. The northern study area is amid Himalayan
foothills. Metamorphic rocks with active faults, high elevation, Vs30,
and frequent far- and near-source events characterize the Darjeeling
Himalayas. It was estimated by Yadav and Tiwari (2018) that the
maximum horizontal stress (S
Hmax
) and its direction and size in the
Himalayan area, which includes the Darjeeling Himalayas, are consis-
tent with the measured maximum horizontal stress. Our study also
shows that Kolkata comes under very high seismic hazards due to the
presence of the Eocene Hinge Zone and the occurrence of earthquakes
near the source zones. However, the city is also affected by far-source
zone earthquakes. Mohanty et al. (2013) computed the seismic ground
motion along 2-D geological cross-sections in Kolkata for the earthquake
Table 4
Vulnerability, seismic hazard, and risk in terms of areas and population in West
Bengal.
Classes Area % Population %
Vulnerability Very Low 19.41 12.04
Low 20.06 07.11
Moderate 19.76 11.01
High 21.42 06.88
Very High 19.35 62.95
Seismic Hazard Very Low 19.95 06.81
Low 21.41 43.49
Moderate 21.24 22.33
High 21.40 12.70
Very High 15.99 14.66
Seismic Risk Very Low 18.59 02.52
Low 20.91 05.01
Moderate 21.81 08.06
High 19.92 14.31
Very High 18.77 70.10
Fig. 4. Seismic hazard map of the study region.
Fig. 5. Coping capacity map of the study region.
Fig. 6. Seismic risk map of the study region.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
6
that occurred on June 12, 1897, in the Shillong plateau to examine the
impact of source and epicentral distance on the local seismic response in
Kolkata city. A scenario earthquake in the Shillong plateau causes var-
iations in the peak ground acceleration (PGA) in the Kolkata metropolis,
ranging from 0.11 to 0.18 g. The area where extreme PGA are computed
may be entirely devastated, with signicant life and property damage, if
the examined earthquake scenario occurs.
By propagating the bedrock ground motion through a 1-D sediment
column with a 10% probability of exceedance in 50 years using an
equivalent linear analysis, Nath et al. (2014) conducted a probabilistic
seismic hazard assessment at a surface-consistent level for Kolkata. This
assessment of the local seismic hazard related to site amplication
predicts a PGA range of 0.176–0.253 g in Kolkata. 50% of Kolkata is in
the prospective liqueable zone according to a deterministic liquefac-
tion scenario based on the geographical distribution of the liquefaction
potential index matching the surface PGA distribution. Soil liquefaction
is commonly observed in cohesionless saturated soil when dynamic
stress and a rise in pore water pressure lead the soil’s shear strength to
drop to zero. The eastern and northeastern parts of Kolkata city feature
large patches of zones with a high risk for liquefaction (Nath et al.,
2014). The Mw 8.1 earthquake in Bihar-Nepal caused signicant dam-
age to life and property in the form of cracks in buildings, subsidence,
and collapse, as reported, presumably due to the effect of soil liquefac-
tion caused by the intensifying ground motion combined with the
shallow groundwater table and the thick alluvial-lled Bengal Basin.
However, the expansion of hospitals and communication networks
and the rise in the number of educated people are some of the primary
causes of the high coping capacity of cities. Metropolitan cities have
government multi-specialty and private hospitals and extensive road
and rail networks. However, areas like the Sunderban coastal region are
economically disadvantaged regions affected by a number of natural
calamities (Mondal et al., 2022). Waterways primarily serve this region
and have insufcient healthcare and education facilities, making most
sites remote and inaccessible and having extremely low coping capacity.
The hilly terrain of the northern study area makes the development of
transportation networks complicated, and the region’s coping capacity
is very low to moderate. However, with appropriate planning and pre-
paredness strategies, the seismic hazard can be mitigated to an extent.
6. Validation and sensitivity analysis of the model
A sensitivity analysis was conducted on the vulnerability, seismic
hazard, and coping capacity to better comprehend the inuence of
various parameters used in this work (Table 5). For this purpose, we
adjust the ’
α
’ to assess the inuence of weights on the used parameters.
The result shows that the ranking of the parameters used for estimating
vulnerability, seismic hazard, and coping capacity remains consistent
for
α
≥0.6, validating the assumption of Chuansheng et al. (2012) of
setting the
α
as 0.6.
Converting the vulnerability and risk map to a probability map
validated the results (Pradhan et al., 2014). Our proposed model was
validated by developing the receiver operating characteristic curve,
where the area under the curve (AUC) measures the accuracy of prob-
ability assessment where the dot black diagonal line serves as the
reference line and represents a random classier (Fig. 7). Our results
indicate that the AUC is 0.663, corresponding to a prediction accuracy of
66.30% (Fig. 7). At AUC =0.6–0.7, the GIS-based models may have
satisfactory prediction accuracy (Trifonova et al., 2013). This prediction
may be enhanced by introducing more high-resolution datasets affecting
the region’s seismicity.
7. Conclusion
This study integrated objective AHP and subjective Entropy MCDM
models to estimate seismic risk. Twenty-four seismic risk inuencing
parameters were used to assess vulnerability, seismic hazard, and coping
capacity in West Bengal, India, and to estimate seismic risk.
Our analysis shows that the northern and eastern part of West Bengal
is relatively highly vulnerable. Numerically, it is found that approxi-
mately 70% of the population may live in a highly vulnerable zone.
Table 5
Analysis of sensitivity for inuence levels of the
α
in the AHP-Entropy method
for vulnerability, hazard, and coping capacity.
Selected Layers
α
=0.2
α
=0.4
α
=0.6
α
=0.8
α
=1
Population Density 0.103
(6)
0.138
(3)
0.174
(1)
0.209
(1)
0.245
(1)
Building Density 0.145
(1)
0.157
(1)
0.169
(2)
0.180
(2)
0.192
(2)
Transportation
Terminals
0.138
(2)
0.141
(2)
0.143
(3)
0.146
(3)
0.149
(3)
Popular Places 0.137
(3)
0.131
(4)
0.125
(4)
0.120
(4)
0.114
(4)
Visiting Places 0.124
(4)
0.116
(5)
0.108
(5)
0.100
(5)
0.092
(5)
Historical Places 0.106
(5)
0.096
(6)
0.087
(6)
0.077
(6)
0.068
(6)
Museums 0.083
(7)
0.075
(7)
0.067
(7)
0.059
(7)
0.051
(7)
Religious Places 0.070
(8)
0.062
(8)
0.054
(8)
0.046
(8)
0.038
(8)
Park Density 0.055
(9)
0.048
(9)
0.042
(9)
0.035
(9)
0.029
(9)
Dams 0.039
(10)
0.035
(10)
0.031
(10)
0.026
(10)
0.022
(10)
Selected Layers
α
=0.2
α
=0.4
α
=0.6
α
=0.8
α
=1
Distance from Fault 0.125
(5)
0.164
(2)
0.203
(1)
0.241
(1)
0.280
(1)
Distance from
Epicentres
0.182
(1)
0.189
(1)
0.196
(2)
0.202
(2)
0.209
(2)
Seismic Density 0.164
(2)
0.162
(3)
0.160
(3)
0.159
(3)
0.157
(3)
Vs30 0.163
(3)
0.152
(4)
0.141
(4)
0.130
(4)
0.119
(4)
PGA 0.127
(4)
0.117
(5)
0.107
(5)
0.097
(5)
0.087
(5)
Sediment Thickness 0.111
(6)
0.100
(6)
0.089
(6)
0.079
(6)
0.068
(6)
Lithology 0.083
(7)
0.074
(7)
0.066
(7)
0.057
(7)
0.048
(7)
Elevation 0.046
(8)
0.042
(8)
0.039
(8)
0.036
(8)
0.033
(8)
Selected Layers
α
=0.2
α
=0.4
α
=0.6
α
=0.8
α
=1
Hospital 0.166 (4) 0.209 (2) 0.252 (1) 0.295 (1) 0.338 (1)
Educated People 0.248 (1) 0.248 (1) 0.249 (2) 0.249 (2) 0.249 (2)
Road Network 0.200 (2) 0.193 (3) 0.187 (3) 0.180 (3) 0.174 (3)
Rail Network 0.186 (3) 0.168 (4) 0.149 (4) 0.131 (4) 0.113 (4)
Service Centre 0.125 (5) 0.112 (5) 0.100 (5) 0.087 (5) 0.075 (5)
Waterways 0.077 (6) 0.070 (6) 0.064 (6) 0.057 (6) 0.051 (6)
Fig. 7. Receiver operating characteristic curve associated with seismic risk. The
dot black diagonal line represents a random classier.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
7
According to the seismic hazard results, the northern region of West
Bengal has a relatively high level of seismic hazard, and about 28% of
the population may reside in highly seismic hazard zones. However,
coping capacity may be essential in minimizing the threat against the
potential event. Some major metropolitan cities in the east of the study
region may have high coping capacity. Interestingly, 0.77% of the
population in the study area may reside under very low coping capacity,
whereas 78.01% have very high coping capacity. The seismic risk is
estimated using seismic hazard, vulnerability, and coping capacity. The
study found that 18.59% of the region may have very low seismic risk,
20.91% low, 21.81% moderate, 19.92% high, and 18.77% very high. In
terms of population, the result shows that 2.52% of the population may
live in very low seismic risk zones, 5.01% in low seismic risk zones,
8.06% in moderate seismic risk zones, 14.31% in high seismic risk zones,
and 70.10% may be in very high seismic risk zones. The outcome has
been validated through a receiver operating characteristic curve, dis-
playing satisfactory performance in seismic risk estimation. The ndings
could assist disaster mitigation agencies in identifying seismic risk zones
and planning for potential hazards in the region.
CRediT authorship contribution statement
Monalisa Malakar: Validation, Methodology, Formal analysis, Data
curation, Conceptualization. Sukanta Malakar: Writing – review &
editing, Writing – original draft, Validation. Mohd Sayeed Ul Hasan:
Writing – review & editing, Supervision, Project administration, Inves-
tigation. Abhishek K. Rai: Project administration, Investigation. Vijay
K. Kannaujiya: Writing – review & editing, Writing – original draft.
Declaration of competing interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
The authors thank the University Institute of Technology, Aliah
University, Adamas University and the Indian Institute of Technology
Kharagpur for providing all the resources. We also thank the agencies
that made the data accessible to the public.
References
Alam, M.S., Haque, S.M., 2021. Multi-dimensional earthquake vulnerability assessment
of residential neighborhoods of Mymensingh City, Bangladesh: a spatial multi-
criteria analysis based approach. J. Urban Manag. 11, 37–58. https://doi.org/
10.1016/j.jum.2021.09.001.
Alexander, D.C., 2017. Natural Disasters. Routledge, Abingdon, UK.
Alizadeh, M., Ngah, I., Hashim, M., Pradhan, B., Pour, A.B., 2018. A hybrid analytic
network process and articial neural network (ANP-ANN) model for urban
earthquake vulnerability assessment. Rem. Sens. 10, 975. https://doi.org/10.3390/
rs10060975.
Apollo, M., 2017. The population of Himalayan regions-by the numbers: past, present
and future. Contemporary Studies in Environment and Tourism. Camb. Sch. Pubs.
Bai, X., McPhearson, T., Cleugh, H., et al., 2017. Linking urbanization and the
environment: conceptual and empirical advances. Annu. Rev. Environ. Resour. 42
(1), 215–240. https://doi.org/10.1146/annurev-environ-102016-061128.
Bhattacharya, A., Geraghty, J., Young, P., 2010. Supplier selection paradigm: an
integrated hierarchical QFD methodology under multiple-criteria environment.
Appl. Soft Comput. 10, 1013–1027.
Bureau of Indian Standards (BIS), 2002. IS 1893-2002 (Part 1) Indian Standard Criteria
for Earthquake Resistant Design of Structures, Part 1-General Provisions and
Buildings. Bureau of Indian Standards, New Delhi.
Caprio, M., Tarigan, B., Worden, B.C., Wiemer, S., Wald, D.J., 2015. Ground motion to
intensity conversion equations (GMICEs): a global relationship and evaluation of
regional dependency. Bull. Seismol. Soc. Am. 105, 1476–1490. https://doi.org/
10.1785/0120140286.
Census, 2011. Census of India. Available at: http://censusindia.gov. last access: 27th July
2023.
Chuansheng, X., Dapeng, D., Shengping, H., Xin, X., Yingjie, C., 2012. Safety evaluation
of smart grid based on AHP-Entropy method. Sys. Eng. Proc. 4, 203–209.
Centre for Research on the Epidemiology of Disasters (CRED), 2015. The Human Cost of
Natural Disasters: A Global Perspective.
Cui, Y., Feng, P., Jin, J., Liu, L., 2018. Water resources carrying capacity evaluation and
diagnosis based on set pair analysis and improved the Entropy weight method.
Entropy 20.
Dunbar, P.K., Bilham, R.G., Laituri, M.J., 2003. Earthquake loss estimation for India
based on macroeconomic indicators. Risk Sci. Sustain. 163–180. https://doi.org/
10.1007/978-94-010-0167-0_13.
Dyer, R.F., Forman, E.H., 1992. Group decision support with the analytic hierarchy
process. Decis. Support Syst. 8, 99–124. https://doi.org/10.1016/0167-9236(92)
90003-8.
Flores, K.L., Escudero, C.R., Zamora-Camacho, A., 2021. Multicriteria seismic hazard
assessment in Puerto Vallarta metropolitan area, Mexico. Nat. Haz. 105, 253–275.
https://doi.org/10.1007/s11069-020-04308-x.
GSI, 1939. The Bihar-Nepal Earthquake of 1934. Memoirs of the Geological Survey of
India.
Gleeson, T., Smith, L., Moosdorf, N., et al., 2011. Mapping permeability over the surface
of the Earth. Geophys. Res. Lett. 38. https://doi.org/10.1029/2010GL045565.
Hasan, M.S.U., Rai, A.K., 2020. Groundwater quality assessment in the Lower Ganga
Basin using entropy information theory and GIS. J. Clean. Prod. 274. https://doi.
org/10.1016/j.jclepro.2020.123077.
Jena, R., Pradhan, B., Beydoun, G., 2020. Earthquake vulnerability assessment in
Northern Sumatra province by using a multi-criteria decision-making model. Int. J.
Disaster Risk Reduc. 46, 101518. https://doi.org/10.1016/j.ijdrr.2020.101518.
Jena, R., Pradhan, B., Naik, S.P., Alamri, A.M., 2021. Earthquake risk assessment in NE
India using deep learning and geospatial analysis. Geosci. Front. 12 (3). https://doi.
org/10.1016/j.gsf.2020.11.007.
Karaman, H., Erden, T., 2014. Net earthquake hazard and elements at risk (NEaR) map
creation for the city of Istanbul via spatial multi-criteria decision analysis. Nat. Haz.
73, 685–709. https://doi.org/10.1007/s11069-014-1099-2.
Kannaujiya, V.K., Rai, A.K., Malakar, S., 2024. Coastal shoreline change in eastern Indian
metropolises. Photogramm. Fernerkund. GeoInf. https://doi.org/10.1007/s41064-
024-00286-y.
Maiti, S.K., Nath, S.K., Adhikari, M.D., Srivastava, N., Sengupta, P., Gupta, A.K., 2017.
Probabilistic seismic hazard model of West Bengal, India. J. Earthq. Eng. 21,
1113–1157. https://doi.org/10.1080/13632469.2016.1210054.
Malakar, S., Rai, A.K., 2022a. Earthquake vulnerability in the Himalaya by integrated
multi-criteria decision models. Nat. Haz. 111, 213–237. https://doi.org/10.1007/
s11069-021-05050-8.
Malakar, S., Rai, A.K., 2022b. Seismicity clusters and vulnerability in the Himalayas by
machine learning and integrated MCDM models. Arabian J. Geosci. 15, 1674.
https://doi.org/10.1007/s12517-022-10946-1.
Malakar, S., Rai, A.K., Gupta, A.K., 2023. Earthquake risk mapping in the Himalayas by
integrated analytical hierarchy process, entropy with neural network. Nat. Haz. 116,
951–975. https://doi.org/10.1007/s11069-022-05706-z.
Martins, V.N., e Silva, D.S., Cabral, P., 2012. Social vulnerability assessment to seismic
risk using multi-criteria analysis: the case study of Vila Franca de Campo (Miguel
Island, Azores, Portugal). Nat. Haz. 62, 385–404. https://doi.org/10.1007/s11069-
012-0084-x.
Mohanty, W.K., Walling, M.Y., 2008. Seismic hazard in mega city Kolkata, India. Nat.
Haz. 47, 39–54. https://doi.org/10.1007/s11069-007-9195-1.
Mohanty, W.K., Verma, A.K., Vaccari, F., Panza, G.F., 2013. Inuence of epicentral
distance on local seismic response in Kolkata City, India. J. Earth Syst. Sci. 122,
321–338. https://doi.org/10.1007/s12040-013-0275-1.
Molla, S.H., Rukhsana, Hasan, M.S.U., 2023. Deployment of entropy information theory
in the Indian Sundarban region using hydrogeochemical parameters and GIS for
assessment of irrigation suitability. Environ. Monit. Assess. 195, 1227. https://doi.
org/10.1007/s10661-023-11847-w.
Mondal, M., et al., 2022. Climate change, multi-hazards and society: an empirical study
on the coastal community of Indian Sundarban. Nat. Haz. Res. 2, 84–96. https://doi.
org/10.4236/jgis.2011.31004.
Nath, S.K., Adhikari, M.D., Maiti, S.K., Devaraj, N., Srivastava, N., Mohapatra, L.D.,
2014. Earthquake scenario in West Bengal with emphasis on seismic hazard
microzonation of the city of Kolkata, India. Nat. Hazards Earth Syst. Sci. 14,
2549–2575. https://doi.org/10.5194/nhess-14-2549-2014.
Nath, S.K., Adhikari, M.D., Devaraj, N., Maiti, S.K., 2015. Seismic vulnerability and risk
assessment of Kolkata City, India. Nat. Haz. 15, 1103–1121. https://doi.org/
10.5194/nhess-15-1103-2015.
Nyimbili, P.H., Erden, T., Karaman, H., 2018. Integration of GIS, AHP and TOPSIS for
earthquake hazard analysis. Nat. Haz. 92, 1523–1546. https://doi.org/10.1007/
s11069-018-3262-7.
Pal, I., Nath, S.K., Shukla, K., Pal, D.K., Raj, A., Thingbaijam, K.K.S., Bansal, B.K., 2008.
Earthquake hazard zonation of Sikkim Himalaya using a GIS platform. Nat. Haz. 45,
333–377. https://doi.org/10.1007/s11069-007-9173-7.
Panahi, M., Rezaie, F., Meshkani, S.A., 2014. Seismic vulnerability assessment of school
buildings in Tehran city based on AHP and GIS. Nat. Hazards Earth Syst. Sci. 14,
969–979. https://doi.org/10.5194/nhess-14-969-2014.
Pelletier, J.D., Broxton, P.D., Hazenberg, P., et al., 2016. Global 1-km Gridded Thickness
of Soil, Regolith, and Sedimentary Deposit Layers. ORNL DAAC, Oak Ridge,
Tennessee, USA. https://doi.org/10.3334/ORNLDAAC/1304.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
8
Pradhan, B., Hasan, H.A., Jebur, M.N., Tehrany, M.S., 2014. Land subsidence
susceptibility mapping at Kinta Valley (Malaysia) using the evidential belief function
model in GIS. Nat. Haz. 73 (2), 1019–1042. https://doi.org/10.1007/s11069-014-
1128-1.
Rai, A.K., Malakar, S., Goswami, S., 2023. Active source zones and earthquake
vulnerability around Sumatra subduction zone. J. Earth Syst. Sci. 132, 66. https://
doi.org/10.1007/s12040-023-02070-9.
Ramachandra, T.V., Aithal, B.H., Sowmyashree, M.V., 2014. Urban structure in Kolkata:
metrics and modelling through geoinformatics. Appl. Geomat. 6, 229–244. https://
doi.org/10.1007/s12518-014-0135-y.
Ray, R., Das, A., Hasan, M.S.U., Aldrees, A., Islam, S., Khan, M.A., Lama, G.F.C., 2023.
Quantitative analysis of land use and land cover dynamics using geoinformatics
techniques: a case study on Kolkata metropolitan development authority (KMDA) in
West Bengal, India. Rem. Sens. 15, 959. https://doi.org/10.3390/rs15040959.
Rodcha, R., Tripathi, N.K., Shrestha, R.P., 2019. Comparison of cash crop suitability
assessment using parametric, AHP, and F-AHP methods. Land 8 (5), 79.
Saaty, T.L., 1980. The Analytic Hierarchy Process: Planning, Priority Setting, Resource
Allocation. McGraw, New York.
Shannon, C.E., 1948. A mathematical theory of communication. Bell Syst. Tech. J. 27,
379–423.
Sinha, N., Priyanka, N., Joshi, P.K., 2016. Using spatial multi-criteria analysis and
ranking tool (SMART) in earthquake risk assessment: a case study of Delhi region,
India. Geomatics, Nat. Hazards Risk 7 (2), 680–701. https://doi.org/10.1080
/19475705.2014.945100.
SRTM, 2013. Shuttle radar topography mission (SRTM) global. Distributed by
OpenTopography. https://doi.org/10.5069/G9445JDF.
Trifonova, O.P., Lokhov, P.G., Archakov, A.I., 2013. Metabolic proling of human blood.
Biochem. Moscow Suppl. Ser. B7, 179–186. https://doi.org/10.1134/
S1990750813030128.
Wang, D., Singh, V., Zhu, Y., Wu, J., 2009. Stochastic observation error and uncertainty
in water quality evaluation. Adv. Water Resour. 32, 1526–1534.
Wisner, B., Blaikie, P., Cannon, T., Davis, I., 2008. At Risk: Natural Hazards, People’s
Vulnerability and Disasters, second ed. Routledge, New York, p. 275.
World Health Organization (WHO), 2009. Vulnerability and risk analysis and mapping
(VRAM) platform for health risk reduction. Ninth United Nations Regional
Cartographic Conference for the Americas.
Yadav, R., Tiwari, V.M., 2018. Numerical simulation of present day tectonic stress across
the Indian subcontinent. Int. J. Earth Sci. 107, 2449–2462. https://doi.org/10.1007/
s00531-018-1607-9.
Zhu, J., Zhang, Y., Zhang, J., Chen, Y., Liu, Y., Liu, H., 2023. Multi-Criteria seismic risk
assessment based on combined Weight-TOPSIS Model and CF-Logistic Regression
model-A case study of Songyuan City, China. Sustain. Times 15 (14), 11216. https://
doi.org/10.3390/su151411216.
M. Malakar et al.
Evolving Earth 2 (2024) 100036
9
