Natural Hazards Research 4 (2024) 423–433

Contents lists available at ScienceDirect

Natural Hazards Research

journal homepage: www.keaipublishing.com/en/journals/natural-hazards-research

Comprehensive seismic hazard assessment for Guwahati City, Northeast

India: Insights from probabilistic and deterministic seismic hazard analysis
Satyaprakash Mishra a, *, Aman Kumar b, Arjun Sil a
a
Dept. of Civil Engineering, National Institute of Technology Silchar, Assam, India
b
Dept. of Civil Engineering, Motilal Nehru National Institute of Technology, UP, India

A R T I C L E I N F O A B S T R A C T

Keywords: This paper presents a comprehensive seismic hazard assessment for Guwahati city, located in the northeast region
Seismicity parameter of India. The earthquake data considered for the study is from 1762 to 2020, which was sourced from the
Earthquake published data of various seismological agencies. The fault information comprises of the latest studies and
Ground motion prediction models
Seismotectonic Atlas (SEISAT) data. The density of earthquakes close to the faults was used to define the fault
Peak ground acceleration (PGA)
Spectral acceleration (SA)
zones. The G-R relationship was utilized to calculate the seismicity parameters of the identified faults. We have
used PSHA and DSHA to estimate the ground motion parameters such as peak ground acceleration and spectral
acceleration at various probabilities of exceedance to calculate the level of seismic hazard. The results indicate
that the seismic hazard in Guwahati is moderate to high, with several active faults in the vicinity and a high
potential for strong ground motions. This study includes identifying the controlling source for Guwahati and
plotting the seismic hazard curve and uniform hazard spectra for various probabilities of exceedance, which will
give a basic idea of the seismic hazard of the city. The findings of this study can be used as a basis for earthquake
risk reduction and mitigation strategies in Guwahati.

1. Introduction enables us to estimate all possible horizontal forces due to an earthquake

event. Guwahati city has an average PGA value of 0.36g as per the Bureau
Seismic hazard analysis of Guwahati city is crucial as its localized in of Indian Standards (2002)(Bureau of Indian Standards, New Delhi,
an area with high human activity and seismic activity. Guwahati lies in 2016).
the northeastern part of India, which is prone to earthquakes due to its Some significant past seismic events that occurred around the study
proximity to the Himalayan seismic belt. Seismic hazard analysis is area are Cachar-1869(Mw7.4), Shillong-1897 (Mw 8.1), Assam-1950 (Mw
necessary to assess the potential damage that can be caused by earth- 8.6,8.7), Shrimangal-1918 (Ms 7.6), Dhubri1930 (Ms 7.1) and Man-
quakes of varying magnitudes and to determine the level of preparedness ipur1988 (Mw 7.2). Similarly, some of the recent major earthquakes are
required for emergency response. The study area is the most populated Dibrugarh-2015 (Mw 5.5), Manipur-2016 (Mw 6.8), and Sikkim-2011 (Mw
city in North East India is Guwahati, which is located on the banks of the 6.9). Many lives are lost as a result of the ground motion caused by these
Brahmaputra river shown in Fig. 1. With a growing population and earthquakes, and important structures like bridges, buildings, dams, and
infrastructure development, the need for seismic hazard analysis is roads also sustain damage. Thus, it has become increasingly important to
becoming increasingly important to ensure the safety and resilience of design earthquake-resistant structures that can withstand ground motion
Guwahati city. thereby minimizing causalities and property loss.
When an earthquake happens, energy is released in the form of waves Several researchers have worked in seismic hazard analysis (SHA) for
that move the ground. Estimating probable ground motion in this area is various cities like (Misliniyati et al., 2018; Mase et al., 2021) did the
the main goal of this seismic hazard study. Peak ground acceleration seismic hazard analysis of Bengkulu city and also considered the local soil
(PGA), peak ground velocity (PGV), peak ground displacement (PGD), characteristics to access the surface level PGA value. (Somantri et al.,
and spectral acceleration (SA) are common terms used to estimate 2023; Mase et al., 2023) conducted the seismic ground response analysis
ground motion(L. Kramer (1996)). The values of these ground motion for the Java province of Indonesia to calculate the potential seismic
variables are useful in the earthquake-resistant design of structures. This damage to the region due to predicted earthquakes. (Nath and

* Corresponding author. Department of Civil Engineering, National Institute of Technology Silchar, Assam, 788010, India.
E-mail addresses: spmishra.civil@gmail.com, satyaprakash_rs@civil.nits.ac.in (S. Mishra).

https://doi.org/10.1016/j.nhres.2023.10.005
Received 31 August 2023; Received in revised form 10 October 2023; Accepted 21 October 2023
Available online 8 November 2023
2666-5921/© 2023 National Institute of Natural Hazards, Ministry of Emergency Management of China. Publishing services provided by Elsevier B.V. on behalf of
KeAi Communications Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Fig. 1. Study region map (Guwahati City).

Thingbaijam, 2012; Sharma and Malik, 2006) have carried out proba- data from various agencies. Similarly, the major fault data is identified
bilistic analysis (PSHA) in India, (Devi and Kalita, 2014) have used PSHA using SEISAT 2000; Geological Survey of India (2000). On the basis of
method for northeast India and its connecting region, (Baro et al., 2018) these sources, we have gathered data on the fault zones, the recurrence
have performed hazard study for Shillong plateau region, (Kanth and pattern, the maximum observed magnitude and fault data (latitude,
Iyengar, 2006)have performed SHA for Mumbai city, (Joshi et al., 2007) longitude, length and depth) for each fault. Then the seismicity param-
have plotted seismic Hazard map for North East India, (Sil et al., 2013) eters (a & b) are calculated for each fault based upon Gutenberg (g-r)
have performed hazard assessment for Tripura and Mizoram state, recurrence law using the least square method. The maximum magnitude
(Bahuguna and Sil, 2020) have performed hazard study for Assam state, for each fault zone method has been taken using the data given by
(Kumar et al., 2023) for Silchar city, Assam and (Das et al., 2016b) have (Jaiswal et al., 2002; Kijko, 2004). Subsequently, appropriate equations
created seismic hazard map for northeastern India in a probabilistic were selected on the basis of the comparison between the observed PGA
manner. data from the COSMOS data center and the result of the attenuation
The seismicity of any region changes as both the frequencies of equation given by various researchers. Finally, by using the selected
earthquakes and maximum magnitude change over time in that region. ground motion prediction equation (GMPE), the PGA value has been
This necessitates a newer seismic hazard study of the region of interest calculated for Guwahati city in the North East India region by using the
(Guwahati) by incorporating the study with the latest seismic data which DSHA method and PSHA method. Results of the Seismic Hazard analysis
will help make better estimates allowing us to do better disaster man- comprise of identification of controlling sources, plotting of seismic
agement and also helping us to mitigate the hazard appropriately. In the hazard curve and uniform hazard spectra for the selected city. The
current study, Guwahati city is taken as the study region, this is because bedrock level PGA value resulted from the study can be further used as
Guwahati has the highest population density in Northeast India and has the input parameter for the site-specific ground response analysis and
many important structures like; institutes, universities, ancient heritage calculate the surface level ground acceleration considering the local soil
temples, and other important installations. The presence of these heavy conditions.
structures and populations makes cities more susceptible to damage from Earthquake catalogue has been collected from longitude 86
to 100

natural hazards. and latitude 18
to 30
encompassing the North East region boundary of
For purpose of this study, the earthquake data from the year India (Fig. 6). The event details and data have been sourced from various
1763–2020 are taken, which include pre-instrumental and instrumental International seismological agencies such as Incorporated Research

424
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Institutions for Seismology (IRIS), United States of Geological Survey marking and colors. The process of collecting seismic source data en-
(USGS), European-Mediterranean Seismological Centre (EMSC), Indian compasses gathering crucial fault parameters, including fault length,
Meteorological Department (IMD), National Center for Seismology depth, geographical coordinates, and the maximum recorded earthquake
(NCS), International Seismological Centre (ISC), and Incorporated magnitude associated with each fault. An assumption is made by
Research Institutions for Seismology (IRIS), National Oceanic and At- considering the point sources near any fault line are attribute to the ac-
mospheric Administration (NOAA) etc. The magnitude of a seismic event tivity originating from the respective fault. All such point sources are
is provided by various agencies is in different magnitude scale, such as encompassed within the defined fault zones. The seismic zones are
moment magnitude (Mw), surface magnitude (Ms), body wave magnitude defined based on the earthquake density in close proximity to the faults.
(Mb), and local magnitude (Ml). All other scales have been converted into Consequently, if an earthquake event occurs within this defined zone, we
(Mw) moment magnitude in order to maintain consistency using the re- consider the associated fault as the source of that event. Furthermore, in
lationships given by (Sitharam and Sil, 2014). This study takes Moment order to ensure precision in our analysis, we calculate the minimum
Magnitude (Mw) into account because other scales have saturation for source-to-site distance by taking into account the characteristics of the
larger size magnitudes and thus do not provide more precise measure- line sources considered in the study.
ments of the larger earthquake size. The moment magnitude scale, on the A total of 28 faults are present in the study region of which some
other hand, does not saturate even for larger magnitude of earthquakes. faults are active while others are inactive. Hence in the present study, 8
In the earthquake catalog 1936 events were recorded for the major faults which are active and produce reasonable ground motion at
magnitude (Mw) range 4–4.9, 1342 events in the magnitude (Mw) range the selected site are considered for the seismic hazard analysis. The
5–5.9, 144 events in the magnitude (Mw) range Mw 6–6.9, 21 events for average depth and length comparison of the faults are shown in Fig. 4.
magnitude (Mw) above 7 as shown in Fig. 2, which shows the data The Mishmi Thrust trends in WNW-ESE direction in the eastern Hima-
completeness using CUVI method. Further, these data are used in GIS layas. Specifically situated in Northeastern part of India and extending
tools for visualization and deciding the fault zones. From the given figure into Bhutan, Tibet, and Nepal. The Oldham fault has a length of 110 km.
it can be observed that data completeness for the magnitude range 4–4.9 It falls to the north of Shillong Plateau. The great assam earthquake of 8.1
is 56 years (1964–2020), for magnitude range 5–5.9 is 98 years magnitude in the year 1897 is associated with this fault. The Lohiti thrust
(1922–2020), for magnitude range 6–6.9 is 100 years (1920–2020) and lies parallel to the Mishmi thrust and has a length of around 200 km.
above 7 is 258 years (1762-2020). A detailed description of faults has given in Table 1 comprising of
Fault Data (latitude, longitude, length) and Maximum Magnitude for that
1.1. Seismic source zone.
These tectonic features (fault data) and seismicity event (earthquake
The seismic source identification and characterization forms a critical data) which was superimposed on the study area map using GIS tools to
component of our research, allowing for a more accurate assessment of define the fault zones as per the earthquake density near the faults. The
SHA in the study area. To accomplish this, we have obtained seismic seismic events (  4), seismic sources and fault zones are shown in Fig. 5.
source data from the seismotectonic atlas (SEISAT2000) (Geological
Survey of India, 2000) and various previous studies ((Sharma and Malik, 1.2. Estimation of seismicity parameters (“a” & “b”)
2006a,b; Sil and Bahuguna, 2018; Ghione et al., 2021)) shown in Fig. 3.
The different types of fractures on the crustal bedrock such as Thrusts, Any fault region's seismicity can be described by two parameters (a &
Faults and lineaments are shown in the figure with distinguishable b) as these two parameters can elucidate the relationship of the Moment

Fig. 2. Plot of data completeness by CUVI method.

425
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Fig. 3. Seismotectonic map of North East India (SEISAT, 2000).

Table 1
Details of the various faults considered in the study.
Fault Starting Ending Length (km) Depth

Long. Lat. Long. Lat.

A1 90.24 24.08 90.83 24.9 110 29

Oldham 90.72 25.77 91.73 26.11 110 31
Samin 90.53 25.76 90.67 25.9 20 26
Lohiti 95.68 27.71 97.6 28.74 200 34
Dauki 92.97 25.12 89.94 25.26 320 54
A3 93.04 26.31 93.65 26.48 80 38
Sylhet 90.69 23.99 91.9 24.94 161 28
Mishmi Thrust 94.35 26.92 97.44 28.9 230 32

logλ ¼ a  bM (1)
M is the Moment magnitude and λ shows the mean annual rate of
exceedance of a particular magnitude, The intercept coefficient (a) and
slope coefficient (b) determine the seismicity of any region. The value of
‘a’ signifies the number of earthquakes per year; and ‘b’ provides the ratio
of the number of lower magnitude earthquakes to the number of high
magnitude earthquakes. The value of ‘a’ fluctuates between 0.46 and 4.3
and the value of ‘b’ fluctuates between 0.36 and 1.01. The value of ‘b’
parameter at Mishmi Thrust fault is 1.01, which signifies that this fault
Fig. 4. Comparison of length and depth of considered seismic sources. has a high chance to cause low-magnitude earthquakes than earthquakes
with high magnitude. A1, Sylhet faults have a very low value of ‘a’ which
Magnitude (Mw) with the cumulative frequency of occurrence. Gutenberg implies the earthquake occurrences are very low as compared to other
and Richter (Beitr, 1945) established a relation between moment faults.
magnitude and frequency of occurrence as,

426
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Fig. 5. Earthquake Events and seismic zones in the study region.

1.3. Maximum magnitude (Mmax) E1 ðn2 Þ þ E1 ðn1 Þ

Mmax ¼ Mmax
obj
þ þ Mmin en (2)
βen2
Maximum Magnitude (Mmax) is a key input for the study of Seismic

 
hazards where a high magnitude value signifies high-ground motion. n1 ¼ n 1  eβðMmax Mmin Þ (3)
While (Kijko, 2004) claims that there is no proper relation for estimation
of the exact maximum magnitude for a source, however, some re-
n2 ¼ n1 eβðMmax Mmin Þ (4)
searchers have asserted that it is possible to make estimates based on
rupture length, rupture area and fault displacement which can be esti- 
 
mated by past earthquake data and field geological investigation. Among E1 ¼ n2 þ a1 n þ a2 n n2 þ b1 n þ b2 en (5)
them, some of the popular methods for calculating maximum magnitude
where Mmax is the predicted maximum magnitude, n is the total number
are ((Wells and Coppersmith, 1994; Kijko, 2004; Jaiswal et al., 2002;
of events in the region with a magnitude above threshold magnitude,
Raghu Kanth and Iyengar, 2007) and the Gutenberg-Richer(G-R)) trun-
n1and n2, E1(n) is the exponential integral function and a1, a2, b1, b2 are
cated law (Gutenberg and Richter, 1944).
coefficients given by (Ghione et al., 2021).
For the present study, the methods of (Jaiswal et al., 2002; Kijko,
According to (Jaiswal et al., 2002) the maximum magnitude of any
2004) are used to estimate the maximum magnitude (Mmax). Kijko
seismic source zone can be found by an increment of some units on the
method (Kijko, 2004) is based upon only the earthquake data and can be
observed maximum magnitude of that zone. This method represents one
used when we have limited or no information about the seismic source
of the most straightforward approaches for determining the maximum
and/or the nature of the earthquake. This approach is only based on
magnitude of the source zone by gradually increasing the observed
earthquake data and may be used when only limited data about the
magnitude of the source zone by 0.5 units. When the available data are
seismic source and characteristics of the earthquake is known. This
not adequate to define the frequency-magnitude relationship for a
approach primarily requires the number of earthquakes caused by that
seismic source, the maximum earthquake magnitude is sometimes also
source and the maximum observed magnitude. This approach relies
found simply by adding an increment to the largest historical earthquake.
solely on seismic data and does not rely on subjective assumptions. It can
This approach is one of the most basic and provides an unarguable
be utilized when there is no knowledge about the distribution of earth-
maximum magnitude. These method has been used by many researchers
quake magnitudes, and it is also applicable in situations where the
such as ((Bahuguna and Sil, 2020; Menon et al., 2010; Sil et al., 2013;
earthquake database is incomplete or when higher magnitude events are
Kumar et al., 2023)). The maximum magnitude estimation for each fault
scarce.
by using (Jaiswal et al., 2002; Kijko, 2004) methods are shown in

427
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Fig. 6. Flow chat of the methodology.

Table 2. The estimated maximum magnitude (Mmax) for a source is taken 2021). Ground motion prediction equations (GMPEs) are mathematical
as the maximum from both approaches. models that predict the characteristics of ground motion caused by
earthquakes. These equations are important because they allow us to
2. Ground motion attenuation relationships assess the potential risk posed by earthquakes to buildings, infrastruc-
ture, and other structures. These equations can be generated by regres-
To explore a suitable attenuation model, this section provides a sion analysis on a set of strong ground motion data. Typically, GMPEs rely
concise overview of the development of ground attenuation models. The on four key parameters: earthquake magnitude, source-to-site distance,
initial attenuation models, known as Ground Motion Prediction Equa- fault type, and local site conditions, all of which are essential for esti-
tions (GMPEs), originated in 1964. These early models started as mating both PGA and SA (Tanapalungkorn et al., 2020). In 2008, the
straightforward empirical equations designed to estimate both Peak Pacific Earthquake Engineering Research (PEER) Center published the
Ground Acceleration (PGA) and Spectral Acceleration (SA) (Ghione et al., Next Generation Attenuation (NGA) model project. This set of GMPEs

Table 2
Seismicity parameters and maximum magnitude of Faults.
Fault Name ‘a’ value ‘b’ value Length(km) Mmax (Observed) (Mw) Kijko Method Gupta Method Considered for Study

A1 0.46 0.36 110 7.2 7.3 7.7 7.7

Oldham 2.25 0.7 110 8.1 8.2 8.6 8.6
Samin 2.17 0.64 20 6.2 6.3 6.7 6.7
Lohiti 3.13 0.8 200 8.6 8.6 9.1 9.1
Dauki 2.4 0.68 320 6.4 6.5 6.9 6.9
A3 2.73 0.7 80 6.4 6.5 6.9 6.9
Sylhet 1.07 0.45 161 7.6 7.6 8.1 8.1
Mishmi Thrust 4.34 1.01 230 7.1 7.2 7.6 7.6

428
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

was based on data from shallow crustal earthquakes in regions with the worst-case scenario, which takes into account the maximum magni-
active tectonics, specifically the NGA-West1 database, encompassing tude possible for the source and the minimum distance between source
earthquakes with magnitudes ranging from 4.2 to 7.9 in the western and site. As a result, the DSHA method offers the maximum ground ac-
United States. Later, in 2014, these models underwent significant en- celeration value, which ordinarily does not happen during the structure's
hancements through the integration of updated global strong motion lifetime. The steps involved in the DSHA method are as follows:
data, referred to as the NGA-West2 database. This expanded dataset in-
cludes earthquakes with magnitudes ranging from 3.0 to 7.9 and com-  Identify and categorize seismic sources near the site.
prises approximately 21,000 records. These records encompass seismic  For each source zone, determine the lowest source-to-site distance
events occurring globally from 1935 to 2011, spanning a wide range of and maximum magnitude.
magnitudes(Mase, 2018). Furthermore, these models are used worldwide  Using the ground motion attenuation equation, identify controlling
considering the regional parameters tailored to specific geographic re- sources based on their ability generate strong ground shaking.
gions in various studies such as ((Misliniyati et al., 2018; Mase et al.,  PGA from the controlling source or an earthquake represent site
2021; Somantri et al., 2023)). In essence, the progression from early hazards.
empirical equations to the NGA-West2 models represents a significant
evolution in the field of ground motion prediction, facilitating more ac- But for the PSHA all possible scenarios are considered for hazard
curate assessments of seismic risk and the design of earthquake-resistant analysis. DSHA is a simple and quick method for hazard analysis. This has
structures across diverse geographic regions (Tanapalungkorn et al., been previously used in the design of large dams, power plants and other
2020). Therefore, it is very important to select the appropriate ground critical facilities. DSHA offers a metric for assessing the worst-case
motion model so that the correct variation of ground motion can be ground motion, but it also doesn't offer any details about when that
estimated. The study region mainly is part of Himalayan subduction earthquake would happen. Additionally, it ignores the ambiguity sur-
zone, so it is necessary to choose attenuation models addressing to rounding the magnitude, location, and timing of the earthquake.
similar tectonic set up. We have selected models based on the comparison The PSHA method takes into account every scenario that could
between the various ground motion attenuation models with the ground possibly take place in the source, and this method has also enabled using
motion data collected from available seismological agencies for various uncertainties in earthquake size, location, and frequency. PSHA is based
time periods. In this study 7 attenuation models proposed by various on the assumption that earthquakes are random events that follow a
researchers ((Kanno et al., 2006; Gupta, 2010; Anbazhagan et al., 2013; probabilistic pattern. It employs statistical models and computer simu-
Bajaj and Anbazhagan, 2019; Ndma, 2011; Sharma et al., 2011; Nath lations to estimate the likelihood of earthquakes of varying magnitudes
et al., 2012)) for North-East India region are considered to select the best occurring at various distances from the point of interest. The basic steps
suitable ground motion attenuation model. Then comparison of the involved in PSHA are as follows:
models has been performed with the station observed strong motion data
for different time periods. To select a suitable ground motion model,  Identify and categorize seismic sources near the site.
(Kumar et al., 2023) compared past earthquake data collected from  For each source, determine the seismicity parameters (a & b).
various stations in Northeast India with the available Ground Motion  Calculate site-specific ground motion (PGA) using ground motion
Prediction Models for the region such as (Ndma, 2011; Gupta, 2010; attenuation equations for different earthquake magnitude.
Sharma et al., 2011; Nath et al., 2012; Kanno et al., 2006; Bajaj and  Assess uncertainties using probability functions for earthquake
Anbazhagan, 2019). 70 seismic records of magnitudes of 4.2, 5.2, 6, and characteristics (size, location, and timing) and combine them to plot
6.9 were considered for validation purposes. After comparison, the seismic hazard curve.
models proposed by (Ndma, 2011; Bajaj and Anbazhagan, 2019) are  Calculate the probability that ground motion parameters will be
found to be the best fitting model for various periods (0-4 s) in North East exceeded over a specified period.
India region. Based on this finding, the study incorporates these two
models with a 0.5 impact factor. (Kumar et al., 2023). The considered Uncertainty can be identified, computed, and combined in this
models are mentioned in equation (6) and equation (7) respectively. method to provide a complete scenario of the seismic Hazard (Kramer,
  1996). The result of PSHA is generally provided in the form of seismic
Sa   hazard curve. The seismic hazard curve is a graphical representation of
ln ¼ C1 þ C2 M þ C3 M 2 þ C4 r þ C5 ln r þ C6 eC7 M þ C7 logðrÞf0 þ lnðεÞ
g the likelihood of experiencing ground shaking at or above a specific in-
(6) tensity level within a defined time frame. It is a fundamental tool used to
quantify the likelihood of ground shaking at or exceeding a certain level
In equation (6), Sa is spectral acceleration in g; M is moment magnitude; r of intensity over a specific period of time, typically expressed as a per-
is hypo-central distance and C1, C2, C3, C4, C5, C6, C7 and C8 are constants; centage or probability. This curve is an essential component of seismic
‘ε’ is the standard deviation. risk assessment and informs various aspects of earthquake engineering
and emergency preparedness. The vertical axis of a POE curve represents
lny ¼ a1 þ a2 ðM  6Þ þ a3 ð9  MÞ2 þ a4 lnR þ am lnRðM  6Þ þ a7 R þ σ
the probability of exceedance whereas the horizontal axis of the curve
(7) represents the level of ground motion intensity, often measured in terms
of a ground motion parameter. Firstly, all the uncertainties and indi-
In equation (7), ‘y’ is spectral acceleration in terms of acceleration due to
vidually calculated and finally merged to prepare the final result.
gravity(g), ‘M’ signifies moment magnitude, R is hypo-central distance
Different probability functions are used to estimate uncertainty in time,
and a1, a2, a3, a4 a5, a6, a7, and are constants, σ is the standard deviation.
magnitude, and distance. These functions are discussed further below.
3. Seismic hazard analysis
3.1. Estimation of magnitude uncertainty
Seismic hazard analysis is the process of evaluating the likelihood and
severity of potential earthquakes in a specific area. It is an essential step
in assessing the risks of earthquakes and developing effective measures to βeβðmm0 Þ
fM ðmÞ ¼ ; ðm0  m  mmax Þ (8)
reduce potential damage and loss of life. There are two method Deter- 1  eβðmm0 Þ
ministic Seismic hazard analysis and Probabilistic seismic hazard anal-
ysis for estimation Seismic Hazard of any region. DSHA method considers Where fM(m) is the probability density function for magnitude m, β ¼
2:303*b

429
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

3.2. Estimation of spatial uncertainty

r
fR ðrÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ; (9)
Lf r2  rmin 2

Where fR(r) is the probability density function for distance r, Lf is length

of fault in km, rmin is minimum source to site distance in km.

3.3. Estimation of temporal uncertainty

For the time uncertainty Poisson distribution model is used and the
probability of at least one exceedance in a period of t years is given by

P½N  1¼ 1eλm t : (10)

The probability of a particular ground motion parameter y exceeds a
certain value y*, for an earthquake of magnitude m, occurring at a given
distance r is given by:

P½y > y* jm; r¼ 1Fy ðy* Þ: (11)

In order to plot a seismic Hazard curve, all of the uncertainties should be

merged as shown below. Fig. 7. DSHA result due to different faults.

X
Ns ZZ
λ y* ¼ νi P½y > y* jm; rfMi ðmÞfRi ðrÞdmdr (12) controlling source because it produces the highest peak ground acceler-
i¼1 ation (PGA) compared to other sources as shown in the Fig. 8. The figure
explains the potential hazard associated to each seismic sources with
νi ¼ eαi βi m : (13) varying magnitudes up to the estimated maximum magnitude level.
Apart from Oldham fault, Lohiti and A1 faults can also produce high PGA
Where Ns is number of earthquake sources and λy* is the mean annual rate value for the study area.
of exceedance of particular value y*. In PSHA method spatial, temporal and magnitude uncertainties are
The seismic hazard of the site is typically described by using the peak considered. The PSHA results are generally represented by hazard curves.
acceleration (PGA) and spectral acceleration (SA). Fig. 6 shows a The seismic hazard curves are the plots that show the probability of
graphical representation of the flow chart of the methodology used in the experiencing a particular level of ground shaking at a given site over a
study. specified period of time. It is a fundamental tool used by seismologists
and engineers to assess the seismic hazard of a region and design struc-
4. Results and discussion tures that can withstand expected earthquake forces. The seismic hazard
curve is calculated by considering the seismicity of the region and the
In the present study, the seismic events are collected homogenized potential magnitude and distance of earthquakes that can occur. It takes
and declustered for the study region Guwahati city. The seismic sources into account the probability of earthquakes occurring at different levels
present in the study area are identified and the fault zones are charac- of ground shaking and the probability of the ground shaking intensity
terized by earthquake density around the faults. The seismicity param- being exceeded during a specified period of time. The seismic hazard
eter i.e., ‘a’ and ‘b’ for different sources have been estimated. The curve is typically plotted on a logarithmic scale, with the horizontal axis
maximum magnitude value (Mmax) for each source is calculated using the representing the ground motion parameter, such as peak ground accel-
previously explained methods. The seismicity parameters and the eration or spectral acceleration, and the vertical axis representing the
maximum magnitude for each seismic source are given in Table 2. After probability of exceedance over a given period of time, usually in per-
that, 8 faults out of 28 faults in the study region are selected that are centages or annual rates. The hazard curve for the present study is shown
responsible for generating higher ground motion at Guwahati. Appro- in Fig. 9. From the figure it can be clearly seen that the Oldham fault is
priate ground motion attenuation models are chosen after comparing the most contributing source along with the A1 fault and the Lohiti
with the past earthquake-strong motion data. For the study, (Ndma, Thrust for the Study area whereas, the Dauki Fault is the least contrib-
2011; Bajaj and Anbazhagan, 2019) equations are selected with an equal uting source.
weightage factor of 0.5. Then the DSHA analysis is performed and the The comparison of ground motion results from DSHA method and
results show that the Oldham fault is the controlling source for the PSHA method are shown in Table 3. In PSHA, PGA value have been
Guwahati region and can generate PGA value of 1.01g. The results for the estimated for the 2%, 5% and 10% for 50years and 100 years respec-
DSHA analysis are shown in Fig. 7. tively. It is clear from the Table 3, that Oldham fault is main controlling
In the DSHA analysis, every source near the city has been taken into source for Guwahati region, which can generate PGA 1.01g (DSHA
consideration. For this study, ground motion values resulting from each method) and PGA 0.51g (PSHA (2% in 50 years) method) respectively.
source have been examined, taking into account the maximum magni- Fig. 10 illustrates PGA values in the PSHA method for each fault at 2%,
tude earthquake associated with each source. Maximum Magnitude are 5%, and 10% probability of exceedance over 50 and 100 years. This
determined from historical earthquake data. The earthquake with the figure aids in estimating PGAs at different intensity levels (DBE and MCE)
highest credibility for a particular source is considered the controlling due to various faults. For MCE level earthquake A1 and the Oldham
earthquake. Faults produce the maximum Damage intensity i.e., 0.51g while for DBE
For example, in the Oldham fault zone, the maximum observed level earthquake only the Oldham Fault produces the maximum intensity
earthquake is the Shillong 1897 earthquake, which had a magnitude of i.e., 0.28g. For both the cases Dauki fault produces the least damage
(Mw 8.1). Based on this earthquake, the maximum magnitude for this intensity (see Fig. 10).
source has been estimated as 8.6. Therefore, this source is considered the

430
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Fig. 8. De-aggregation plot of different sources for varying magnitudes.

Fig. 10. The PGA values calculated for Guwahati at probabilities of exceedance
of 2%, 5%, and 10% over 50 and 100 years for each source.

5. Uniform hazard spectra

The graphical representation of spectral acceleration with different

time periods is known as Uniform Hazard Spectra (UHS). It is a tool used
by engineers and seismologists to evaluate the seismic hazard of a region
and design structures that can withstand the expected earthquake forces.
UHS is calculated by considering the seismic hazard at a specific site and
then estimating the response spectrum for that location. For response
spectrum analysis of structures, UHS is required. The UHS can be used to
evaluate the design of structures, such as buildings, bridges, and dams.
Engineers can use the UHS to determine the design forces that a structure
must be able to withstand and select appropriate materials and con-
struction methods. The UHS can also be used to evaluate the potential
damage to existing structures during an earthquake and to develop
strategies to mitigate that damage. The UHS have been plotted for 2475
years, 475 years, 4950 years, 950 years, 1975 years, and 975 years return
period for Guwahati city, shown Fig. 11. At period 0.2 s, spectral accel-
eration value is highest and then it decreases with increasing time period.
Fig. 9. Plot of seismic hazard curves for Guwahati City. Comparing different recurrence intervals, 4950years return period shows
maximum spectral acceleration and 475 years shows minimum spectral
The findings of the present study are compared with works of other acceleration as compared to others.
researchers and the comparison for 475 years return period is shown in Velocity response spectra refer to the graphical representation of the
Table 4. The results of the present study results shows higher results than response of a structure or system to ground motion, as a function of the
the Indian Standard Code (IS-1893:2016); (Bureau of Indian Standards, vibration frequency or time period. They are commonly used in earth-
New Delhi, 2016),evenly matches with the works of ((Bandyopadhyay quake engineering and seismic design to predict the response of struc-
et al., 2022; Das et al., 2016a; Raghukanth, 2010)) and are lesser than tures to earthquakes or other ground motions. In a velocity response
((Ghione et al., 2021; Bahuguna and Sil, 2020; Nath and Thingbaijam, spectrum, the horizontal axis represents the vibration frequency/time
2012; Sharma and Malik, 2006)). The differences in the results are due to period, while the vertical axis represents the maximum velocity response
the selection of different attenuation models and the change in seismicity of the structure to ground motion. Whereas, in displacement response
parameters due to change in earthquake catalogue. spectrum, the horizontal axis represents the natural time period of the
structures, while the vertical axis represents the maximum displacement
response of the structure to ground motion at each time period. These can

Table 3
Comparison of DSHA & PSHA results for Guwahati City.
CITY DSHA PSHA

PGA(g) 2% 50yr 5% 50yr 10%50yr 2% 100yr 5% 100yr 10%100yr Controlling Source

Guwahati 1.01 0.51 0.36 0.28 0.65 0.47 0.37 Oldham

431
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

Table 4
PGA value comparison with other studies for PSHA method (10% in 50 years).
Srijit et al. Raghukanth Das et al. Bahuguna and Sil Sharma and Malik Nath et al. Ghione et al. IS-1893 Present
(2022) (2010) (2016) (2020) (2006) (2012) (2021) (2016) Study

0.25g 0.23g 0.24g 0.46g 0.5g 0.66g 0.35g 0.18g 0.28g

Fig. 12 (b). Displacement response spectra For Guwahati.

Fig. 11. Uniform hazard spectra of Guwahati region for different Recur-
rence interval.
6. Conclusions

be generated using recorded ground motion data from earthquakes or by This paper presents a comprehensive study of seismic hazard assess-
using artificial ground motion records. The spectra can be used to ment for Guwahati city using a combination of earthquake data analysis
determine the potential damage or structural response of a building or and probabilistic and deterministic seismic hazard analysis methods. The
other structure to a given ground motion. Design codes often specify study includes the collection of earthquake data from various seismo-
target velocity response spectra or displacement response spectra that logical agencies, and the seismic source information comprises of the
structures must be able to withstand in order to be considered safe in a latest studies and Seismotectonic Atlas (SEISAT) data. The estimation of
given seismic hazard zone. The velocity spectra and the displacement seismicity parameters and maximum magnitude using the Gutenberg-
spectra for Guwahati city are shown in Fig. 12(a) and (b) respectively. Richter (G-R) relationship and results of seismic hazard analysis by
These response spectra will further be subjected modification with to both DSHA and PSHA method for Guwahati city of Assam state of India
select appropriate design parameters, such as the stiffness and damping have been presented in the form of controlling source, seismic hazard
of the structures. curves and uniform hazard spectra for updated seismicity data.
Furthermore, the study presents acceleration, velocity, and displace-
ment response spectra for the study region which are essential tools used
in structural and seismic engineering to understand and evaluate the
seismic performance of structures in the city. These spectra provide
valuable information about how ground motion affects buildings and
other infrastructure. So, the study is very much important to ensure the
public safety, infrastructure resilience, better urban planning, engineer-
ing design and risk management. Notably, our analysis has identified
relatively high PGA values for the study area. Consequently, we recom-
mend that the next step in research should involve incorporating site-
specific effects and local soil conditions to obtain surface-level hazard
assessments. This will provide essential site-specific design parameters
for creating earthquake-resistant structures. It's worth mentioning that,
according to Indian Standard Code (IS-1893:2016), the study region falls
under seismic Zone-V, with a zone factor of 0.36g. However, our obser-
vations reveal that the study area experiences higher PGA values than
expected. Therefore, this study's results can serve as critical inputs for
conducting ground response analysis, liquefaction analysis and seismic
micro-zonation studies in the area, enabling more efficient and cost-
effective designs for infrastructure within the study area.

Fig. 12 (a). Velocity response spectra for Guwahati.

432
S. Mishra et al. Natural Hazards Research 4 (2024) 423–433

 Peak ground acceleration (PGA) value obtained from deterministic assessment for northeast India Region. Pure Appl. Geophys. 173 (8), 2653–2670.
https://doi.org/10.1007/s00024-016-1333-9.
seismic hazard analysis method is 1.01g and controlling source is
Devi, A., Kalita, S., 2014. Seismic hazard analysis of Northeast India and its adjoining
Oldham fault. region. Int. J. Environ. Sci. 4 (5), 859–874.
 Peak ground acceleration (PGA) value obtained from probabilistic Geological Survey of India, 2000. Seismotectonic Atlas of India and its Environs.
seismic hazard analysis method are 0.51g, 0.28g, 0.65g and 0.37g for Ghione, F., Poggi, V., Lindholm, C., 2021. A hybrid probabilistic seismic hazard model for
Northeast India and Bhutan combining distributed seismicity and finite faults. Phys.
2% and 10% probability of exceedance in 50 years and 100 years Chem. Earth 123, 103029. Parts A/B/C.
respectively. Gupta, I.D., 2010. Response spectral attenuation relations for in-slab earthquakes in Indo-
 Spectral acceleration value is maximum at 0.2 s time period for all Burmese subduction zone. Soil Dynam. Earthq. Eng. 30 (5), 368–377.
Gutenberg, B., Richter, C.F., 1944. Frequency of earthquakes in California. Bull. Seismol.
different recurrence intervals. Soc. Am. 34 (4), 185–188.
Jaiswal, K., Gupta, S., Sinha, R., 2002. Estimation of maximum magnitude earthquakes in
We recommend that the next step in research should involve incor- Peninsular India using extreme value statistics. In: 12th Symposium on Earthquake
Engineering. IIT Roorkee, India.
porating site-specific effects and local soil conditions to perform site Joshi, A., Mohan, K., Patel, R.C., 2007. A deterministic approach for preparation of
characterization, site specific ground response analysis and obtain the seismic hazard maps in North East India. Nat. Hazards 43 (1), 129–146.
surface-level hazard assessments and liquefaction potential. These results Kanno, T., et al., 2006. A new attenuation relation for strong ground motion in Japan
based on recorded data. Bull. Seismol. Soc. Am. 96 (3), 879–897.
can be used as an input for seismic micro zonation and Hazard calcula- Kanth, S.T.G.R., Iyengar, R.N., 2006. Seismic Hazard Estimation for Mumbai City. Current
tion of north east India for economical and efficient design of structures. Science, pp. 1486–1494.
Kijko, A., 2004. Estimation of the maximum earthquake magnitude. Mmax, Pure Appl.
Geophys. 161, 1–27.
Funding declaration Kramer, S.L., 1996. Geotechnical Earthquake Engineering. Pearson Education India.
Kumar, Mishra, Satyaprakash, Arjun Sil, A., 2023. Seismic hazard analysis of North East
The authors received no specific grants or financial support from any India and hazard assessment of capital cities in the Region. Int. J. Reliab. Saf. 17 (2),
143–166.
funding agencies to carry out the present study. Kumar, A., et al., 2023. Seismic hazard analysis of Silchar city located in North East India.
Geomatics, Nat. Hazards Risk 14 (1), 2170831.
CRediT authorship contribution statement Kramer, L., 1996. In: Hall, W.J. (Ed.), S. (No Date) Geotechnical Earthquake Engineering.
University of Washington: Prentice-Hall International Series in Civil Engineering and
Engineering Mechanics.
Satyaprakash Mishra: Data curation, Formal analysis, Investigation, Mase, L.Z., 2018. Reliability study of spectral acceleration designs against earthquakes in
Methodology, Validation, Writing – original draft, Writing – review & Bengkulu City, Indonesia. Int. J. Technol. 9 (5).
Mase, L.Z., et al., 2021. Local site investigation and ground response analysis on
editing, Conceptualization. Aman Kumar: Formal analysis, Data cura- downstream area of Muara Bangkahulu River, Bengkulu City, Indonesia. Indian
tion, Methodology, Investigation. Arjun Sil: Conceptualization, Super- Geotech. J. 1–15.
vision, Writing – review & editing. Mase, L.Z., et al., 2023. Analysis of Ground Response and Potential Seismic Damage to
Sites Surrounding Cimandiri Fault, West Java, Indonesia. Natural Hazards, pp. 1–41.
Menon, A., et al., 2010. Probabilistic seismic hazard macrozonation of Tamil Nadu in
Declaration of competing interest Southern India. Bull. Seismol. Soc. Am. 100 (3), 1320–1341.
Misliniyati, R., et al., 2018. ‘Seismic hazard mitigation for Bengkulu coastal area based on
site class analysis. IOP Conf. Ser. Earth Environ. Sci. 212 (1), 012004’.
The authors declare that there is no conflict of interest. Nath, S.K., et al., 2012. Ground-motion predictions in Shillong region, northeast India.
J. Seismol. 16 (3), 475–488.
Nath, S.K., Thingbaijam, K.K.S., 2012. Probabilistic seismic hazard assessment of India.
Appendix A. Supplementary data Seismol Res. Lett. 83 (1), 135–149.
Ndma, 2011. Development of Probabilistic Seismic Hazard Map of India Technical
Supplementary data to this article can be found online at https:// Report. National Disaster Management Authority, p. 126.
Raghu Kanth, S.T.G., Iyengar, R.N., 2007. Estimation of seismic spectral acceleration in
doi.org/10.1016/j.nhres.2023.10.005. peninsular India. J. Earth Syst. Sci. 116 (3), 199–214.
Raghukanth, S.T.G., 2010. Estimation of seismicity parameters for India. Seismol Res.
References Lett. 81 (2), 207–217.
Sharma, M., et al., 2011. Ground-motion Prediction Equations Based on Data from the
Himalayan and Zagros Region to Cite This Version : HAL Id : Hal-00502687.
Anbazhagan, P., Kumar, A., Sitharam, T.G., 2013. Ground motion prediction equation Sharma, M.L., Malik, Shipra, 2006a. Probabilistic Seismic Hazard Analysis and Estimation
considering combined dataset of recorded and simulated ground motions. Soil of Spectral Strong Ground Motion. 4th International Conference on Earthquake
Dynam. Earthq. Eng. 53, 92–108. https://doi.org/10.1016/j.soildyn.2013.06.003. Engineering, Taipei, Taiwan, p. 15, 015.
Bahuguna, A., Sil, A., 2020. Comprehensive seismicity, seismic sources and seismic Sharma, M.L., Malik, Shipra, 2006b. Probabilistic seismic hazard analysis and estimation
hazard assessment of Assam, North East India. J. Earthq. Eng. 24 (2), 254–297. of spectral strong ground motion on bed rock in north east India. In: 4th International
Bajaj, K., Anbazhagan, P., 2019. Regional stochastic GMPE with available recorded data Conference on Earthquake Engineering, pp. 12–13.
for active region – application to the Himalayan region. Soil Dynam. Earthq. Eng. 126 Sil, A., Sitharam, T.G., Kolathayar, S., 2013. Probabilistic seismic hazard analysis of
(August), 105825. https://doi.org/10.1016/j.soildyn.2019.105825. Tripura and Mizoram states. Nat. Hazards 68 (2), 1089–1108.
Bandyopadhyay, S., Parulekar, Y.M., Sengupta, A., 2022. Generation of seismic hazard Sil, D.A., Bahuguna, A., 2018. Comprehensive seismicity, seismic sources and seismic
maps for Assam region and incorporation of the site effects. Acta Geophys. 70 (5), hazard assessment of Assam, North East India. J. Earthq. Eng.
1957–1977. Sitharam, T.G., Sil, A., 2014. Comprehensive seismic hazard assessment of Tripura and
Baro, O., Kumar, A., Ismail-Zadeh, A., 2018. Seismic hazard assessment of the Shillong Mizoram states. J. Earth Syst. Sci. 123 (4), 837–857.
Plateau, India. Geomatics, Nat. Hazards Risk 9 (1), 841–861. Somantri, A.K., et al., 2023. Analysis of ground response of Bandung Region subsoils due
Beitr, G., 1945. Frequency of earthquakes in California. Nature 156 (3960), 371. https:// to predicted earthquake triggered by Lembang Fault, West Java Province, Indonesia.
doi.org/10.1038/156371a0. Geotech. Geol. Eng. 41 (2), 1155–1181.
Bureau of Indian Standards, New Delhi, I, 2016. IS 1893 (2016). Indian Standard Criteria Tanapalungkorn, W., et al., 2020. Verification of attenuation models based on strong
for Earthquake Resistant Design of Structures IS 1893 (Part-I): 2016. ground motion data in Northern Thailand. Soil Dynam. Earthq. Eng. 133, 106145.
Das, R., Sharma, M.L., Wason, H.R., 2016a. Probabilistic seismic hazard assessment for Wells, D.L., Coppersmith, K.J., 1994. New empirical relationships among magnitude,
northeast India region. Pure Appl. Geophys. 173 (8), 2653–2670. rupture length, rupture width, rupture area, and surface displacement. Bull. Seismol.
Das, R., Sharma, M.L., Wason, H.R., 2016b. Probabilistic seismic hazard assessment for Soc. Am. 84 (4), 974–1002.
northeast India Region pure and applied geophysics probabilistic seismic hazard

433