The attenuation characteristics of NW Himalaya and NE India region has been estimated using recorded accelerograms. In this analysis frequency dependent attenuation of P-wave, S wave and coda wave was estimated and the scattering & intrinsic attenuation are separated using Wennerberg (1993) method. The frequency dependent relation of scattering attenuation (QS) and intrinsic attenuation (QI) have also been developed for these seismically active regions. The frequency dependent relations for NW Himalaya are Qα = (67 ± 0.22) f (0.96 ± 0.001), Qβ = (68 ± 0.29) f (0.95 ± 0.002), QC = (107 ± 0.10) f (1.06 ± 0.0003), QI = (95 ± 0.06) f (1.01 ± 0.0002) and QS = (306 ± 6.85) f (0.80 ± 0.009) for P-wave, S-wave, coda wave, intrinsic and scattering attenuation respectively. Similarly, for NE India region the frequency dependent relations are Qα = (89 ± 0.22) f (0.85 ± 0.001), Qβ = (92 ± 0.10) f (0.86 ± 0.0004), QC = (141 ± 0.34) f (1.02 ± 0.0009), QI = (120 ± 0.21) f (0.98 ± 0.0004) and QS = (306 ± 0.66) f (0.77 ± 0.0009) for P-wave, S-wave, coda wave, intrinsic and scattering attenuation respectively. The ratio of Qβ and Qα is greater than 1 for entire frequencies range considered in this analysis which indicates that both NW Himalaya and NE India regions are partially saturated with fluids. The estimated values of intrinsic (QI) and scattering (QS) attenuation shows that intrinsic attenuation is dominant in both NW Himalaya as well as NE India region. The important fact is that in both regions’ QC lies in between QI and QS, which agrees with theoretical and experimental measurements. The frequency dependent of Q’s in this study suggests that both regions are highly heterogenous and low value of Q0’s (i.e. Qα, Qβ and QC) for both regions exhibits high degree of seismicity.
Research Article
Attenuation characteristics of seismic waves beneath North-West Himalaya and North-East India region
https://doi.org/10.21203/rs.3.rs-2889932/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Seismic Attenuation
Body waves attenuation (Qα and Qβ)
Coda wave Attenuation (QC)
Intrinsic (QI) and Scattering (QS) attenuation
In addition to the source and site, another factor that affects the ground motions at surface is attenuation of the seismic waves while traveling in the medium. The amplitude of a seismic signal which travels away from its source decreases with increasing distance and is referred to as attenuation. There are various factors causing attenuation of seismic waves including geometrical spreading, scattering, multi-pathing, and anelasticity. For the propagation of seismic wave beneath the earth, all these four processes are important. The first three describe the elastic property of the seismic wave, in which, the energy during the propagation of seismic wave through a crustal rock is conserved, while in anelasticity attenuation, which is also known as intrinsic attenuation, seismic energy is converted into heat and there is loss of seismic energy. This dissipation of seismic energy takes place in term of heat which is due to the presence of folds, cracks, micro-fractures and grains in the earth, which leads the crustal structure weaken. Therefore, to understand the heterogeneity of Earth’s internal structure, attenuation of seismic wave plays important role. The attenuation of seismic wave is described by a quality factor Q, which is a dimensionless quantity and describes the decay of wave amplitude during propagation of the wave in a medium. In general, it was observed that the region which are seismically active generally have a high heat flow and more attenuating than that of colder regions that are seismically inactive or less active. The quality factor, Q, can be estimated using different parts of the seismograms i.e. using decay rate of direct P-waves (Qα), S-waves (Qβ) & coda waves (QC). The coda wave attenuation is the result of combined effect of scattering and intrinsic attenuation. Intrinsic attenuation is a result of small-scale crystal dislocation, frictional heating and viscous strain between rock and interstitial fluids (Goric and Muller 1987; Akinci and Eydogen 2000; Giampiccolo et al. 2006). On other hand, the scattering process involves an elastic process in which energy is redistributed by reflection, refraction and by changing the phases due the presence of irregularities in the medium (Sato and Fehler 1998). The relative contribution of scattering and intrinsic attenuation is helpful to understand the tectonics processes of any region (Del Pezzo et al., 1995; Bianco et al., 1999) such as faulting, folding, large-scale crustal movements associated with plate tectonics.
Several models were proposed to estimate the Q values of seismic waves. One such model was Aki and Chouet (1975) single backscattering model to estimate the coda-attenuation. This model is successfully used by many researchers in their study and has been correlate with geology and seismotectonic of the region (Rautian & Khalturin, 1978; Roecker et al., 1982; Singh & Herrmann, 1983; Pulli, 1984; Jin and Aki, 1988, 1989; Havskov et al., 1989; Ibanez et al., 1990; Pujades et al., 1991; Akinci et al., 1994; Latchman et al., 1996; Zelt et al., 1999). The coda Q has been estimated at different part of India region by Gupta et al. (1995) for Garhwal Himalayas, Mandal and Rastogi (1998) for Koyna region, Tripathi and Ugalde (2004) for southern India, Mandal et al. (2004) for Bhuj, Kutch area, Kumar et al., 2016 for Arunachal Himalaya and Mishra et al., 2020 for NW Himalaya. The quality factor for S-wave and P-wave has been estimated by different techniques but the most famous one is coda-normalization method (Aki, 1980), which was further extended by Yoshimoto et al., 1993. A number of studies have reported the attenuation of P-wave and S-wave in various region of the world (Hough and Anderson, 1988; Yoshimoto et al., 1993; Chung et al., 2001; Horasan & Boztepe-Guney 2004; Bindi et al., 2006; Hauksson et al., 2006). These body wave attenuation parameters have also been estimated in different parts of India by Sharma et al., 2007, 2009 for Koyna and Chamoli region, India; Padhy et al., 2010 for NE Himalaya; Chopra et al., 2010, for Gujarat, India, Padhy et al., 2011, Andaman region, India; Hazarika et al., 2013 for Sikkim Himalaya; Kumar et al., 2014 for Kinnaur Himalaya; Kumar et al., 2016 for Arunachal Himalaya and Mishra et al., 2020 for NW Himalaya. All these studies show that the attenuation is more in the seismically active region than in the stable region and helpful to evaluate the attenuation mechanism and heterogeneity present in the crust, under various geological regimes.
The attenuation of seismic wave is the combine effect of both scattering and intrinsic attenuation. However, it is very hard to separate out the relative contribution of these two attenuations from the total seismic attenuation. Wu (1985) developed a technique to separate the scattering and intrinsic attenuation from the total attenuation, which is based on radiative transfer theory. Wennerberg (1993) proposed a model to estimate the scattering attenuation (QS) and intrinsic attenuation (QI) which is based on Zeng (1991) scattering model. The objective of this study is to understand the attenuation properties beneath north-west Himalaya region and north-east India region by developing frequency dependent relationships for Q using coda wave, P-waves & S-waves using strong motion accelerograms and separate out the intrinsic and scattering attenuation, which is useful to known the type of attenuation mechanism exist in the study regions. The extended coda normalization method has been used to develop the relationships for Qα and Qβ while single back scattering model of Aki and Chouet (1975) has been used to obtain the relationship for Qc. The relationships for intrinsic attenuation (QI) and scattering attenuation (QS) have also been developed in order to measure the relative contribution of intrinsic and scattering attenuation with the help of Qβ and Qc. Wennerberg’s (1993) formulation has been used for this purpose. The outcome of this study will be helpful for understanding the seismotectonic and estimating the seismic hazard assessment of these regions.
In this section, we discuss the seismotectonic of the North-West Himalaya and North-East India region, separately.
2.1) NW Himalaya
On the basis of seismotectonic (as shown in Fig. 1), the NW Himalaya is divided as Kashmir Himalaya, Himachal Himalaya and Garhwal & Kumaon Himalaya. The evolution of Himalaya is understood by the tectonic models proposed by different research groups (Seeber et al., 1981; Ni and Barazangi, 1984; Yin and Harrison, 2000). Seeber (1981) in his model shows the existence of a gently dipping detachment thrust plane at depths 20–40 km. This detachment plane is known as Main Himalayan thrust (MHT), which separating the Indian plate and Eurasian plate. All the large/great earthquakes happened due to periodical stick slip displacements along this sub-horizontal decollement interface, which reaches the surface along the foothills and coincides with the Himalayan Frontal Thrust (HFT) fault (Nakata, 1989 and Yeats et al, 1992). The remaining two fault planes Main Boundary Thrust (MBT) and Main Central Thrust (MCT), meet this detachment at steep angles. Further Ni and Barazangi (1984) have suggested a model in which earthquakes occurred along this detachment plane and the ruptures of large/great earthquakes are propagated toward south of this detachment plane. All these Himalayan thrust planes serve as interfaces for various physiographic provinces of Himalaya. It is observed that MBT is the main potential source of seismic activity in the region. It comprises of straight and arc shaped faults about 150 km in length, which are separated by transverse faults. In North-West Himalaya region, intensive thrust type dislocations present from the preceding tectonic development stages occurred in the Late Quaternary period along the faults of the MBT system (Krishnaswamy et al., 1970; Rajal et al., 1986). But there is a difference between the geological evidence and the seismological results but still MBT is considered to be seismically active. The Main Frontal Thrust (MFT) makes the boundary of the alluvium deposits with the Siwaliks. At several places in this region, Neotectonic activity has been observed. The MCT is considered to be largest dislocation tectonic division in the Himalaya. In this, the strata of the crystalline rocks comprising the High Himalaya are thrusted over the metamorphic deposits of the lesser Himalaya. The ruptures in the MCT in the NW Himalaya are steep reverse faults with uplifting southern side. Also, a different alignment patterns for the processes of Quaternary activity in MCT are observed. These are created by different tectodynamic situation and are totally different with movements along the detachment thrust plane. The movement of faults in MCT is oblique having northeastern and northwestern strike.
Two big devastating earthquakes of magnitude greater than 7.5 have occurred in this region, one is 1905 Kangra earthquake of Mw 7.8 (Wallace et al., 2005) and the another, 2005 Kashmir earthquake of Mw 7.6. Besides this, many small, moderate and large earthquakes have occurred in this region. The seismicity of the NW Himalaya generally follows the NW-SE trend and the region between MCT and MBT shows high level of seismicity.
2.2) NE India region
Figure 2 shows the seismotectonic map of the North-East India region. The region lies at the junction of the Himalayan arc to the north and the Burmese arc to the east and is one of the six most seismically active regions of the world. The high seismicity of the North-East Himalayan region has been attributed to a complex tectonic province displaying association of the E–W trending Himalaya and the N–S trending Arakan-Yoma belt. The movement along the Po Chu fault, in the north-eastern part of the region is believed to have caused the 1950 Great Assam Earthquake of MW 8.7 (Ben-Menahem, 1974). The formation of the Shillong Plateau has been concerned with a pop-up tectonics, which is associated with the 1897 Great Earthquake of MW 8.1 (Bilham and England, 2001). On the southern end of the Kopili fault, the 1869 Cachar earthquake of MW 7.4 was generated. The Indo-Myanmar arc has been associated with the 1988 Manipur earthquake of Mw 7.2. Besides this, 20 large earthquakes (Ms > 7) and hundreds of moderate and small earthquakes have occurred in the region (Nandy, 2001), which have caused widespread destruction in the region.
The earthquakes in the Himalayan arc are due to the collision tectonics and mostly occurred at shallow depths. However, in Arunachal Himalaya earthquakes originates upto depth of 80 km (Kayal, 1993). The intersection point, where the Himalayan and Burmese arcs collided, also known as eastern syntaxis zone, is seismically active region. The great 1950 Assam earthquake has occurred in this region. On the other side, earthquakes in the Indo-Burma subduction zone are entrenched up to the depth ~ 200 km. Besides these active collision tectonics, there are Assam valley and Shillong Plateau, which are also seismic active regions and most of the earthquakes that have occurred in this region has depth ranges 30–35 km. Just east to Shillong Plateau, there is Mikir Hills, which is also seismically active region and separated by Kopili fault (Nandy, 2001) and have earthquakes originates upto the depth of 45–50 km. All the above facts show that the region is one of the highly seismic active regions of the world and most vulnerable to moderate and large earthquakes in the future.
In the present analysis we used the Strong Motion Accelerograms (SMA) of 31 earthquakes occurred in North-West Himalaya and 45 earthquakes occurred NE India region. They are recorded from the period December 2005 to 2014 having magnitude range varies from 3.4 to 5.8 for NW Himalaya, while for NE India region the magnitude range varies from 3.5 to 5.7 for the period of 2009 to 2015. The location of earthquakes and recording stations of NW Himalaya and NE India region are shown in Fig. 3(a) & (b). The data used were recorded by strong motion accelerometers networks maintained by Department of Earthquake Engineering, IIT Roorkee, under the National Strong Motion Instrumentation Project, sponsored by the Ministry of Earth Sciences, New Delhi. All the strong motion accelerographs consists of internal AC-63 GeoSIG triaxial forced balance accelerometers based on the latest Micro Electro-Mechanical Systems (MEMS) technology. All accelerographs have GSR-18 GeoSIG 18-bit digitizers with external GPS. The GSR-18 strong motion recorder has a dynamic range 200 SPS (Mittal et al., 2012) and operated on triggered mode. The main criteria for selecting the earthquakes is the Signal-to-Noise (SNR) should be greater than 3 for the events consider in this analysis and the accelerograms of those earthquakes have used for the analysis which have been recorded at least five seismic stations. The epicenter location, focal depth and magnitude (ML) has been obtained from the National Center for Seismology, Ministry of Earth Sciences (MoES), New Delhi. The vertical (Z-component) and horizontal components have been used for the analysis. The selected accelerograms have been filtered at six central frequencies (C.F.) of 1.5 Hz (1–2 Hz), 3 Hz (2–4 Hz), 6 Hz (4–8 Hz), 12 Hz (8–16), 24 (16–32 Hz), 36 Hz (24–48 Hz) using Butterworth band-pass (6 pole) filter. Figure 4 shows the example of original and filtered seismograms.
4.1) Estimation of Coda- Q (QC):
The values of QC have been estimated using the single back-scattering model proposed by Aki and Chouet (1975). This model states that, the coda waves are considered as a superposition of the secondary waves through single back-scattering of the primary body waves at randomly distributed heterogeneities present in the Earth’s crust and upper mantle. This shows that scattering is weak process and that outgoing waves are scattered only once before reaching the receiver. Using this assumption, the coda amplitudes, amplitude AC (f, t), in a seismogram can be expressed as a function of central frequency (f) over a narrow bandwidth signal and the lapse time t (Aki and Chouet, 1975)
where S(f) represents the source function at frequency f and independent of time and radiation pattern, a is geometrical spreading factor whose value for body waves is taken as 1. Taking logarithm on both sides, the Eq. (1) can be written as
$$\text{ln}\left[{A}_{c}\left(f,t\right) t\right]=lnS\left(f\right)- \frac{\pi ft}{{Q}_{c}}$$
2
The slope of this linear equation is (-πf/QC), from which the values of QC can be estimated.
Rautian and Khalturin (1978) modified this model and states that above relation is valid for lapse time greater than twice the S-wave travel time so that the data of the direct S-wave are avoided and for the validation of the model that the source of the earthquake and the receiver are coincident.
4.2) Estimation of Qα and Qβ:
It has been observed that the spectral amplitude of coda wave at lapse times greater than twice the S-wave travel time is proportional to the source spectral amplitude of the S waves within 100 km. This observation is the basis of the coda normalization method (Aki, 1980; Yoshimoto et al., 1993) and used for the estimation of Q for S-waves. Yoshimoto et al., (1993) extended this method further to estimate Q for P-waves by assuming that the earthquakes within a magnitude range have the same spectral ratio of P- to S-waves. The main advantage of using normalization of S-wave or P-wave spectra by coda is that it removes the effects of source, site and instrument and enables us to calculate the values of Qβ and Qα using single station data. The spectral amplitude of coda can be expressed as (Aki, 1980):
$${A}_{c}\left(f,{t}_{c}\right)= {S}_{s}\left(f\right)P(f,{t}_{c})G\left(f\right)I\left(f\right)$$
3
where f is the frequency, SS(f) is source spectral amplitude of the S-waves, P (f, tC) is the coda excitation factor, G(f) is site amplification factor and I(f) is the instrumental response.
For the direct S-wave, the spectral amplitude AS (f, r) can be written as (Yoshimoto et al., 1993)
$${A}_{s}\left(f,r\right)= {R}_{\theta \phi }{S}_{s}\left(f\right){r}^{-\gamma }\text{exp}\left[-\pi fr/{Q}_{\beta }\left(f\right){V}_{s}\right]G\left(f,Ѱ\right)I\left(f\right)$$
4
where, Rθφ is the radiation pattern, γ is the geometrical exponent, Qβ is the quality factor for S-waves, VS is the average velocity of S-wave and Ѱ is angle of incident of S-waves.
Dividing Eq. (4) by Eq. (3) and taking logarithm on both the sides, we get the expression (Yoshimoto et al., 1993):
$$\{\text{ln}\left[\frac{{A}_{s}\left(f,r\right){r}^{\gamma }}{{A}_{c}\left(f, {t}_{c}\right)}\right]{\}}_{r\pm \varDelta r}=-\frac{\pi fr}{{Q}_{\beta }{V}_{s}}+constant\left(f\right)$$
5
The left hand side of Eq. (5) represents the average for a hypocentral distance range r ± Δr. Eq. (4) is obtained under the assumption that Rθφ is neglected if we are averaging over many different fault plane solutions and also by averaging over many different focal plane solutions the ratio \(\frac{G\left(f,Ѱ\right)}{G\left(f\right)}\) becomes independent of Ѱ. Eq. (5) represents the linear equation of a straight line between \(\text{ln}\left[\frac{{A}_{s}\left(f,r\right){r}^{\gamma }}{{A}_{c}\left(f, {t}_{c}\right)}\right]\) and ‘r’, slope (-πf/ QβVS) of which gives the value of Qβ at a particular central frequency.
The similar expression for P-waves is given as (Yoshimoto et al., 1993):
$$\{\text{ln}\left[\frac{{A}_{P}\left(f,r\right){r}^{\gamma }}{{A}_{c}\left(f, {t}_{c}\right)}\right]{\}}_{r\pm \varDelta r}=-\frac{\pi fr}{{Q}_{\alpha }{V}_{P}}+constant(f)$$
6
where \({A}_{P}\left(f,r\right)\) is the amplitude spectra of the direct wave at hypocentral distance \(r\)(km) and VP is the wave velocity of P-wave. The slope of line gives the value of \({Q}_{\alpha }\).
4.3) Estimation of QS and QI
The QS and QI have been estimated using the formulation given by Wennerberg (1993) which is based on Zeng et al. (1991) model. The relationship between Qc, QI and QS is given by (Wennerberg, 1993):
$$1/{Q}_{c}= 1/{Q}_{I}+ \{1-2\delta (\tau \left)\right\}/{Q}_{S}$$
7
where
$$\delta \left(\tau \right)= \frac{0.72}{(4.44+0.738\tau )}-0.5$$
8
$$\tau =\frac{\omega t}{{Q}_{S}}$$
9
Ω and t are the angular frequency and lapse time respectively. QS and QI are expressed as follows (Wennerberg, 1993):
$$\frac{1}{{Q}_{S}}= \left\{\frac{1}{2\delta \left(\tau \right)}\right\}\left\{\frac{1}{{Q}_{d}}- \frac{1}{{Q}_{C}\left(\tau \right)}\right\}$$
10
$$\frac{1}{{Q}_{I}}= \left\{\frac{1}{2\delta \left(\tau \right)}\right\}\left\{\frac{1}{{Q}_{C}\left(\tau \right)}+\frac{[2\delta \left(\tau \right)-1]}{{Q}_{d}}\right\}$$
11
From equations (7) to (11), QS is obtained from the positive root of the quadratic equation:
$$4.44\left(\frac{1}{{Q}_{d}}-\frac{1}{{Q}_{C}}\right){Q}_{S}^{2}+\left[0.738\left(\frac{1}{{Q}_{t}}-\frac{1}{{Q}_{C}}\right)\omega t-5.88\right]{Q}_{S}-0.738\omega t=0$$
12
and QI is similarly obtained from the positive value of QS as:
$${Q}_{I}={Q}_{t}\left(\frac{{Q}_{S}}{{Q}_{S}-{Q}_{t}}\right)$$
13
The accelerograms have been filtered at six different central frequencies of 1.5 Hz (1–2 Hz), 3 Hz (2–4 Hz), 6 Hz (4–8 Hz), 12 Hz (8–16 Hz), 24 Hz (16–32 Hz) and 36 Hz (24–48 Hz) using Butterworth filter. Both vertical and horizontal components of selected events have been used for the analysis.
5.1) North-West Himalaya
Q α and Qβ (Body wave attenuation)
The plot of \(\{\text{ln}\left[\frac{{A}_{p}\left(f,r\right){r}^{\gamma }}{{A}_{c}\left(f, {t}_{c}\right)}\right]{\}}_{r\pm \varDelta r}\) versus hypocentral distance (r) for the six different central frequencies for Chamba station is shown in Fig. 5 (a). Figure 5 (b) shows the corresponding plot for S-waves. The estimated values along with average values of Qα and Qβ at different central frequencies are given in Tables 1 and 2 respectively. From these values of Qα and Qβ obtained in this study, it has been observed that the values of Qα is less than Qβ, which shows that P-wave attenuates faster than the S-waves. It also observed that both Qα and Qβ increases with the frequency. The average values of Qα varies from 87 to 1870 at frequency ranges from 1.5 Hz to 36 Hz and the average values of Qβ varies from 99 to 2240 at the same frequencies range. The power law \(Q={Q}_{0}{f}^{\eta }\) (where Q0 is the Q value at 1 Hz and η is the frequency dependent parameter) fitted to the estimated values of station are given in Table 3. The overall average power law relation for the NW Himalaya region is given as Qα = (60 ± 1.47) f (0.99 ± 0.006) and Qβ = (66 ± 1.42) f (0.97 ± 0.009) for P-wave and S-wave respectively.
| C.F. (Hz) | Bageshwar | Barkot | Chamba | Chamoli | Champawat | Dharamshala | Kapkot | Rampur | Munsiari | Average |
|---|---|---|---|---|---|---|---|---|---|---|
| 1.5 | 136 | 119 | 64 | 71 | 70 | 97 | 80 | 58 | 92 | 87 |
| 3 | 197 | 146 | 185 | 92 | 308 | 114 | 168 | 134 | 133 | 164 |
| 6 | 560 | 532 | 256 | 227 | 844 | 253 | 298 | 267 | 298 | 393 |
| 12 | 812 | 742 | 620 | 416 | 998 | 743 | 557 | 384 | 584 | 651 |
| 24 | 1266 | 1025 | 1063 | 1175 | 1670 | 1301 | 900 | 749 | 977 | 1125 |
| 36 | 2009 | 2131 | 2071 | 2116 | 1947 | 2051 | 1808 | 1149 | 1546 | 1870 |
| C.F. (Hz) | Bageshwar | Barkot | Chamba | Chamoli | Champawat | Dharamshala | Kapkot | Rampur | Munsiari | Average |
|---|---|---|---|---|---|---|---|---|---|---|
| 1.5 | 148 | 124 | 86 | 95 | 56 | 75 | 93 | 92 | 119 | 99 |
| 3 | 223 | 199 | 177 | 107 | 221 | 175 | 204 | 167 | 185 | 184 |
| 6 | 597 | 635 | 264 | 245 | 794 | 257 | 373 | 352 | 356 | 430 |
| 12 | 922 | 937 | 370 | 377 | 864 | 645 | 701 | 574 | 625 | 668 |
| 24 | 1325 | 1447 | 835 | 1277 | 1430 | 1297 | 1010 | 1165 | 1235 | 1225 |
| 36 | 2144 | 2390 | 2309 | 2818 | 1879 | 2537 | 2181 | 1829 | 2079 | 2241 |
| Station | Qα = Qofη | Qβ = Qofη |
|---|---|---|
| Bageshwar | (94 ± 0.92) f (0.96±0.004) | (106 ± 1.46) f (0.84±0.004) |
| Barkot | (76 ± 1.96) f (1.02±0.003) | (88 ± 1.21) f (0.93±0.006) |
| Chamba | (47 ± 1.60) f (1.04±0.005) | (55 ± 1.91) f (0.92±0.014) |
| Chamoli | (34 ± 1.23) f (1.09±0.007) | (41 ± 1.97) f (1.07±0.019) |
| Champawat | (82 ± 5.75) f (0.97±0.027) | (62 ± 4.63) f (1.03±0.03) |
| Dharamshala | (46 ± 1.12) f (1.08±0.009) | (47 ± 0.39) f (1.07±0.003) |
| Kapkot | (57 ± 0.29) f (0.92±0.002) | (68 ± 0.54) f (0.92±0.003) |
| Rampur | (45 ± 0.23) f (0.90±0.002) | (58 ± 0.33) f (0.94±0.002) |
| Munsiari | (58 ± 0.21) f (0.91±0.002) | (74 ± 0.30) f (0.90±0.003) |
The frequency dependent of Qα and Qβ for this region is comparable to another active region of the world and India as shown in Fig. 6(a) and Fig. 6(b) respectively. Figure 6(a) shows that at lower frequencies, the values of Qα for NW Himalaya are closer to the values obtained for Kachhch, India (Sharma et al., 2008) and Kinnaur Himalaya (Kumar et al., 2014). From Fig. 6(b) it is observed that at lower frequencies, the values of Qβ is closer to the values obtained for Kinnaur Himalaya (Kumar et al., 2014), Chamoli, India (Sharma et al., 2009), East-Central Iran (Ma’hood et al., 2009) and Kanto region of Japan (Yoshimoto et al., 1993). All these regions are highly seismic active regions and witnessed the large earthquakes in past. However, the values obtained for both Qα and Qβ for the present study is lower than the values obtained by the Imtiyaz et al. (2012) for same region. The estimated average frequency dependent relation given by Imtiyaz et al. (2012) are Qα = (97 ± 3) f (1.06±0.06) and Qβ = (127 ± 6) f (0.96±0.06), by using three seismic stations and 43 local events. Kumar et al., 2014 estimated these values for Kinnaur Himalaya as Qα = (47 ± 2) f (1.04±0.04) and Qβ = (86 ± 4) f (0.96±0.03). While Sharma et al., 2009 obtained Qα = (44 ± 1) f (0.82±0.02) and Qβ = (87 ± 3) f (0.71±0.03) for Chamoli region. Both Kinnaur Himalaya and Chamoli are the parts of NW Himalaya. The average value of Qα and Qβ for NW Himalaya obtained in this study is lie within the range of values obtained for other seismically active region. Figure 7 shows the comparison between the ratio \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) (at 1 Hz) estimated in this study at different frequencies with the other studies done at different regions like Kanto, Japan (Yoshimoto et al., 1993); Kinnaur Himalaya (Kumar et al., 2014); Kachchh, India (Sharma et al., 2008); Chamoli, India (Sharma et al., 2009); East-Central Iran (Ma’hood et al., 2009) and NW Himalaya (Imtiyaz et al., 2012). In the present study, it is observed that the ratio \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) ≥ 1 for the entire frequency range considered in this analysis. Imtiyaz (2012) estimated the ratio \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) relation for this region which lies between 1.33 to 1.82 for the frequency range 1–18 Hz. Kumar (2014) estimated the ratio \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) for Kinnaur Himalaya, which is subset of NW Himalaya region. They estimated that the ratio \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) lies in between 1.52 to 1.85 for the frequency range 1.5–12 Hz. The ratio values estimated by Sharma (2008) for Kachhch region varying from 0.99 to 1.53 for frequency range from 1.5–24 Hz. Similarly, Yoshimoto (1993) estimated the values for \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) for Kanto region of Japan which lies in between 1.2 to 2.9. For this study, the ratio value varies from 1.02 to 1.20 with varying frequency from 1.5–36 Hz. Thus, the values of \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) estimated in this study shows good agreement with the values obtained for other seismically active regions. The results of this analysis are also compared with the laboratory measurement of Qα and Qβ. Vassiliou et al., 1982 gives a general relation between Qα and Qβ for sedimentary rocks, which is summarized as Qα ≤ Qβ for partially saturated rocks, Qα ≥ Qβ for fluid saturated rocks and Qα = Qβ for dry rocks. In the present study we obtained Qα ≤ Qβ, which indicates the presence of partially saturated rocks and high degree of lateral heterogeneities in the North-West Himalaya region.
Coda-Q
The single backscattering model proposed by Aki and Chouet (1975) has been used to estimate the frequency-dependent attenuation characteristic of the coda waves in the NW Himalaya region using both vertical (Z) as well as horizontal component (N). The accelerograms have been filtered at six different frequency bands using a Butterworth band-pass filter. From the filtered accelerograms, the root-mean square (rms) amplitude of coda waves in a moving window length of 512 samples have been used to estimate QC. Figure 5 (c) shows the plot for the estimation of Qc for Chamba station and the estimated values of Qc for all stations considered in this study are given in Table 4. The average value of the QC has been obtained by averaging all individual QC values for the region. The estimated value of Q0 (Q at 1 Hz) and the frequency factor η along with standard deviation, for the lapse time of 30 seconds for vertical as well as for horizontal component are shown in Table 5. The average values of QC for all stations varies from 177 to 3979 for vertical (Z) component and from 169 to 4571 for horizontal component (N), for the frequencies corresponding to 1.5 Hz to 36 Hz, respectively. Hence, the QC values increases with an increase of frequency. The frequency dependent relationships for Qc are estimated as QC (Z) = (121 ± 0.28) f (0.98 ± 0.0009) QC(N) = (107 ± 0.10) f (1.06 ± 0.0003).
| Station | 1.5 Hz | 3 Hz | 6 Hz | 12 Hz | 24 Hz | 36 Hz | |
|---|---|---|---|---|---|---|---|
| Bageshwar | Z | 243 | 502 | 1365 | 1807 | 2024 | 3169 |
| N | 293 | 392 | 1182 | 1738 | 3421 | 4792 | |
| Barkot | Z | 229 | 338 | 1182 | 1330 | 2844 | 4270 |
| N | 183 | 268 | 723 | 1546 | 3096 | 4368 | |
| Chamba | Z | 94 | 336 | 803 | 1199 | 2524 | 5353 |
| N | 93 | 421 | 503 | 1486 | 2618 | 4695 | |
| Chamoli | Z | 156 | 295 | 513 | 1207 | 2613 | 3692 |
| N | 142 | 302 | 525 | 1069 | 2158 | 4312 | |
| Champawat | Z | 151 | 309 | 749 | 1006 | 2983 | 4438 |
| N | 97 | 278 | 613 | 1042 | 2758 | 5333 | |
| Dharamshala | Z | 123 | 267 | 722 | 1114 | 1688 | 5211 |
| N | 169 | 287 | 716 | 1873 | 1781 | 7140 | |
| Kapkot | Z | 161 | 313 | 528 | 1234 | 2200 | 3246 |
| N | 168 | 384 | 501 | 1121 | 2220 | 3213 | |
| Rampur | Z | 198 | 390 | 910 | 1881 | 6938 | |
| N | 181 | 368 | 1007 | 2018 | 9804 | ||
| Munsiari | Z | 240 | 302 | 493 | 918 | 1820 | 2459 |
| N | 193 | 285 | 530 | 994 | 1924 | 2719 |
| Station | QC (Z) = Qofη | QC (N) = Qofη |
|---|---|---|
| Bageshwar | (227 ± 6.39) f (0.76±0.011) | (196 ± 2.59) f (0.87±0.005) |
| Barkot | (154 ± 2.77) f (0.93±0.007) | (106 ± 0.66) f (1.05±0.003) |
| Chamba | (75 ± 1.54) f (1.16±0.008) | (77 ± 2.07) f (1.15±0.01) |
| Chamoli | (96 ± 0.25) f (1.02±0.001) | (92 ± 0.33) f (1.02±0.001) |
| Champawat | (98 ± 0.89) f (1.05±0.004) | (65 ± 0.50) f (1.19±0.003) |
| Dharamshala | (83 ± 2.12) f (1.07±0.01) | (98 ± 4.16) f (1.08±0.02) |
| Kapkot | (107 ± 0.17) f (0.95±0.006) | (118 ± 0.68) f (0.91±0.002) |
| Rampur | (104 ± 1.62) f (1.25±0.008) | (86 ± 2.51) f (1.30±0.016) |
| Munsiari | (145 ± 1.07) f (0.77±0.003) | (122 ± 0.28) f (0.86±0.001) |
Figure 8 shows the comparison of QC values of present study with other studies of worldwide and India. It is observed that the values of QC estimated in this study shows similar trend as other active tectonic regions like Italy, Norway, South Central Alaska, South Korea, Greece and Garhwal-Kumaun Himalaya etc. and lies in between them. In this analysis the value of Q0 is low and the frequency parameter (η) of the NW Himalaya is obtained closed to unity, which indicates that the upper lithosphere of the region is more heterogeneous and highly seismically active. Kumar et al. (2005) determined the attenuation of NW Himalaya region and obtained the relation QC = 158 f 1.05 while using 36 local earthquakes which is recorded at three seismic stations. Mukhopadhaya et al. (2008) estimated the relationship for the Garhwal-Kumaun Himalaya as QC = 122 f 0.98 at lapse time 40 s and Gupta et al. (1995) estimated QC = 126 f 0.95 for Garhwal Himalaya, both of these are the subset of NW Himalayan region.
To understand the attenuation mechanism of NW Himalayan region we have known the geology of the region. The NW Himalaya region is characterized by its complexity, which is represented by a deformed and collision-produced lithologies. This is the most seismically active region of the great Himalayan arc. The region is composed of Sedimentary rocks, which are obtained from the uplift magmatic rocks and their subsequent erosion. It is also composed of deep-sea and flysch sedimentary rocks, blue schist metamorphic rocks, ultra-basic and submarine volcanic rocks, and plutonic intrusions. The igneous rocks are consisting of gabbro’s, diorites and granites. The region also constitutes of Mica schist, quartzite, paragneiss, migmatite, and leucogranite bodies. All these shows the multiphase deformation event, firstly at normal geothermal gradient conditions and then low pressure and high temperature conditions (Sorkhabi, 1999). The metamorphism is observed by mineral accumulations of upper greenschist to amphibolite facies deformation. Besides this Proterozoic and lower Paleozoic sedimentary rocks also found in this region. These rocks are metamorphosed to greenschist facies. The region also consists of Paleocene-Eocene limestone, shale (the Subathu formation), synclinal (klippen), gneiss, mica schists and anticlines outliers of Greater Himalayan metamorphic rocks. The region also consists of clastic sediments, which were formed during the uplift and erosion of the Himalaya and deposited by rivers in a foreland basin. These above facts show that the formation of upper lithosphere of the NW Himalaya is highly heterogenous and having complex geometry, which makes this region more attenuative, as shown in the present study where we obtained low Q-values.
The estimated values of QC are more than those of Qβ for the frequencies range considered in this study. This observation satisfies Zeng et al (1991) model which shows that the effects of intrinsic and scattering attenuation combine in such a way that QC is greater than the Qβ. Similar results have been reported for the regions of India (Padhy et al., 2010; Hazarika et al.,2013; Kumar et al., 2014, Kumar et al., 2016 etc.).
Q I and QS
Wennerberg (1993) approach has been used to estimate QI and QS values for both vertical and horizontal component separately. The estimated average values of QI and QS are given in Tables 6. The average value of QI is varying from 148 at 1.5 Hz to 3515 at 36 Hz. Similarly, the average value of QS varies from 462 at 1.5 Hz to 7344 at 36 Hz. The power laws for QI and QS are estimated as QI (Z) = (102 ± 0.22) f (0.97 ± 0.0009); QI (N) = (95 ± 0.06) f (1.01 ± 0.0002) and QS (Z) = (238 ± 3.00) f (0.90 ± 0.005); QS (N) = (306 ± 6.85) f (0.80 ± 0.009). Figure 9 shows the variation of Qα, Qβ, QC, QI and QS with frequency. It has been observed that QC is close to QI (e.g. Frankel and Wennerberg, 1987; Matsunami, 1991) and found to be intermediate between QI and QS at lower frequencies and close to QI at higher frequencies (Mayeda et al., 1992). This is consistent agrees with the theoretical as well as the laboratory experiments of measuring the QI and QS. From the present study, the higher values of QS as compare to QI indicates that intrinsic attenuation is dominated over scattering attenuation in the NW Himalaya region. Also, the seismic albedo (Bo) (Wu, 1985) is an important parameter which is helpful to know the attenuation mechanism in the region; if Bo ≤ 0.5 then intrinsic attenuation is dominate in the region and vice-versa (Wu and Aki, 1988). In this study it has been observed that Bo is less than 0.5 in the frequency range 1.5–36 Hz (as shown in Table 6), which also shows that intrinsic attenuation is dominate in entire NW Himalaya region.
| C.F (Hz) | QI | QS | Bo |
|---|---|---|---|
| 1.5 | 148 | 462 | 0.243 |
| 3 | 279 | 580 | 0.325 |
| 6 | 626 | 1320 | 0.322 |
| 12 | 1093 | 1837 | 0.373 |
| 24 | 2166 | 3539 | 0.380 |
| 36 | 3515 | 7345 | 0.324 |
5.2) North-East India
Qα and Qβ (Body wave attenuation)
The body wave attenuation i.e. Qα and Qβ (quality factor of P-wave and S-wave) are estimated using extended coda normalization method proposed by Yoshimoto (1993). The rms amplitudes for P and S waves have been picked from the filtered accelerograms and then normalized by the coda wave amplitude. We used vertical component of the filtered accelerograms for P wave and horizontal component of filtered accelerograms for S wave for this purpose. The plot of ln [(AP or AS/AC) r] versus the hypocentral distance r (km) at six different central frequencies for Guwahati station are shown in Fig. 10(a) and (b). The gradient of the least squared fitted line of these plots gives us the values of Qα and Qβ respectively. The estimated values of Qα and Qβ are given in Table 7 and Table 8 respectively. The average values of Qα varies from 131 to 2046 for the frequencies 1.5 Hz to 36 Hz. And the average values of Qβ varies from 139 to 2159, corresponding to the central frequencies from 1.5 Hz to 36 Hz respectively. All these values of Qα and Qβ increases with increase in the frequency at all stations, which shows that Qα and Qβ have frequency dependent nature for the region. Table 9 gives the frequency dependent relationships Q = Qofη for of Qα and Qβ for each station. The average power law relations Qα and Qβ for the NE India region are estimated as Qα = (89 ± 0.22) f (0.85 ± 0.001) and Qβ = (92 ± 0.10) f (0.86 ± 0.0004).
| Station | 1.5 Hz | 3 Hz | 6 Hz | 12 Hz | 24 Hz | 36 Hz |
|---|---|---|---|---|---|---|
| Bongaigaon | 103 | 173 | 256 | 387 | 900 | 1466 |
| Golaghat | 167 | 232 | 263 | 397 | 888 | 1507 |
| Guwahati | 96 | 366 | 719 | 657 | 1432 | 1956 |
| Jorahat | 97 | 247 | 446 | 782 | 1293 | 2074 |
| Kokrajhar | 69 | 126 | 183 | 545 | 1172 | 2063 |
| Mokochung | 155 | 255 | 419 | 1297 | 2453 | 3164 |
| Naugaon | 261 | 264 | 831 | 395 | 1094 | 2231 |
| Tezpur | 103 | 232 | 351 | 561 | 991 | 1825 |
| Tura | 107 | 140 | 164 | 379 | 672 | 1368 |
| Ziro | 151 | 177 | 653 | 1087 | 1925 | 2803 |
| Average | 131 | 221 | 429 | 649 | 1282 | 2046 |
| Station | 1.5 Hz | 3 Hz | 6 Hz | 12 Hz | 24 Hz | 36 Hz |
|---|---|---|---|---|---|---|
| Bongaigaon | 126 | 146 | 299 | 478 | 1525 | 1837 |
| Golaghat | 157 | 243 | 279 | 599 | 1206 | 2092 |
| Guwahati | 134 | 325 | 826 | 1001 | 1773 | 2018 |
| Jorahat | 89 | 177 | 330 | 871 | 1356 | 2228 |
| Kokrajhar | 96 | 117 | 227 | 518 | 1057 | 1958 |
| Mokochung | 153 | 282 | 427 | 1122 | 1951 | 3113 |
| Naugaon | 124 | 143 | 391 | 585 | 906 | 1595 |
| Tezpur | 111 | 207 | 394 | 622 | 937 | 1782 |
| Tura | 145 | 199 | 216 | 493 | 1385 | 2393 |
| Ziro | 252 | 274 | 777 | 1239 | 2109 | 2576 |
| Average | 139 | 211 | 417 | 753 | 1421 | 2159 |
| Station | Qα = Qo fη | Qβ = = Qofη |
|---|---|---|
| Bongaigaon | (67 ± 0.70) f (0.81±0.004) | (69 ± 1.35) f (0.90±0.008) |
| Golaghat | (104 ± 2.57) f (0.67±0.01) | (94 ± 1.81) f (0.80±0.008) |
| Guwahati | (104 ± 4.69) f (0.84±0.018) | (125 ± 2.78) f (0.83±0.009) |
| Jorahat | (79 ± 1.24) f (0.92±0.002) | (58 ± 2.33) f (1.01±0.003) |
| Kokrajhar | (38 ± 1.36) f (1.08±0.006) | (48 ± 2.82) f (0.98±0.007) |
| Mokochung | (90 ± 1.00) f (1.01±0.005) | (96 ± 0.51) f (0.95±0.002) |
| Naugaon | (126 ± 9.41) f (0.71±0.029) | (77 ± 0.97) f (0.82±0.005) |
| Tezpur | (79 ± 5.22) f (0.84±0.003) | (83 ± 3.90) f (0.82±0.002) |
| Tura | (62 ± 3.30) f (0.76±0.02) | (72 ± 3.37) f (0.89±0.019) |
| Ziro | (89 ± 1.69) f (0.98±0.008) | (158 ± 2.17) f (0.80±0.006) |
Figure 11 (a) and (b) shows the comparison of the relations for Qα and Qβ estimated in the present analysis with those of obtained for other seismic active regions of the world. From the Fig. 11 (a), we observed that the average values of Qα for NE India region show similar trend as shown by the Kachchh, India (Sharma et al., 2008), Sikkim Himalaya (Hazarika et al., 2013) and Arunachal Himalaya (Kumar et al., 2016). All these regions are highly seismic active. While in Fig. 11 (b), the average values of Qβ is low as compared to other active region except the East-Central part of Iran and the Kanto region of Japan, whose values are comparable to the values of NE India region. These comparisons of values of body wave attenuation show that NE India is highly seismic active region.
Figure 12 shows the comparison between the ratio \(\frac{{Q}_{\beta }}{{Q}_{\alpha }}\) estimated in this study at different frequencies with the other studies done at different regions like Kanto, Japan (Yoshimoto et al., 1993); Kachchh, India (Sharma et al., 2008); East-Central Iran (Ma’hood et al., 2009), Sikkim Himalaya (Hazarika et al., 2013) and Arunachal Himalaya (Kumar et al., 2016). It is observed that the values of Qβ/Qα of present study show similar trend as that of study done for Arunachal Himalaya (Kumar et al., 2016) i.e. at higher frequency the values of Qβ/Qα falls down and approaches to unity. The ratio of Qβ/Qα varies from 0.97 to 1.16 with average of ~ 1.1 for the frequencies considered in the present study. The observed value of Qβ/Qα > 1 may be due to scattering from shallow heterogeneities in the crust beneath the NE India region. The results of this analysis are also compared with the laboratory measurement of Qα and Qβ. The laboratory experiment (Toksoz et al., 1979, Mochizuki, 1982 and Winkler & Nur, 1982) shows that the ratio Qβ/Qα > 1 for dry rocks and Qβ/Qα< 1 for water saturated rocks. The ratio Qβ/Qα found in this study indicates the presence of partially saturated rock in the crust of NE India region.
Coda-Q
The root-mean squared (rms) amplitudes of coda waves of filtered accelerograms in a window length of 512 samples and lapse time window of 30 sec have been used to estimate coda-Q using single scattering method given by Aki and Chouet (1975). Figure 10 (c) shows the plot of ln [Ac (f, t) t] with lapse time t for Guwahati station along with least-square fitted line. The slope of least square fitted line gives the value of Qc. Table 10 shows the estimated values of QC at different central frequencies for the stations used in this study for vertical as well as horizontal components. We note that QC shows show the frequency dependence and it increases with frequency. The average values of QC for all stations vary from 219 to 6151 for vertical component and from 201 to 5720 for horizontal component, for the frequencies corresponding to 1.5 Hz to 36 Hz, respectively. Table 11 shows the average values of QC in terms of power law i.e. QC = Q0fη, which also indicates that QC increases with frequency. The average QC relationship for the NE India region is QC (Z) = (136 ± 0.24) f (1.04 ± 0.0007) and QC (N) = (141 ± 0.34) f (1.02 ± 0.0009) for vertical and North-South components. The obtained value of η being greater than 1, which also indicates a strong dependence of QC on frequency. The Fig. 13 shows the comparison of QC values obtained for NE India region with results from other active regions of world, which shows that the QC value of NE India is similar to another active tectonic region of the world. These values lie in between other studies done in various seismic active regions of India as well in the world such Andaman region (Padhy et al., 2010), Sikkim Himalaya region (Hazarika et al., 2013), Arunachal Himalaya (Kumar et al., 2016), Garhwal & Kumaun Himalaya (Mukhopadhaya et al.,2008), Kachhch (Sharma et al., 2008), Italy (Tuve et al., 2006) & Turkey (Akinei et al., 1996). From the Fig. 13, it is observed that the coda attenuation trend for NE India lies closer to Kachchh, India and Arunachal Himalaya; both these regions are highly seismic active regions and Arunachal Himalaya is subset of NE India region. These results indicate that NE India region is highly seismically active region.
| Station | 1.5 Hz | 3 Hz | 6 Hz | 12 Hz | 24 Hz | 36 Hz | |
|---|---|---|---|---|---|---|---|
| Bongaigaon | Z | 180 | 233 | 847 | 1265 | 3199 | 5863 |
| N | 274 | 363 | 604 | 958 | 3393 | 6032 | |
| Golaghat | Z | 343 | 384 | 388 | 1078 | 3800 | 4925 |
| N | 208 | 397 | 630 | 1383 | 3509 | 4845 | |
| Guwahati | Z | 328 | 699 | 1296 | 2084 | 4124 | 5730 |
| N | 248 | 641 | 1047 | 1509 | 3408 | 5281 | |
| Jorahat | Z | 149 | 645 | 868 | 1731 | 3845 | 6603 |
| N | 189 | 526 | 688 | 1439 | 3301 | 6573 | |
| Kokrajhar | Z | 134 | 215 | 607 | 2001 | 3558 | 5832 |
| N | 134 | 256 | 684 | 1524 | 3651 | 6048 | |
| Mokochung | Z | 293 | 355 | 1163 | 1756 | 3184 | 7502 |
| N | 283 | 356 | 893 | 1568 | 3208 | 7097 | |
| Naugaon | Z | 245 | 341 | 782 | 1502 | 3184 | 6754 |
| N | 192 | 436 | 923 | 1499 | 2689 | 5250 | |
| Tezpur | Z | 166 | 547 | 1325 | 1890 | 3242 | 4685 |
| N | 172 | 553 | 1230 | 1875 | 3063 | 4420 | |
| Tura | Z | 145 | 309 | 991 | 1936 | 3967 | 8168 |
| N | 100 | 627 | 1020 | 2104 | 4166 | 7296 | |
| Ziro | Z | 203 | 250 | 1137 | 2167 | 3436 | 5453 |
| N | 209 | 479 | 1761 | 2143 | 4044 | 4360 |
| Station | QC (Z) = Q0fη | QC (N) = Q0fη |
|---|---|---|
| Bongaigaon | (94 ± 1.75) f (1.12±0.008) | (132 ± 5.37) f (0.98±0.016) |
| Golaghat | (146 ± 11.47) f (0.92±0.031) | (125 ± 0.79) f (1.01±0.003) |
| Guwahati | (248 ± 0.41) f (0.88±0.0007) | (195 ± 1.62) f (0.91±0.003) |
| Jorahat | (126 ± 2.87) f (1.09±0.009) | (128 ± 1.90) f (1.04±0.006) |
| Kokrajhar | (70 ± 0.89) f (1.25±0.005) | (75 ± 0.13) f (1.22±0.0007) |
| Mokochung | (160 ± 3.69) f (1.01±0.009) | (147 ± 2.58) f (1.02±0.007) |
| Naugaon | (128 ± 1.69) f (1.04±0.005) | (139 ± 0.79) f (0.98±0.002) |
| Tezpur | (156 ± 4.02) f (0.99±0.01) | (162 ± 3.54) f (0.96±0.008) |
| Tura | (88 ± 0.54) f (1.24±0.002) | (98 ± 4.36) f (1.22±0.018) |
| Ziro | (117 ± 3.48) f (1.09±0.012) | (183 ± 6.81) f (0.96±0.015) |
In the present analysis we obtained low Q-value’s which suggests that the upper lithosphere of NE India region is heterogenous and very attenuative. The high attenuative property of NE India is understood by complex geological formation, as geology plays important role in understanding the heterogeneity and seismicity of any region, which ultimately helps in seismic hazard assessment of the region. The whole region is overlapped by a number of active faults & thrusts and large-scale thrust movements have occurred from the northeast to the southeast direction, which resulted in crustal shortening. About 70% of the area of the North-East India is covered by Cenozoic rocks. About one-fourth area of the entire North-East Himalaya region is occupied by the Himalayan ranges and their foothills. These Himalayan ranges consists of some unclassified rocks from late Proterozoic to early Paleozoic in age. These rocks consist of low-grade metamorphic in the southern part to high-grade schists toward the top of the mountain. The NE India region also composed of grey sandstones, greenish shales, coarse-grained sandstones, pebble-cobble beds, lower facies of the sandstone & upper clay horizon. The Shillong Plateau consist of large body of rocks known as Khasi greenstone (amphibolite), which mainly occur along a shear zone. Besides this it also consists of quartzite, phyllites and pebbly conglomerate, deposited over the gneissic basement during the Proterozoic period (Sarma et al., 1998). The region is covered by a layer of Ganga-Brahmaputra alluvium and formed a basin known as Bengal basin, which has a thick Quaternary sediment of about 15–17 km thickness (Singh et al., 2016). Thus, due to this complicated geological formation, and high degree of heterogeneity or high attenuative property, the region witnessed large to moderate earthquakes every year.
Estimation of Q I and QS
We note that QC > Qβ for the frequency range 1.5 Hz to 36 in the region. This implies that the attenuation mechanism in the region follows Zeng et al., (1991) model which says that the effect of intrinsic and scattering attenuation combines in such a manner that QC should be greater than Qβ. The total attenuation of a region is a combined effect of intrinsic attenuation (QI) and scattering attenuation (QS). To separate the contribution of the scattering and intrinsic attenuation, Wennerberg (1993) approach has been used as discussed in the methodology section. The values of QI and QS are estimated from QC and Qβ. In this analysis, we estimated QI and QS for both vertical as well horizontal components. The average values of QI and QS are given in the Table 12. The average values of QI vary from 183 at 1.5 Hz to 4274 at 36 Hz and average values of QS vary from 412 at 1.5 Hz to 4527 at 36 Hz. The results also show that both QI and QS are frequency dependent and their power law equation has been estimated as QI = (120 ± 0.21) f (0.98 ± 0.0004) and QS = (306 ± 0.66) f (0.77 ± 0.0009). Figure 14 shows that QC lies in between QI and QS and the values of QC is relatively closer to QI at low frequencies. Also, from the Table 12, it has been observed that the seismic albedo (Bo) values varies from 0.307 to 0.486 for the given frequency ranges. This shows that Bo values is less than 0.5 for entire frequency ranges consider in this analysis. Hence, the higher value of Qs and low value of Bo (≤ 0.5) shows that intrinsic attenuation mechanism is more dominated than scattering attenuation in NE India region.
| C.F. (Hz) | QI | QS | Bo |
|---|---|---|---|
| 1.5 Hz | 183 | 412 | 0.307 |
| 3 Hz | 341 | 653 | 0.343 |
| 6 Hz | 721 | 1187 | 0.378 |
| 12 Hz | 1290 | 2093 | 0.381 |
| 24 Hz | 2667 | 3396 | 0.44 |
| 36 Hz | 4274 | 4527 | 0.486 |
The present study has emphasized on the frequency dependent attenuation relationships for seismically active regions of Himalaya – NW and NE regions. The frequency dependent relationships for different Q – Qα, Qβ, Qc, QI and QS have been developed for the NW region of Himalaya. The estimated quality factors are dependent on frequency and estimated as : Qα = (67 ± 0.22) f (0.96 ± 0.001), Qβ = (68 ± 0.29) f (0.95 ± 0.002), and QC = (107 ± 0.10) f (1.06 ± 0.0003). Similarly, for NE India region the estimated average values of different types of Q are: Qα = (89 ± 0.22) f (0.85 ± 0.001), Qβ = (92 ± 0.10) f (0.86 ± 0.0004) and QC = (141 ± 0.34) f (1.02 ± 0.0009). The ratio Qβ/Qα for NW Himalaya and NE India region is found to be more than 1. Both the region has observed that QC > Qβ is consistent with that the Zeng et al (1991) model. The intrinsic and scattering attenuation of NW Himalaya region are determined as QI = (95 ± 0.06) f (1.01 ± 0.0002) and QS = (306 ± 6.85) f (0.80 ± 0.009) whereas for NE India region QI = (120 ± 0.21) f (0.98 ± 0.0004) and QS = (306 ± 0.66) f (0.77 ± 0.0009). The results of this analysis show that in both NW Himalaya region and NE India region intrinsic attenuation is more dominated than that of the scattering part of the total seismic attenuation for these two regions. However, if we see all Q-values of both regions we observed that the estimated Q-values for NW Himalaya has been less as compared to the NE India region, which shows that NW Himalaya region exhibit more attenuative and heterogenous property as compared to the NE India region, although both regions are consider to be highly seismically active region of the world.
The attenuation relationships estimated in this study are useful for near-field simulation of earthquake ground motions, estimating the earthquake source parameter of the region and are useful to understand the physical properties of the propagation of seismic wave in both the regions. Thus, attenuation study of both the regions are helpful to evaluate the seismic hazard assessment of these regions.
Data availability
All data were obtained from (PESMOS) https://pesmos.org/
Acknowledgements
The authors would like to thanks their respective Institutions for support and encouragement to conduct this research. The first author express thanks to the Director, NCS, MoES, New Delhi for their constant support and provide facilities for carrying out this study. The authors are thankful to Department of Earthquake Engineering, IIT Roorkee for providing the data of Indian strong motion network through PESMOS. Some of figures in this work are produced by using the Generic Mapping Tool (GMT) package (Wessel and Smith, 1998).
Conflict of interest
The authors declare no competing interests.
Funding
The authors declare that no funds, grants, or other support were received during the preparation of this manuscript.
Authors’ Contributions
All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by Vikas Kumar, Dinesh Kumar and Sumer Chopra. The first draft of the manuscript was written by Vikas Kumar and other authors commented on previous versions of the manuscript. All authors read and approved the final manuscript
- Aki, K. and Chouet, B. (1975) Origin of coda waves: Source, attenuation and scattering effects, J. Geophys. Res. 80, 3322–3342.
- Aki, K. (1980) Attenuation of shear waves in the lithosphere for frequencies from 0.05 to 25 Hz, Phys. Earth Planet. In. 21, 50–60.
- Akinci, A., Taktak, A.G., Ergintav, S., 1994. Attenuation of coda waves in Western Anatolia. Phys. Earth Planet Inter. 87, 155–165.
- Akinci, A. and Eyidogan, H. (1996) Frequency-dependent attenuation of S and coda waves in Erzincan region (Turkey), Phys. Earth Planet. Inter., 97: 109-119.
- Akinci A, Eydogen H (2000) Scattering and anelastic attenuation of seismic energy in the vicinity of north Anatolian fault zone, eastern Turkey. Phys Earth Planet Inter 122:229–239.
- Ben-Menahem, A., Aboodi, E., and Schild, R. (1974) The source of the great Assam earthquake—an interplate wedge motion. Physics of the Earth and Planetary Interiors, 9(4), 265-289.
- Bianco F, Castellano M, Del Pezzo E, Ibañez JM (1999) Attenuation of short-period seismic waves at Mt. Vesuvius, Italy. Geophys J Int 138:67–76.
- Bilham, R., England, P. (2001) Plateau pop-up during the 1897 Assam earthquake. Nature (London);410: 806–9.
- Bindi D, Parolai S, Grosser H, Milkereit C, Karakisa S (2006) Crustal attenuation characteristics in northwestern Turkey in the range from 1 to 10 Hz. Bull Seismol Soc Am 96:200–214.
- Chopra S, Kumar D, Rastogi BK (2010) Attenuation of high frequency P and S waves in the Gujarat Region, India. Pure Applied Geophys 168:797–813.
- Chung, T.W. & Sato, H., 2001. Attenuation of high frequency P and S waves in the crust of southeastern South Korea, Bull. seism. Soc. Am., 91,1867–1874.
- De Lorenzo, S., Bianco, F. & Del Pezzo, E. (2013a) Frequency dependent Q-alpha and Q-beta in the Umbria-Marche (Italy) region using a quadratic approximation of the coda-normalization method, Geophys. J. Int., 193(3), 1726–1731.
- Del Pezzo E, Ibañez J, Morales J, Akinci A, Maresca R (1995) Measurements of intrinsic and scattering seismic attenuation in the crust. Bull Seism Soc Am 85:1373–1380.
- Dutta, U., N. N. Biswas, D. A. Adams, and A. Papageorgiou (2004) Analysis of S-wave attenuation in south-central Alaska, Bull. Seismol. Soc. Am. 94, 16–28.
- Frankel, A., and L. Wennerberg (1987) Energy flux model of the seismic coda: Separation of scattering and intrinsic attenuation, Bull. Seismol. Soc. Am. 77, 1223–1251.
- Giampiccolo E, Tuvè T, Gresta S, Patan D (2006) S wave attenuation and separation of scattering and intrinsic absorption of seismic energy in southeastern Sicily (Italy). Geophys J Int 165:211–222
- Goric M, Muller G (1987) Apparent and intrinsic Q. The one-dimensional case J Geophys Res 61:46–54.
- Gupta, S. C., Singh, V. N. and Kumar, A. (1995). Attenuation of coda waves in the Garhwal Himalaya, India, Phys. Earth Planet. Inter. 87, 247–253.
- Hauksson E, Shearer PM (2006) Attenuation models (Qp and Qs) in three dimensions of the southern California crust: inferred fluid saturation at seismogenic depths. J Geophys Res 111:B05302. doi:10.1029/2005JB003947.
- Havskov, J., Malone, S., McClurg, D., Crosson, R. (1989) Coda Q for the state of Washington. Bulletin of the Seismological Society of America, 79, 1024-1038.
- Hazarika, P., M. Ravi Kumar, Dinesh Kumar (2013) Attenuation character of seismic waves in Sikkim Himalaya. Geophysical Journal International, Volume 195, Issue 1, 1 October 2013, Pages 544–557, https://doi.org/10.1093/gji/ggt241.
- Horasan, G. & Boztepe-G¨uney, A. (2004) S-wave attenuation in the Sea of Marmara, Turkey, Phys. Earth planet. Inter., 142, 215–224.
- Hough, S. E., and J. G. Anderson (1988). High frequency spectra observed at Anza, California: implications for Q structure, Bull. Seismol. Soc. Am. 78, 672–691.
- Ibanez, J.M., Delpezzo, E., Demiguel, F., Herraiz, M., Alguaicil, G., Morales, J. (1990) Depth dependent seismic attenuation in the Granada Zone (Southern Spain). Bull Seismol Soc Am 80:1232–1244.
- Imtiyaz A. Parvez, Preeti Yadav, K. Nagaraj (2012) Attenuation of P, S and Coda Waves in the NW-Himalayas, India. International Journal of Geosciences, Vol. 3 No. 1, 2012, pp. 179-191. doi: 10.4236/ijg.2012.31020.
- Jin, A. and Aki, K. (1988) Spatial and Temporal Correlation between Coda Q and Seismicity in China. Bulletin of the Seismological Society of America, 78, 741-769.
- Jin, A. and Aki, K. (1989) Temporal correlation between coda Q−1 and seismicity evidence for a structural unit in the brittle-ductile transition zone, J. Geodyn., 17, 95-120.
- Kayal, J. R., De, R., and Chakraborty, P. (1993). Microearthquakes at the Main Boundary Thrust in eastern Himalaya and the present day tectonic model. Tectonophysics 218, 375–381.
- Krishnaswamy, V.S., Jalote, S.P., Shome, S.K. (1970) Recent crustal movements in northwest Himalaya and the Gangetic foredeep and related pattern of seismicity. In: Proceedings of the fourth symposium on earthquake engineering, University of Roorkee, p. 419–39.
- Kvamme, L.B., Havskov, J. (1989) Q in southern Norway. Bull. Seismol. Soc. Am. 79:1575–1588.
- Kumar, Naresh & Parvez, Imtiyaz & Virk, Hardev (2005) Estimation of coda wave attenuation for NW Himalayan Region using local earthquakes. Physics of the Earth and Planetary Interiors. 151. 243-258. 10.1016/j.pepi.2005.03.010.
- Kumar, N., Mate, S. & Mukhopadhaya, S. (2014) Estimation of Qp and Qs of Kinnaur Himalaya. J Seismol (2014) 18: 47. https://doi.org/10.1007/s10950-013-9399-7.
- Kumar, R, Gupta, S.C., Singh, S.P., and Kumar, A. (2016a) The Attenuation of High-Frequency Seismic Waves in the Lower Siang Region of Arunachal Himalaya: Qα, Qβ, Qc, Qi, and Qs, Bulletin of the Seismological Society of America 106(4).
- Latchman, J.L., Ambeh, W.B., Lynch, L.L., 1996. Attenuation of seismic waves in the Trinidad and Tobago area. Tectonophysics 253, 111–127.
- Ma’hood, M., H. Hamzehloo, and G. J. Doloei (2009) Attenuation of high frequency P and S waves in the crust of the east-central Iran, Geophys. J.Int. 179, no. 3, 1669–1678, doi: 10.1111/j.1365-246X.2009.04363. x.
- Mandal, P., Rastogi, B.K., 1998. A frequency-dependent relation of coda Qc for Koyna-Warna region. India. Pure Appl. Geophys.153, 163–177.
- Mandal, P., Jainendra, Joshi, S.,Kumar, S., Bhunia, R., Rastogi, B.K., 2004. Low coda Qc in the epicentral region of the 2001 Bhuj earthquake of Mw 7.7. Pure Appl. Geophys. 161, 1635–1654.
- Matsunami, K. (1991) Laboratory tests of excitation and attenuation of coda waves using 2-D models of scattering media, Phys. Earth Planet. Inter., 67, 36–47.
- Mayeda, K., Koynagi, S., Hoshiba, M., Aki, K. & Zeng, Y., 1992. A comparative study of scattering, intrinsic and coda Q for Hawaii, Long Valley, and Central California between 1. 5 and 15 Hz, J. geophys. Res., 97, 6643–6659.
- Mishra, O. P., Vandana, Kumar, V., & Gera, S. K. (2020). A new insight into seismic attenuation characteristics of Northwest Himalaya and its surrounding region: Implications to structural heterogeneities and earthquake hazards. Physics of the Earth and Planetary Interiors, 106500. doi:10.1016/j.pepi.2020.106500.
- Mittal, H., Kumar, A., Ramhmachhuani, R. (2012) Indian National Strong Motion Instrumentation Network and site characterization of its stations, Int. J. Geosci. 3, 1151–1167, doi: 10.4236/ijg.2012.326117.
- Mukhopadhaya, S., Sharma, J., Massey, R., Kayal, J.R. (2008) Lapse time dependence of coda Q in the source region of the 1999 Chamoli earthquake. Bull Seismol Soc Am 98:2080–2086.
- Nandy, D. R. (2001) Geodynamics of Northeastern India and the Adjoining Region. ACB publication, New Delhi, 209 pp.
- Ni, J. and Barazangi, M. (1984) Seismotectonics of the Himalayan collision zone. Journal of Geophysical Research, 89: 1147-1163.
- Padhy, S. and Subhadra, N. (2010) Frequency-dependent attenuation of P and S waves in northeast India. Geophysical Journal International, 183: 1052–1060. doi: 10.1111/j.1365-246X.2010.04783. x.
- Padhy, S., Subhadra, N., Kayal, J. R. (2011) Frequency-Dependent Attenuation of Body and Coda Waves in the Andaman Sea Basin. Bulletin of the Seismological Society of America; 101 (1): 109–125. doi: https://doi.org/10.1785/0120100032.
- Pujades, L., Canas, J.A., Egozcue, J.J., Puigvi, M.A., Pous, J., Gallart, J., Lana, X., Casas, A. (1991) Coda Q distribution in the Iberian Peninsula. Geophysical Journal International, 100, 285-301.
- Pulli, J.J. (1984) Attenuation of coda waves in New England. Bull. Seism. Soc. Am. 74, 1149–1166.
- Rajal, B.S., Virdi, N.S., Hasija, N.L. (1986) Recent crustal uplift in the Doon Valley. International Symposium on Neotectonics in South Asia, Dehradun, India,146–59.
- Rautian, T. G., and Khalturin, V. I. (1978). The use of the coda for the determination of the earthquake source spectrum, Bull. Seismol. Soc. Am. 68, 923–948.
- Roecker, S.W., Tucker, B., King, J., Hatzfeld, D. (1982) Estimation of Q in Central Asia as a function of frequency and depth using the coda of locally recorded earthquakes. Bull Seismol Soc Am 72:129–149.
- Sarma, K.P., Bharali, T.K. and Chowdhary, P.K. (1998) Rotation of micro-structural fabric in the Barapani shear zone, Meghalaya; Proc. Reg. Sem. Dev. Geol. Res. in NE India, Gauhati University, 57–67.
- Sato, H. (1977). Energy propagation including scattering effect, J. Phys. Earth. 25, 27–41.
- Sato H, Fehler M (1998) Seismic wave propagation and scattering in the heterogeneous earth. AIP Press/Springer, New York, p 308.
- Seeber, L., Armbruster, J., Quittmeyer, R. (1981) Seismicity and continental subduction in the Himalayan Arc. In: H.K. Gupta and F.M. Delany (Eds.), Zargos-Hindukush- Himalaya: Geodynamic Evolution, Geodynamic Series, American Geophysical Union, 3: 215 - 242.
- Sharma B, Teotia SS, Kumar D (2007) Attenuation of P, S, and coda waves in Koyna region, India. J Seismol 11:327–334.
- Sharma, B., Gupta, A. K., Devi, D. K., Kumar, D., Teotia, S. S. and Rastogi, B. K. (2008) Attenuation of high-frequency seismic waves in Kachchh region, Gujarat, India, Bull. Seismol. Soc. Am. 98, no. 5, 2325–2340.
- Sharma, B., Teotia, S. S., Kumar, D. and Raju, P. S. (2009) Attenuation of P- and S-waves in the Chamoli region, Himalaya, India, Pure Appl. Geophys. 166, no. 12, 1949–1966.
- Singh, A., Bhushan, K., Singh, C., Steckler, M. S., Akhter, S. H., Seeber, L., Kim, W.Y., Tiwari, A. K., and Biswas, R. (2016) Crustal structure and tectonics of Bangladesh: New constraints from inversion of receiver functions, Tectonophysics.
- Singh, C., V. K. Srinivasa Bharathi, and Chadha, R. K. (2012) Lapse time and frequency-dependent attenuation characteristics of Kumaun Himalaya, J. Asian Earth Sci. 54, 64–71.
- Singh, S. & Herrmann, R.B. (1983) Regionalization of crustal coda Q in the continental United States, J. Geophys. Res., 88, 527–538.
- Sorkhabi, Rasoul B., and Macfarlane, A. (1999) Himalaya and Tibet: Mountain roots to mountain tops, in Macfarlane, A., Sorkhabi, R.B, and Quade, J., eds., Himalaya and Tibet: Mountain Roots to Mountain Tops, Geol. Survey of America Special Paper 328, 1-8.
- Tripathi, J.N., Ugalde, A., 2004. Regional estimation of Q from seismic coda observation by the Gauribidanur seismic array (southern India). Phys. Earth Planet Inter. 145, 115–126.
- Tuvè, T., Bianco, F., Ibáñez, J., Patanè, D., Del Pezzo E., Bottari A. (2006) Attenuation study in the Straits of Messina area (southern Italy), Tectonophysics, 421, 173–185.
- Wallace, K., Bilham, R., Blume, F., Gaur, V.K., Gahalaut, V.K. (2005) Surface deformation in the region of the 1905 Kangra Mw = 7.8 earthquake in the period 1846–2001. Geophys. Res. Lett.,32, L15307, doi:10.1029/2005GL022906.
- Wennerberg, L. (1993) Multiple-scattering interpretations of coda-Q measurements. Bull. Seismol. Soc. Am. 83, 279–290.
- Wessel P, Smith WHF. New, improved version of Generic Mapping Tools released, Eos Trans. AGU 1998;79:579.
- Wu, R. S. (1985) Multiple scattering and energy transfer of seismic waves Separation of scattering effect from intrinsic attenuation—I. Theoretical modelling, Geophys. J. Int. 82, no. 1, 57–80.
- Wu, R. S., and K. Aki (1988). Scattering and attenuations of seismic waves, part I, in Pure and Applied Geophysics, Vol. 128, Springer Verlag, Heidelberg, Germany, 1–6.
- Yin, A., Harrison, T.M. (2000) Geologic evolution of the Himalayan– Tibetan orogen. Annu Rev Earth Planet Sci 28:211–280.
- Yoshimoto, K., H. Sato, and M. Ohtake (1993) Frequency-dependent attenuation of P and S waves in the Kanto area, Japan, based on the coda normalization method, Geophys. J. Int. 114, no. 1, 165–174.
- Zelt, B.C., Dotzev, N.T., Ellis, R.M., Roger, G.C., 1999. Coda Q insouthwestern British Columbia. Can. Bull. Seismol. Soc. Am.89, 1083–1093.
- Zeng, Y. (1991) Compact solutions for multiple scattered wave energy in time domain, Bull. Seismol. Soc. Am. 81, 1022–1029.