Investigation of crustal thickness and uppermost mantle velocity beneath Gujarat, western India, utilizing Moho reflected P phases

https://doi.org/10.1016/j.pepi.2020.106619Get rights and content

Highlights

  • PmP phases from local earthquakes has been used to estimate crustal thickness and uppermost mantle velocity across Gujarat.

  • The methodology has been shown to work for a comparatively smaller dataset for the first time.

  • The results indicate highest crustal thickness and uppermost mantle velocity in the Kachchh rift zone.

Abstract

Crustal thickness and its variation can provide important insights into the tectonics of a region. The most popular methodology to determine crustal thickness – receiver function analysis utilizes converted phases from the Moho discontinuity. Although the reflection co-efficient is comparable to the transmission co-efficient, reflected phases are rarely used for determining crustal thickness. We tested and utilized such a rare methodology to determine the crustal thickness and uppermost mantle P-velocity beneath Gujarat, a seismically active region of the northwestern Deccan Volcanic Province. We used waveforms of ~2900 local and regional earthquakes (ML≥2.0) recorded by a network of broadband seismometers spanning the region. After manually marking the P arrivals from all the seismograms, computing and applying several signal-to-noise ratio criteria and aligning each seismogram to their P-onsets, a total of 9134 seismograms were stacked together. This stack enabled identification of the Moho reflected phases (PmP), based on the amplitude. Forward modelling of the observed PmP data yielded estimates of crustal thickness (H) and uppermost mantle velocities (v), given a crustal velocity structure or equivalent average velocity of the region. We divided the region into 0.5° × 0.5° grids and estimated the values of H and v at each grid point, using stacks of seismograms corresponding to Moho reflection points (within 50 km) of the grid. These values provided the lateral variation of the crustal thickness and uppermost mantle velocities across Gujarat. We found that the average uppermost mantle velocity is higher in the Kachchh region compared to that in Saurashtra, probably indicating past high-temperature events that transported and accumulated high velocity materials from greater depths beneath Kachchh. The crustal thickness also follows the same trend and maximum thickness of 43 km was observed within the center of the Kachchh rift with values gradually decreasing outwards from it. It is very likely that a thick crust beneath Kachchh includes a root. The average thickness and uppermost mantle velocity in Saurashtra were found to be 38.5 km and 7.9 km/s, respectively.

Introduction

The western margin of India consists of a number of failed-rift systems that evolved sequentially from north to south (Biswas, 1982, Biswas, 1987). Although the association of these paleo-rifts with the present tectonics of this region is well documented, their exact role remains obscure. Two of these failed-rift systems, namely Kachchh and Cambay are situated in the state of Gujarat (Fig. 1a). The devastating 2001 MW 7.7 Bhuj earthquake occurred within the Kachchh basin (shown by a star in Fig. 1a), which diverted the attention of the global earth science community to this region. Since then, several geological and geophysical studies have been carried out (e.g. Antolik and Dreger, 2003; Bodin and Horton, 2004; Copley et al., 2011; Kayal et al., 2002; Schmidt and Bürgmann, 2006) to unravel the causative tectonic processes. The Saurashtra peninsula lying south of both the Kachchh and Cambay basins, is also a part of Gujarat.

The Kachchh region witnessed initiation of continental rifting in the late Triassic (Biswas, 1982). Rifting processes greatly alter the characteristics of the crust and lithosphere (e.g. Corti, 2012; Thompson et al., 2001). Moreover, western India interacted with a plume in the late Cretaceous (Storey, 1995). Therefore, it is likely that the crust of Kachchh is highly deformed and heterogeneous in nature, having preserved signatures of all these deformation processes. Also, it is crucial to understand whether the crust beneath adjacent regions like Saurashtra and Cambay are as deformed as that of Kachchh. For the above reasons, determination of crustal thickness and its lateral variation are important. In addition, it enables deciphering the seismogenic depth. Moreover, in a rift basin, crustal thickness can provide direct evidence of the vertical extent of rifting.

The most prevalent methodology for determining crustal thickness, in earthquake seismology, is the receiver function (RF) technique (Ammon, 1991; Vinnik, 1977). In this methodology, the P-to-S converted energy (while transmitting through a discontinuity, especially Moho) is used to infer the characteristics of the discontinuity (Moho). Receiver functions have been employed to determine the crustal thickness of Kachchh and adjoining regions (Chopra et al., 2014; Mandal, 2006, Mandal, 2011). However, receiver function results for the Kachchh region are not consistent, as both a thick (Chopra et al., 2014) and thin crust (Mandal, 2011) have been reported. A separate and independent investigation, in this respect, will probably assist in resolving the issue.

It has been analytically shown that P-to-S converted energy for a near-vertically incident ray (which is the case for RFs) is smaller than even 10% (Aki and Richards, 2002; Stein and Wysession, 2003). The energy reflected from a P wave incident at shallower angles is also comparable to this value. Although RFs have been extensively used, only a few studies have made use of the reflected waves to infer the crustal thickness (e.g. Richards-Dinger and Shearer, 1997; Luo et al., 2019). Reflected wave study using local earthquakes is a low-cost equivalent of controlled source reflection surveys, therefore capable of providing detailed images. In receiver function studies, azimuthal coverage of teleseismic earthquakes may become inadequate in some cases, as large earthquakes are mainly concentrated in the plate boundaries. In this respect, a large set of local earthquakes can provide better coverage for reflection studies. Richards-Dinger and Shearer (1997) used PmP (Moho-reflected P) arrival times to study the Moho depth of southern California utilizing a huge dataset of over 350,000 waveforms obtained from a subset of over 200,000 earthquakes. For Gujarat, although the estimates of crustal thickness are available from RF studies, the moho-reflected P waves originating from local earthquakes have never been used. The objective of this study is two-fold – firstly, to test if the moho-reflected phases can be utilized to obtain reliable crustal thicknesss and uppermost mantle velocity with a much smaller dataset (less than one-tenth of the data used in southern California), and secondly, compare the results obtained from this method with those from receiver function studies. Also, this study is aimed at obtaining the variation of crustal thickness and uppermost mantle velocity across Gujarat region and understand their implications.

Section snippets

Data and its stacking

After the devastating 2001 Bhuj earthquake, the Institute of Seismological Research (ISR) was established to prioritize seismological research, assess and mitigate seismic hazard and understand the tectonics of the Gujarat region. In 2006, ISR established a network of 22 broadband seismological stations distributed throughout Gujarat, which was further expanded to 45 broadband stations in 2008 (Chopra et al., 2008). Furthermore, several temporary stations have been operated during various

Stacking results

The result of stacking the entire dataset is shown in Fig. 2a. It can be seen that the normalized amplitude is very small (almost zero) before the start time (zero mark) increases instantaneously from the mark, indicating the P-wave onset. However, we do not observe a unit amplitude for the P-arrival, as expected, but comparatively smaller values (0.35–0.5). This is due to the fact that we aligned the seismograms by the P-onset time while the peak amplitude lies within a 0.5 s window following

Estimation of crustal thickness and uppermost mantle velocity

Estimation of H and v was carried out in two steps – (i) identifying and picking the onset of PmP phase in various epicentral distance ranges, and (ii) obtaining the best fitting theoretical tPmP - tP curve to the observed data points. We observed a distinct pattern of variation of normalized amplitudes with time, wherever PmP arrivals are identified visually. As mentioned earlier, the normalized amplitude is small before the zero mark, but reaches upto 0.5 during the peak P arrival and then

Factors influencing the results

The major factors affecting the estimated results are crustal velocity, dip of the Moho and earthquake locations. The traveltimes of Pg and PmP phases are dependent on crustal velocities while that of the Pn phase depends on both crustal and uppermost mantle velocities. In order to test the influence of crustal velocities, we examined several 1D velocity models to estimate H and v. The velocity model of Rao and Tewari (2005) is specific to Saurashtra region while that of Kayal et al. (2002);

Conclusions

Our study has shown that Moho reflected phases from local and regional earthquakes are very effective in determining the crustal thickness and uppermost mantle velocity using a comparatively smaller dataset of ~9000 waveforms. This methodology offers an alternative for the prevalent technique of receiver function analysis to determine crustal thickness. The quality of the crustal thickness results obtained from this method is at par with receiver function results, or even better considering the

Credit author statement

  • Himangshu Paul: Conceptualization, Methodology, Software, Validation, Formal Analysis, Data Curation, Writing- Original Draft, Writing - Review & Editing, Visualization.

  • Abhay Pandey: Conceptualization, Software, Formal Analysis.

  • M. Ravi Kumar: Conceptualization, Methodology, Resources, Writing – Review & Editing, Supervision, Project administration, Funding acquisition.

  • Santosh Kumar: Investigation, Resources, Data Curation, Project administration, Funding acquisition

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

HP and AP are thankful to the Institute of Seismological Research (ISR), Gandhinagar, for providing the opportunity and dataset to carry out this study. The data used in this study is proprietary of ISR and is available on request to the Director-General (DG), ISR. We are thankful to all the technical and administrative staff of ISR who work very hard to install, maintain and retrieve data from seismological stations, located in various parts of the state of Gujarat as well as other states. We

References (43)

  • K. Aki et al.

    Quantitative Seismology

    (2002)
  • C.J. Ammon

    The isolation of receiver effects from teleseismic P wave forms

    Bull. Seismol. Soc. Am.

    (1991)
  • M. Antolik et al.

    Rupture process of the 26 January 2001MW 7.6 Bhuj, India, earthquake from teleseismic broadband data

    Bull. Seismol. Soc. Am.

    (2003)
  • S.K. Biswas

    Rift basins in the western margin of India and their hydrocarbon prospects with special reference to Kutch Basin

    Bull. Am. Assoc. Pet. Geol.

    (1982)
  • S.K. Biswas

    A review of structure and tectonics of Kutch basin, western India, with special reference to earthquakes

    Curr. Sci. India

    (2005)
  • P. Bodin et al.

    Source parameters and tectonic implications of aftershocks of the MW 7.6 Bhuj earthquake of 26 January 2001

    Bull. Seismol. Soc. Am.

    (2004)
  • W.R. Buck

    Modes of continental lithospheric extension

    J. Geophys. Res.

    (1991)
  • S. Chopra et al.

    The Gujarat (India) seismic network

    Seismol. Res. Lett.

    (2008)
  • Chopra, S., Chang, S. S., T-M, Yadav, R. B. S., Choudhury, P., & Roy, K. S. (2014). Crustal structure of the Gujarat...
  • N.I. Christensen et al.

    Seismic velocity structure and composition of the continental crust: a global view

    J. Geophys. Res.

    (1995)
  • A. Copley et al.

    The 2001 MW 7.6 Bhuj earthquake, low fault friction, and the crustal support of plate driving forces in India

    J. Geophys. Res.

    (2011)
  • Cited by (0)

    View full text