Revisiting the seismogenic characteristics of stable continental interiors: The case of three Indian events
Introduction
Global distribution of shield area earthquakes shows that the stable continental regions (SCR) are exposed to destructive events (Johnston, 1996a,b; Crone et al., 1997; Adams and Basham, 1989; Page and Hough, 2014). Despite negligible tectonic loading rates at these stable regions, significant elastic strain is released through occasional earthquakes (Sykes and Sbar, 1973; Campbell et al., 2015). Best known examples of global SCR events include the 1811–1812 New Madrid, Central US; 1886 Charleston, South Carolina; 1955 Amazon craton, Brazil; 1988 Tennant Creek, Australia; 1819 Kachchh, 1993 Latur and 2001 Bhuj; India earthquakes (Johnston, 1996c; Bowman, 1992; Barros et al., 2009; Bakun and Hopper, 2004; Bilham, 1999; Rajendran et al., 2001). The New Madrid seismic zone (NMSZ) sequence (four earthquakes of M > 7, between December 1811 to February 1812) is considered to be linked with faulting in a failed intracontinental rift (Tuttle et al., 2002; Hough et al., 2000; Johnston and Schweig, 1996; Mueller et al., 2004). The NMSZ is seismically still active and has hosted several earthquakes in the past (e.g., Page and Hough, 2014). Here, the paleo-seismological and instrumental observations suggest a repeat period of several hundred to thousands of years (e.g., Johnston and Nava, 1985; Tuttle et al., 1999; Schweig and Ellis, 1994; Newman et al., 1999). Temporal clustering of events is also observed here, which disrupts very long intervals of seismic quiescence (e.g., Holbrook et al., 2006). Similarly, the South Carolina region, where 1886 Charleston event occurred, also has large repeat period of several hundreds of years, for events of M > 7 (Talwani and Cox, 1985; Talwani and Schaeffer, 2001).
Interestingly, there are SCR regions without any large event occurrences but exhibit moderate seismicity. A typical example is the Amazon basin, a paleozoic rift zone in central Brazil, where a considerable amount of Mw 5–6 earthquakes are observed (e.g., Barros et al., 2009). At the same time, there are SCR regions with no previously observed events in the paleo-earthquake record, like the Meers fault region of southwestern Oklahoma, US. This fault has generated two events, ~1200–1300 years ago, with surface ruptures. But, prior to that, this region never experienced any earthquakes for more than ~120000 years (Madole, 1988; Crone and Luza, 1990). Another example of a SCR event which struck a region with no previously known seismic activity is the 1969 Ceres event of South Africa (Mw 6.3) (Smit et al., 2015). Similarly, the Australian shield also has experienced seismic events in regions with no prior evidence of faulting, like the 1986 Marryat Creek and the 1988 Tennant Creek earthquakes, both of which generated surface ruptures (e.g., Choy and Bowman, 1990; Crone et al., 1997, 2003). Similar to NMSZ, some regions of the Australian shield also show temporal clustering of seismic activity. Here, these clusters are separated by several thousands of years of seismic quiescence (e.g., Crone et al., 1997; Clark et al., 2012). All these point towards the differing nature of SCR seismic sources, where some of the events came as a complete surprise, particularly when the penultimate earthquake might have occurred well beyond the documented history.
Studies stipulate various seismogenic mechanisms for SCR earthquakes, including stress perturbations induced by regional plate boundary forces (e.g., Stein et al., 2002), reservoir loading/unloading (e.g., Gupta, 1992), deglaciation and flexure (e.g., Stein et al., 1979), regional stress concentrations in or around mafic intrusions (Campbell, 1978), stress concentration near the intersections of faults (Talwani, 1988, 1999) etc. The inhomogenities in crustal structure, presence of fluids and effects of topography are also considered to be major influencing factors in the SCR earthquake genesis (Sykes, 1978; Mareschal and Kuang, 1986). It is widely believed that most of the SCR events are associated with regional rifted structures (e.g., Johnston and Kanter, 1990; Talwani and Rajendran, 1991). But, recent studies do not always hold well with these results. For example, studies show that the SCR events are not just limited to rift systems (e.g., Schulte and Mooney, 2005), but can also occur within non-rifted crust or faults which are not considered to be weak (e.g., Seeber et al., 1996). In addition, the SCR events do not always follow stress accumulation and release patterns of plate boundary ruptures (e.g., Calais et al., 2016) and do not always have typical repeat times which are preceded or succeeded by a period of seismic quiescence (Clark et al., 2012).
Though large SCR events are comparatively rare and release only a small percentage of the global seismic energy, they can cause massive damages to life and property. Despite its rarity, there is a significant seismic hazard arising from SCR earthquakes. A major danger being the lack of preparedness within the local communities due to no knowledge of previous event occurrence or the repeat times for an event genesis. Moreover, the SCR events can be of large magnitude, inherently due to a cold and deeper brittle lithosphere (Sloan et al., 2011), where the continental interior has a very low strain, compared with very actively deforming plate boundary zones. As already observed, due to its rarity and limited occurrences within the instrumental observation era, we are restricted with a very limited amount of data regarding SCR event occurrences. Revisiting some of the events, using present-day state-of-the-art modelling techniques may provide a better understanding on the nature of seismic asperity breakage, slip extent, maturity of the fault, etc. This can possibly help us in adding to our knowledge on the SCR faulting processes.
This paper mainly focuses on the Indian SCR seismogenesis, where occasional large earthquakes are reported and have generated a few of the most disastrous earthquakes in the global SCR earthquake history (See Fig. 1). With this perspective, we look at the spatio-temporal slip distribution of the 1993 Mw 6.2 Latur, 1997 Mw 5.8 Jabalpur and the 2001 Mw 7.6 Bhuj earthquakes, through tele-seismic finite-fault models. Further, using these slip distribution models, we estimate the coseismic stress drop values and finally try to link the source process characteristics to the regional tectonics. Slip distribution solutions for the Bhuj earthquake using seismic and geodetic methods are already discussed widely in literature (Antolik and Dreger, 2003; Chandrasekhar et al., 2004; Schmidt and Bürgmann, 2006; Copley et al., 2011), but the finite-fault models of the 1993 Latur and 1997 Jabalpur events are being presented for the first time.
Section snippets
Indian SCR seismicity
The major tectonic driver for plate deformation is the state of lithospheric stress, where the accumulated strain is released through earthquakes. Geodetic studies indicate microstrain rates within the Indian lithosphere (Paul et al., 1995) and despite this, significant elastic strain is released through occasional earthquakes (Fig. 1). Among these, the most significant historical event is the 1819 Mw 7.8 Kachchh earthquake (Bilham, 1999; Rajendran et al., 1998). The other significant Indian
Materials and methods
The source parameters and kinematic slip distribution of the Indian SCR earthquakes are modelled using Kikuchi and Kanamori (1982, 1986, 1991), where the seismic source is characterised by a series of spatially distributed point sources. Good quality tele-seismic body waveforms are retrieved from the Incorporated Research Institutions for Seismology - Data Management Center (IRIS-DMC), considering good azimuthal coverage of the stations. The data are constrained within an epicentral limit of
1993 Latur earthquake
Inversion of azimuthally well distributed 43 tele-seismic waveforms (35 P and 8 SH) of the 1993 Latur earthquake (Figure S1), yields a SE-NW oriented, southwest-dipping reverse faulting mechanism (strike: 134°, dip: 44° and rake: 112°) (Fig. 2.c). Our best-fit model has a shallow earthquake nucleation depth of 7 km (Fig. 2.a and Table 1). Model iterations with varying rupture velocities (1–4 km/s) show a sudden decrease in the NRMS misfit after a Vr of 2.4 km/s, with the least misfit at
Discussion
The finite-fault inversion results from this study on the Indian SCR earthquakes show variations in nucleation depth, seismic slip extends, rupture duration, rupture velocity, stress drop, etc. Our model solutions indicate that the Latur event ruptured a very shallow upper crustal portion, with the rupture breaching the surface. Likewise, both the Jabalpur and Bhuj events have their slip majorly distributed within the lower crustal regions, where the 2001 Bhuj event slip does not breach the
Conclusions
The following are the major conclusions from this study on revisiting the seismogenic characteristics of Indian stable continental interior through kinematic source process modelling of the 1993 Latur, 1997 Jabalpur and 2001 Bhuj earthquakes.
- •
Indian SCR earthquakes generally show a compact zone of asperity breakage restricted within the crust. The Bhuj and Jabalpur events have their rupture mostly restricted within the lower crustal regions, whereas Latur asperity rupture is very shallow with a
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.
Acknowledgements
Tele-seismic data are from the Incorporated Research Institutions for Seismology (IRIS). Seismic Analysis Code (SAC) was used for the data processing (Goldstein and Snoke, 2005). Earthquake catalogs from NEIC, precise locations from EHB bulletin of International Seismological Centre (ISC-EHB) and focal mechanism data from Global Centroid Moment tensor catalog (GCMT) (Ekström et al., 2012) are used in this study. We used the Generic Mapping Tools for drawing the figures (Wessel and Smith, 1998).
References (133)
- et al.
The intraplate porto dos gauchos seismic zone in the amazon craton|Brazil
Tectonophysics
(2009) Source parameters of two rift-associated intraplate earthquakes in peninsular India: the Bhadrachalam earthquake of April 13, 1969 and the Broach earthquake of March 23, 1970
Tectonophysics
(1993)- et al.
Long-term behaviour of Australian stable continental region (SCR) faults
Tectonophysics
(2012) - et al.
The global cmt project 2004-2010: centroid-moment tensors for 13,017 earthquakes
Phys. Earth Planet. In.
(2012) - et al.
Rupture mechanism of the 1993 killari earthquake, India: constraints from aftershocks and static stress change
Tectonophysics
(2003) - et al.
An investigation into the latur earthquake of september 29, 1993 in southern India
Tectonophysics
(1998) - et al.
Stratigraphic evidence for millennial-scale temporal clustering of earthquakes on a continental-interior fault: holocene Mississippi river floodplain deposits, new madrid seismic zone, USA
Tectonophysics
(2006) - et al.
Inversion of complex body waves-ii
Phys. Earth Planet. In.
(1986) - et al.
Paleoearthquake signatures from the south wagad fault (swf), wagad island, kachchh, Gujarat, western India: a potential seismic hazard
J. Struct. Geol.
(2017) Three-dimensional modeling of intraplate stresses in the epicentral zone of the 21 may 1997 jabalpur earthquake of mw 5.8, central India
Tectonophysics
(2010)