April 2015, M_w7.8, Gorkha, Nepal earthquake is one of the disastrous earthquakes which caused widespread damage in Nepal and nearby countries. The earthquake associates itself with a 135 km long rupture length which progressed southeastwards towards its most significant aftershocks (M_w7.3) that occurred after 16 days. It recorded the highest seismic intensity of IX on the MMI scale in the Epicentral region. Prakash et al. (2016) estimated a stress drop of 3.4 MPa for the Nepal earthquake 2015 in the inter-plate region; Han et al. (2017) reported dynamic triggering in southwest China ∼2,200 km away during the 2015 Nepal event. Thus, delineating the link between the Nepal event and its triggering capability in Gujarat, NW India, might lead to new insights into dynamic triggering. The Nepal earthquake occurred ∼1,200 km away from the Gujarat region (Figure 1A). It is, therefore, important to look for signs of triggering seismicity in the region.

The M_w7.3 Landers earthquake of 1992 (Hill et al., 1993) gave birth to the study of remote triggering. It reinforced the idea of remotely triggered earthquakes; now, the study has matured, but the mechanism of such triggering remains elusive. The magnitude of triggered seismicity is generally less than 2 (Harrington and Brodsky, 2006; Wang et al., 2019), which generally global catalogs miss. Velasco et al. (2008) analyzed 15 earthquakes recorded by 500 globally distributed stations and suggested that remote triggering is ubiquitous. Thus, it is an essential criterion for selecting regions with good-quality earthquake catalogs supplemented with a wide range of digital waveforms to comprehensively study and examine dynamic triggering.

The digital waveform network covers Gujarat, Northwestern India well (Figure 1B). The Kachchh Rift Basin (KRB) is a seismically active intraplate region that also hosted the M_w7.7 Bhuj earthquake in 2001 (Figure 1B). The Bhuj earthquake of 2001 renewed seismologists’ interest in understanding the region’s geodynamics. The Institute of Seismological Research (ISR) installed the Gujarat Seismic Network (GSNet) in 2006 (Figure 1B). A digital network of 54 accelerographs and 60 high-resolution three-component seismographs has been installed, stretching the entire state of Gujarat. The newer data furnishes an excellent opportunity to interpret the earthquake’s origin, and it is undergoing geodynamic processes.

The Northwestern Deccan Volcanic Province, India, has been a host of moderate to major intra-plate earthquakes (Rajendran and Rajendran, 2001; Gupta et al., 2001; Singh et al., 2015). The Bureau of Indian Standards 2002 categorized the KRB in Seismic Zone-V (highly vulnerable). On the other hand, the different regions like Saurashtra Horst (SH) and Mainland (ML) regions of Gujarat are less seismically active than KRB and fall in the seismic zone IV/III. The three physiographical units (SH, ML, and KRB) have different tectonic setups, geology, and seismicity (Biswas, 1987; ISR annual report 2016). The area lies approximately 1,000 km off the region of the Himalayan Collision Zone and about 500 km off the plate boundary of India and Arabian.

In peninsular India, the Kachchh region of Gujarat state, although not located on or near any plate boundary, has been experiencing frequent earthquakes. The KRB evolved during 135 Ma and formed due to extensional tectonics (Kayal et al., 2002a; Kothari et al., 2016). The region is the most active intra-continental region globally, and in addition to the 2001 M_w7.7 Bhuj event, the KRB hosted some of the most significant known intraplate earthquakes, viz, the 1819 Allahbund earthquake (Dam-of-God) (M 7.8), the 1,668 Indus delta (MM X), the 1,845 Lakhpat earthquake (M 6.3, MM VIII), and 1956 Anjar earthquake (M_w6.0) (Gaur, 2001; Rajendran and Rajendran, 2001; Rastogi et al., 2011).

Previous studies have indicated that the formation of the KRB is attributed to rifting along with the E-W tectonic trend. The E-W trending Kachchh region has also been ascertained from the highly abnormal values in the Bouguer gravity data (GSI, 2000; Chandrasekhar and Mishra, 2002). Moreover, the E-W alignment of faults within the basin controls the KRB structurally, the faults being Island Belt Fault (IBF), Banni Fault (BF), Kachchh Mainland Fault (KMF), Katrol Hill Fault (KHF), South Wagad Fault (SWF), North Wagad Fault (NWF) and Gedi Fault (GF) as shown in Figure 1 and described in Supplementary Table S1.

In contrast to the KRB, the SH region of Gujarat is seismically less active. A horst structure bounded by faults on all four edges controls the seismicity in the SH. The SH is bounded by the extended Son Narmada Fault (SNF) and the North Kathiawar Fault (NKF) on the southern edge and its northernmost border, respectively. The West Cambay Fault (WCF) and the WNW–ESE trending West Coast Fault (WCF) are on the eastern edge and the Arabian Sea, respectively. The SH has hosted three medium-sized shallow earthquakes in the Talala region since 2007, M_w4.8 and M_w5.0 in 2007 and M_w5.1 in 2011 (Yadav et al., 2011; Singh et al., 2013; Singh and Mishra, 2015). Historically, the area has also experienced many other earthquakes, such as the M_w6.1 event of 1919 in Bhavnagar, considered the biggest in the region (Rajendran and Rajendran, 2001). Historical earthquakes occur away from plate boundaries and are considered intraplate events. Hence, the stresses therein due to plate boundaries are still unclear.

The ML region of Gujarat is limited by two long boundary fault systems, the Cambay rift basin in the north and the Narmada rift basin in the south of Gujarat. Further, the Narmada rift is divided into Narmada North Fault (NNF) and the Narmada South Fault (NSF). The most significant event generated by the Narmada lineament is 1970 M_w5.4, with a depth of 10 km and a strike-slip mechanism (Gupta et al., 1972; Chandra, 1977; Rastogi et al., 2012). On the other hand, the NNW-SSE trending Cambay rift basin is seismically less active and bounded to the east by the Great Boundary Fault, and the Aravalli-Delhi Mobile Belt covers the western edge.

Most cases of dynamically triggered events (i.e., unexpected rise in seismicity after large remote earthquakes) are noticed near plate margins or volcanic/geothermal areas (e.g., Peng and Chao, 2008; Aiken and Peng, 2014; Bansal et al., 2016 and reference therein). In the stable regions, remote earthquake triggering was identified in regions with meagre background seismicity rates (Gomberg et al., 2004; Velasco et al., 2008). On the other hand, some studies documented that regions with high background seismicity or regions that have experienced significant earthquakes (or hosted large historical events) in the past are more susceptible to dynamically triggered earthquakes (e.g., Hough et al., 2003; Peng et al., 2010b; Jiang et al., 2010; Dixit et al., 2022a; b), because these regions denote the zones of weakness having several nucleation points (Hill and Prejean, 2007; Savage and Marone, 2008). Hence, the Gujarat region is susceptible to dynamic triggering, and the KRB has a higher chance of being triggerable among SH and ML. The dynamic triggering in other intraplate regions has also been documented (Peng et al., 2010b; Jiang et al., 2010; Wu et al., 2011; Wu et al., 2012; Yao et al., 2015; Bansal et al., 2018; Li et al., 2019).

This article systematically analyzed the local seismicity following the 2015 Nepal event to determine dynamic triggering in the Gujarat region. The ISR catalog has been used to measure the seismicity rate change in the region. 10 Hz high pass waveform data is studied to find micro-earthquakes missing in the ISR catalog. Along with the Nepal earthquake, 2015, we also examined other recent earthquakes having comparable dynamic stresses in the region (i.e., 16 April 2013, M_w7.7 Iran, 24 September 2013, M_w7.7 Pakistan, and 26 October 2015, M_w7.7 Afghanistan Earthquakes (Figure 1A). It is, therefore, helpful to look for triggering due to large earthquakes having comparable dynamic stresses in the region.

The Kachchh Rift Basin, Saurashtra Horst, and the Mainland Gujarat region differ in geology, seismicity, and tectonics, so we selected and separated events in the ISR catalog for the above three regions. The β value quantifies a change in seismicity from the observed seismicity rate before and after the mainshock (Matthews and Rosenberg, 1988; Aron and Hardebeck, 2009). β = N a – N   T a T N T a T 1 − T a T

Where T_a is the period after the onset of the P wave of the remote mainshock (Triggered Window), T is the total time length of the triggered window plus the background seismicity window (Time window before the arrival of the remote mainshock P wave), N_a is the number of events during the triggered window, and N is the total events during the triggered plus background window. In the study, we used T=48 h and T_a=24 h.

A β≥2 indicates a significant increase in the seismicity, whereas β≤−2 corresponds to a significant decrease in the seismicity (Hill and Prejean, 2015). We computed the β value from the ISR catalog during the 24 h of the main shock. We observed that the β value for the 2015 Nepal event is less than two indicating an insignificant change in the seismicity. A vital portion of triggered events may be missing in the existing catalog because higher amplitude surface waves mask local microearthquakes (Bansal et al., 2018; Dixit et al., 2022a; b). The analyses of continuous waveform data for identifying missing microearthquakes are more viable (e.g., Gomberg et al., 2004; Prejean et al., 2004; Peng et al., 2010a).

Numerous studies examined the waveforms within 1 h after the remote mainshock (e.g., Velasco et al., 2008; Jiang et al., 2010; Li et al., 2019) because most of the seismic activity due to triggering occurred either at the time of the arrival of large surface wave or immediately after it. Hence, we used waveform data in a continuous form recorded by GSNet, 1 h before and after the P-wave of the 2015 Nepal mainshock. Next, we cut the waveforms with surface wave phase velocities between 5 km/s and 2 km/s, covering most of the surface waves, and computed the peak ground velocity (PGV). We measured the Peak Dynamic Stress (PDS) using the relationship σ = μ P G V υ , where σ is dynamic stress, υ is the nominal phase velocity (assumed to be 3.5 km s-1), and μ is the shear rigidity (which is assigned a nominal value of 35 GPa) (Aiken and Peng, 2014).

Next, we corrected the raw waveform data of all three components by removing the instrument response. After applying the instrument correction to the raw waveform data, we rotated the north-south and east-west components to a great circle path to obtain the transverse and radial components. Since, the triggered events are generally observed in the high-frequency range, the waveform data is filtered using a 10 Hz high pass filter to confirm possible triggered events. Further, a spectrogram is also generated using a 0.5 Hz high-pass filter using the waveforms containing the vertical component. The GSNet recorded the 2015 Nepal event at 22 stations, eight in the Kachchh, five in the Mainland, and nine in the Saurashtra region. The analyzed raw and 10 Hz high-pass filtered waveforms of the Nepal event recorded at GSNet are shown in Supplementary Figure S1.

We examined the 10 Hz high-pass filtered waveforms and spectrograms at the permanent station in the KRB, SH, and ML regions during the 2015 Nepal event (Figures 2–4). The peak dynamic stresses at the vertical component in the KRB, SH, and ML is greater than 52 kPa. Despite the region’s high peak dynamic stresses, we found no evidence of remote triggering during the 2015 Nepal mainshock. However, we identified only one event after the 2015 mainshock at station UNA (Figure 3).

To study the remote dynamic triggering in detail, we have also examined the other recent large earthquakes with comparable peak dynamic stresses (>50 kPa), namely, 16 April 2013, M_w7.7 Iran, 24 September 2013, M_w7.7 Pakistan, and 26 October 2015, M_w7.5 Afghanistan Earthquakes (Supplementary Figure S2–S10). We followed the same procedure as the 2015 Nepal event. We found that there is an indication of an increase in seismicity during 2013 Pakistan (Supplementary Figure S6) and the 2015 Afghanistan (Supplementary Figure S8,S10). However, as in the case of the 2015 Nepal event, we found only one or two earthquakes just after the surface waves of the 2013 Pakistan and 2015 Afghanistan earthquakes. Notably, these identified microearthquakes are not listed in the ISR catalog. To check the seismicity increase, we added these identified events to the ISR catalog (green circles in Figures 5–7) and calculated the β values again. In addition to manually picking local events, we identified local peaks using STA/LTA (Short-term algorithm/Long-term algorithm) and then calculated the β. We found that the β value is less than 2 for all four mainshocks at all the stations, suggesting that the triggering is not statistically significant (Figures 5–7, Supplementary Figure S11,S12).

The study suggested remote dynamic triggering does not occur in the Gujarat region following the 2015 Nepal earthquake. Similarly, with their significant peak dynamic stress, other large earthquakes did not trigger seismicity. The research finding suggests that the surface wave amplitude is not the only factor that controls the remote dynamic triggering. The rupture direction and orientation of faults with incoming surface waves may play an essential role. The following factors, either individually or in combination, may be responsible for dynamic triggering: 1) The region should be critically stressed, 2) significant low-frequency surface waves, 3) rupture propagation direction of the mainshocks, 4) compressional tectonic regime, and 5) direction of incoming waves from remote mainshocks.