We use Global Navigation Satellite Systems (GNSS) data to characterize earthquakes for rapid response. We estimate the peak ground velocities (PGV) associated with an earthquake, and subsequently use them to arrive at earthquake magnitudes and ground motion. We have found that seismic data can be interchangeably used with said geodetic data to arrive at these early estimates.  

The figure shows the earthquake magnitude estimates computed using GNSS and seismic PGVs for the 2021Mw 8.2 Chignik event in Alaska , and how comparable they are to one another.

(Parameswaran et al., SRL, 2023)  

Earthquakes release stresses that accumulate through process such as plate motion. This release can either increase the stresses acting on a neighboring fault leading to its failure, or slightly relieve a fault from accumulating stress, pushing it away from failure. 

Here is an instance where we find that the 2021 Mw 8.2 Chignik earthquake in Alaska originated on a fault segment that experienced an increase in stress caused by the 2020 Mw 7.8 Simeonof earthquake. 

(Elliott et al., Science Advances, 2022)

Earthquakes are a reflection of the state of stress in the region. One way to investigate the associated stress field is to use dense background seismicity. 

Here is an example where we analyzed the seismicity along the western segment of the South Iceland Seismic Zone (SISZ) adjacent to the Hengill volcanic system. We used earthquake relocations to show spatio-temporal evolution of seismicity in the region, and performed stress inversions to show positive correlations between seismic and volcanic activities in the SISZ and Hengill. (Parameswaran et al., JGR-Solid Earth, 2020)

When stress regimes in a region change over time, they manidest through varying earthquake mechanisms in the resultant seismicity. The parameters associated with these mechanisms can be inverted to quantify the changes in stress fields. 

Using earthquake focal mechanisms  from the western part of south Iceland we were able to illustrate varying stress fields between 1991 - 1999 in this region, both spatially and temporally. These changes could be attributed to activity in neighboring volcanic and tectonic systems. (Parameswaran et al., JGR-Solid Earth, 2023)

At times, fault geometries of seismic ruptions are not evident or seem complex. Earthquakes also tend to occur in regions with limited seismic history, with data recorded only by far-field stations. In such cases, forward modeling can prove to be an effective tool to test for possible geometries that best replicate observed data. 

Here is an example where the 2012 Indian Ocean intra-plate earthquake (Mw 8.6) is modelled using SPECFEM3D with the mainshock nucleating in the upper mantle on a listric (non-planar) fault extending into the brittle crust. This work has motivated me to enhance communication and collaboration with the larger geoscience community that studies viscoelastic rheologies and failure conditions. (Parameswaran et al., PEPI, 2020)

Seismic source inverse modeling is an effective tool to estimate the kinematic properties of an earthquake. These kinematics often represent the tectonic state of the region and the changes occuring in it.

We analysed over 12 large earthquakes around the Indian plate boundary to study different convergent settings along the collision zone. To the left are the source models for the 2015 Mw 7.8 and Mw 7.3 Nepal earthquake sequence. We found that the Mw 7.3 earthquake was likely a triggered aftershock. The rupture areas clearly delineated locked regions along the Himalaya collision boundary. (Parameswaran et al., JAES 2015; Parameswaran et al., IJES, 2016; Rajendran et al., ESR, 2017)