Ground motion prediction equations for Northeast India: A hybrid approach using observed and simulated data

Abstract

Given the potential for large-magnitude earthquakes to occur in northeastern India in the future, risk mitigation measures such as hazard assessment and strategy implementation are critical for areas with dense populations. Equations for predicting ground motion, termed GMPEs, tailored to specific characteristics play a pivotal role in conducting hazard analyses at both the macro and micro scales. Researchers have leveraging developed GMPEs for various regions, utilizing recorded strong ground motion data. Nevertheless, the limited availability of such data poses a substantial challenge in developing new, region-specific GMPEs that cover all magnitudes and distances. In light of this challenge, the present study aimed to formulate GMPEs by combining both recorded and simulated data. The approach involves utilizing finite fault simulation to create synthetic ground motion data at various surface locations where previously recorded data are accessible. Initially, the consistency of the adopted technique is evaluated by comparing simulated and recorded time histories at specific sites, considering various parameters, namely, response spectra, Fourier spectra, and peak ground acceleration. Subsequently, ground motion is simulated across various magnitude ranges at different locations, accounting for site effects. Moreover, the validity of the developed GMPEs is tested using earthquake records that are not utilized for development. The resulting GMPEs effectively predict ground motion across various magnitude ranges (M_w3.9–8.5) and hypocentral distances (20–560 km). These GMPEs can serve as valuable tools for ground motion estimation in the future. The newly developed GMPE is presented as follows:$\mathrm{Log}\left(Y\right)=-2.0941+0.4991M-1.0123\log (R+e^{-6.4001M)}\pm 0.2277$where M denotes the magnitude, Y represents the peak ground acceleration in g, and R represents the hypocentral distance.