Developing Data-Driven Stochastic Seismological Parameters of CENA From the NGA-East Database

Session: Earthquake Ground Motion and Impacts [Poster]
Type: Poster
Date: 4/30/2020
Time: 08:00 AM
Room: Ballroom
Description: 

The fundamental idea of this project is to obtain a set of self-consistent stochastic seismological parameters for Central and Eastern United States (CENA). We are using median 5% damped pseudo-spectral acceleration (ROTD50) with rupture distances (Rrup) less than 500 km from the NGA-East database and the point source stochastic model (Boore, 2003) to determine seismological parameters that minimize the residuals over the usable frequencies of the data using a genetic algorithm. The largest Vs30 value for the database is 2000 m/s. The final product of this study is a set of well correlated seismological parameters compatible with single and generalized additive and multiplicative double corner frequency models.

Parameters obtained in this study include a frequency-dependent tri-linear geometric spreading model, site kappa for very hard rock in CENA, frequency-dependent quality factor, eight magnitude dependent parameters of generalized additive and multiplicative double corner frequency model (Boore et al., 2014) and magnitude dependent point-source correction factor (also known as pseudo depth). The double corner frequency can easily be converted to single corner frequency model and a magnitude dependent stress drop parameter and corner frequency, which has the same entity of being data-driven and self-consistent. The variability of the parameters is assessed using Bayesian Metropolis-Hastings algorithm. This process will accept or select random proposal sets of parameters by assessing the likelihood and the prior information we already have from the GA analysis or other studies.

The final objective of this research is to develop an updated hybrid empirical ground motion model to map the ground motion of the higher magnitudes from a richer database by scaling the ground motion in the target region (CENA) by the ratio of stochastic over empirical ground motion obtained in a target region.

Presenting Author: Nima Nazemi

Authors