← Back to Sessions

Empirical Correlations of Response Spectral Ordinates, Arias Intensity (AI) and Cumulative Absolute Velocity (CAV) With Fourier Spectral Ordinates of Ground-Motion and Associated Variabilities

Description: 

Empirical correlations between various ground-motion intensity measures (IMs) are of particular interest in engineering applications such as conditional ground-motion selection and vector probabilistic seismic hazard analysis (VPSHA). Past studies have mostly examined the inter-period correlation of response spectral ordinates e.g., pseudo spectral acceleration (PSA) and the correlation of non-spectral IMs such as arias intensity (AI) and cumulative absolute velocity (CAV) with PSA. However, from the perspective of engineering seismology it is rather enigmatic how these engineering IMs are correlated with Fourier spectral ordinates -in other words- with the seismological parameters that determine amplitude and shape of Fourier spectrum of input ground-motion.

Bora et al. (2016) presented a theoretical analysis using random vibration theory (RVT) to decipher the complex relationship between response spectral ordinates (e.g., PSAs) and the Fourier amplitude spectrum of the input ground-motion. In the present study, I present empirical correlations of PSAs, AI and CAV with Fourier spectral amplitudes of acceleration ground-motion. The correlations are examined in terms of residual correlations between the chosen IMs obtained from the NGA-West2 database . The residual correlations are investigated separately for between-event and within-event residuals of ground-motion. For that purpose, the PSA ground-motion model (GMM) of Abrahamson et al. (2014), Fourier model of Bora et al. (2019) and AI-CAV models of Campbell and Bozorgnia (2019) are considered. In addition to validating the inferences from Bora et al. (2016), the correlation results clearly identify the frequency range in the input ground-motion that dominates the IMs such as PGA, PGV, AI and CAV. Furthermore, I also examine the scenario-dependence of such correlations in addition to investigating regional variations. The results presented in this study have significant implications for host-to-target adjustment of ground-motion models mainly for site-specific hazard analysis and physics-based numerical simulation of ground-motions.

Session: High-frequency Ground Motion Measurements, Assessments and Predictions

Type: Oral

Date: 4/18/2023

Presentation Time: 08:30 AM (local time)

Presenting Author: Sanjay Bora

Student Presenter: No

Invited Presentation: 

Authors