← Back to Sessions

Sensitivity Analysis of Conditional Mean Spectrum Ground Motion Selection Procedure

Description: 

Assessing the seismic performance of structures requires ground motions as inputs. Ground motion selection is, therefore, a topic of significant interest in dynamic structural analysis and performance-based earthquake engineering. In this vein, the most common approach to selecting ground motions is to select time series whose response spectra match a target response spectrum. The uniform hazard spectrum and the conditional mean spectrum (CMS) are commonly used target spectra. The CMS is taken by conditioning on spectral acceleration at the period of interest, and the response spectra at other periods have variance. The response spectra at these periods are derived using the correlation coefficients between the spectral accelerations at them and the period of interest. Thus, one of the major steps in ground motion selection based on CMS is to parameterize the correlations between the spectral accelerations at various periods. In this study, we perform a sensitivity analysis to ascertain how the selected ground motions' different ground motion intensity measures are affected by uncertainties of the model inputs when the CMS ground motion selection model is utilized. Moreover, it investigates the importance of the various input parameters for different ground motion intensity measures. For this purpose, we use an efficient information theoretic that utilizes the concepts of entropy and mutual information. The major advantages of the information-theoretic approach are the fast computational time and its ability to generalize to stochastic simulation models, such as ground motion selection procedure based on CMS, compared to the traditional methods, such as variance-based global sensitivity analysis methods.

Session: Understanding and Modeling the Uncertainties in Earthquake Ground Motions [Poster]

Type: Poster

Date: 4/20/2023

Presentation Time: 08:00 AM (local time)

Presenting Author: Shahram Pezeshk

Student Presenter: No

Invited Presentation: 

Authors