Regional Ground Motion Duration Prediction Model for Subduction Regions

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

Developing an understanding of duration of strong-shaking is important when characterizing engineering ground motion because the duration of strong shaking can be correlated to critical damage indices in the performance-based design of civil engineering structures. Most existing empirical models for duration are for crustal earthquakes. Using the Next Generation Attenuation Subduction Zone Database, we derived a regional empirical model for duration for subduction zone earthquakes. In evaluating the subduction data from large magnitudes (M>8), we found that the commonly used measure of duration based on the time between 5% and 75% or the Arias intensity does not work well if the ground motion has multiple sections of strong shaking that are separated in time. In crustal earthquakes, the slope of the Arias intensity is approximately constant over the 5% - 75% range. With the concept that the parameter of interest is the actual duration for which the shaking has significant energy, we propose a revised algorithm to computing significant duration that computes the slope of the Arias intensity in 5% increments and then summed the time increments with the largest slopes to reach 75% range, effectively skipping the time increments with small slopes (low energy). This new method provides a measure of duration that captures the duration of strong shaking in large subduction earthquakes but remains consistent with the current definition for typical ground motions without large separations in the strong shaking in the recording. Using this new definition, we derive empirical models for the duration for subduction earthquakes in Japan, Taiwan, South America, Cascade, Alaska, Central America & Mexico and New Zealand. Compared to the duration models for crustal earthquakes, the subduction duration model shows stronger distance dependence and weaker magnitude dependence.

Presenting Author: Nicolas M. Kuehn

Authors