Project Title/ID Number Empirical Characterization of Basin Effects on Site Response—2252002
Start/End Dates 10/1/02—9/30/03
Project Leader Jonathan Stewart (UCLA/Faculty)
Team Members Yoojong Choi (UCLA/Grad Student), Robert Graves (Industrial Collaborator)
Project goals and objectives

This Year 6 project is an extension of the Year 5 project having the same title. As described in the original project proposal, the general intent of this work is to develop a statistically robust engineering model for characterizing basin effects on ground motion intensity measures (IMs). The approach taken is to evaluate residuals between data and predictions, and correlate these residuals to basin geometric parameters. The predictions used to evaluate the residuals will, to the extent possible, remove the effects of site, path, and shallow ground response in an average sense, so that the deviations from data can be attributed to basin response.

Role of this project in supporting PEER’s vision

The project will allow for more robust evaluations of ground motion intensity measures for use in Probabilistic Seismic Hazard Analysis (PSHA), which is an integral component of Performance-Based Earthquake Engineering (PBEE). The results could be especially significant for regions located near the basin edge or on deep basin sediments.

Methodology employed

To realize the goals/objectives outlined above, the following tasks are being performed:

  1. We have used available basin models to identify basin geometric parameters for strong motion sites and have prepared maps showing the spatial distribution of these parameters for Los Angeles and the San Francisco Bay Area.
  2. We have estimated the uncertainty in basin depth parameters for southern California through analysis of residuals from oil well logs.
  3. We have developed a statistically robust model for site amplification of spectral acceleration based on a “shallow” metric of site condition (Vs-30) – this model is needed to define the “predictions” from which residuals will be evaluated. A journal paper on the topic is in preparation.
  4. As needed, we will develop statistical models for other IMs as needed by PEER through interaction with testbed efforts.
  5. We will calculate residuals for IMs using the available ground motion inventory.
  6. We will perform statistical analyses to identify the most relevant basin geometric parameters by finding those that best describe trends in residuals.
  7. We will prepare a final project report and journal paper.

Work to date has focused principally on Tasks (a)-(c). We have recently begun Task (d) for an IM that we anticipate having basin parameter sensitivity (significant duration). Remaining work is ongoing.

Brief description of past year’s accomplishments and more detail on expected Year 6 accomplishments

The principal outcome of work to date is the amplification model described in Task (c) above as well as the basin parameter uncertainty evaluation in (b).

The amplification model has several unique attributes. First, it is the only available model that provides statistically robust estimates of ground motion amplification as a continuous function of both Vs-30 and reference peak acceleration (PHAr).

The model is described by the following equation:

One of the innovative features of the model is the formulation of b, which reflects the Vs-30-dependence of amplification as shown in the figure below. The line in red represents the model selected, the black dots indicate the results of regressions within specific Vs-30-categories.

Another interesting feature of the model concerns the formulation of error term . We investigated the dependence of on Vs-30, distance (r), and magnitude (m). We identified a significant dependence on r for r < 20 km, no significant dependence on m, and a weak dependence on Vs-30.

Based on this model, we find a bias in existing NEHRP site factors, as shown below.

Other similar work being conducted within and outside PEER and how this project differs

A project funded by SCEC (PI: Ned Field) evaluated the effect of basin depth on spectral acceleration using data from southern California. This project is examining additional basin geometric parameters and an additional region (northern California). Therefore, the results will be more comprehensive, and more thus more useful for performance-based engineering.

An ongoing project funded by CSMIP (PI: Graves) is examining the generation of basin induced surface waves and their correlation with geologic parameters such as distance to basin edge. This work is similar to the present study, but is focusing on long-period ground motions. Moreover, that study formulated correction factors relative to the Abrahamson and Silva attenuation relationship and not to amplification-adjusted ground motion predictions.


Larger View

Plans for Year 7 if this project is expected to be continued

I anticipate that the basin project will be completed within Year 6 for the IMs of spectral acceleration and significant duration. Possible Year 7 topics could include site amplification models for other IMs, interpretation of the Taiwan earthquake data, or work related to the development of vector hazard capabilities. Alternatively, my work could be re-directed into research related to soil-foundation structure interaction (SFSI). Ideally, my research topic would be selected through open discussion between myself and the PEER leadership.

Larger View

Describe any instances where you are aware that your results have been used in industry

My ground motion related work from previous phases of PEER research has been used in PG&E and Caltrans, and is being implemented within the next-generation attenuation (NGA) project. Previous work on soil-structure interaction partially funded by PEER has been implemented into the NEHRP provisions for new structures and is in the process of being implemented into design documents for existing structures (ATC-55 project).

Expected milestones
  1. Assemble database of critical basin parameters for strong motion stations in active regions, with emphasis on the San Francisco Bay Region and Los Angeles.
  2. Develop site amplification model for Vs-30 ground condition – provides consistent baseline results for evaluation of basin effects (see Item c below).
    Develop statistical models through formal regression analysis of ground motion amplification associated with basin effects. These models will quantify amplification beyond that associated with the amplification model from (b) above.

PEER report presenting research results along with summary journal and conference papers.