Performance Based Tsunami Engineering Methodology

Research Team

  • Harry Yeh, Professor, Oregon State University (PI)
  • Patrick Lynett, Associate Professor, University of Southern California
  • Hong Kie Thio, URS Corporation


Recognizing the present status of research activities by NSF, FEMA, NOAA, USGS, and NTHMP, a series of meetings were convened to address how PEER can best contribute to tsunami research, and what the effective research would entail. Discussions at the meetings led us to identify PEER’s unique and critical contribution: To develop an effective methodology for tsunami analyses of ‘critical’ structures and lifelines. The concept is similar to PEER’s Performance Based Earthquake Engineering (PBEE). The methodology requires comprehensive integration of physics-based engineering knowledge and tools together with in-depth analysis. Tsunami analysis requires interrelated information about hydrodynamics, bathymetry and topography, man-made structures, and natural elements such as vegetation. Therefore, performance-based tsunami engineering (PBTE) requires comprehensive local analysis even when we study for a specific individual structure at a specific location. Consequently, it is necessary to develop the tool based on a GIS platform.

The objective of this tsunami research is to develop a methodology for Performance-Based Tsunami Engineering (PBTE) that focuses on the assessment of critical coastal structures and lifelines. Ultimately, the PBTE methodology will be fully integrated with the existing PBEE so that assessments of multi-hazard scenarios can be made. The scenario of multi-hazard associated with a subduction-type earthquake could be a series of very strong ground motions of long duration followed by tsunami inundation. Critical coastal facilities and structures along the subduction zone must be evaluated and designed for such extreme loadings. The related, immediate goal is to develop a user-friendly GIS-based portal that allows us to download necessary tsunami intensity measure (IM,) which consists of the time history data of flow depth and velocity at the location of interest. With those data one can estimate the maximum hydrodynamic forces, impulse forces, debris impact forces, and their moments.

Milestones and Deliverables

This is a two-year research program. In Year 1, we will develop a prototype of Tsunami Hazard Analysis Module to provide IM for a testbed coastal region. In Year 2, we will expand the module for probabilistic tsunami analysis. The deliverable will be a user-friendly portal that can be used to download tsunami data for use in damage analysis. Once completed and validated, the portal will be transferred to PEER and eventually incorporated in OpenSees.

Scope and Project Plan

The proposed research is the initial stage of PBTE development, and is a two-year program.

In Year 1, we obtain DEM and simulated tsunami inundation data for Port Hueneme, the location selected as a testbed by the PEER Tsunami Working Group. Port Hueneme is home to a US Navy port and is also used for commercial shipping and receiving. For the development phase we will select one specific tsunami event (say, for example, a 2500 year return period event). A GIS-based web portal will be developed for data retrieval (DEM and the time series of pre-computed tsunami-inundation data) – note that this will be accomplished by modifying an existing portal that has already developed for analysis of climate data at Oregon State University. This portal will allow viewing and downloading of data at a user-specified location. This portal will also provide functionality for the user to upload his/her analysis and results based on the downloaded tsunami data. Once this initial stage of the portal is established in Yr. 1, we plan to invite several key participants to perform their analyses for specific objectives (to be determined later). This will serve as a test run to validate functionality of the developed PBTE tools. In addition to storing the pre-computed time-series inundation data, we will develop an alternative approach. With the use of offshore tsunami conditions that are parameterized with maximum and minimum water elevations and wave period at the depth of 100 m, a simplified algorithm is used to generate time-series inundation data for a user-specified location. We will attempt to use a 1D or semi-1D numerical model to produce the inundation data. This alternate scheme requires much less data storage, but requires running the simplified numerical computation for the location of interest.

In Year 2, we will expand this portal functionality to handle the probabilistic tsunami analysis. Time series data of inundation depths and velocities for approximately 25 probabilistic scenarios will be stored. The time-series data for the specified probability will be obtained based on the pre-computed 25 probabilistic datasets by interpolation. Likewise, the data at the specified location will be obtained by interpolating the pre-computed data at the neighboring grid points. In addition, the same probabilistic analysis will be made also with the data generated with the alternate method as described earlier: use of the 1D or semi-1D numerical model to produce the data based upon offshore tsunami conditions.
Yeh will coordinate this research effort; Thio will provide the DEM, tsunami inundation data, as well as offshore tsunami parameters for the alternate method; Lynett will provide the 1D or semi-1D model to generate the inundation data from the offshore parameters; Keon will develop the prototype of the PBTE portal.

Other similar work outside PEER, and differentiation of this work

FEMA is currently developing a tsunami methodology in the HAZUS-MH for tsunami hazard-damage-losses analysis. However, FEMA’s objective and the level of estimation are quite distinct from the proposed. Our objective is to develop an effective methodology for credible tsunami analyses of ‘critical’ structures and lifelines.

Interaction of this work to other related PEER work

This PBTE Intensity Measure (IM) functionality will be utilized (even in its development stage) for our companion research projects: namely, evaluation of coastal bridge performance under tsunami loading (Scott, PI), and assessment of uncertainty and model variability (Lynett and Thio, PI). For example, we can perform exercises for evaluation of a (hypothetical) bridge
located in Port Hueneme. Uncertainty of the model performance can be tested with the use of this portal; especially once we prepare the portal for more definitive sites (likely some Japanese sites where accurate and rich field data are available).
With the budget under this research task, Yeh will work on the other two tasks:

  • Assessment of uncertainty and model variability (Lynett and Thio, PI): Yeh will collect and validate field data and DEM for 2~3 Japanese sites for the 2011 East Japan Tsunami, where quality data of velocities are available. The data will be stored and made downloadable to our research collaborators (Lynett ad Thio) with the use of ISEC:
  • Preparation for the UJNR bridge workshop (Scott, PI): With Lynett, Yeh will assist Scott for organizing the workshop.

Potential Impact of Project on Practice

The foregoing development should be considered as an initial study aiming at the full-scale methodology development for PBTE. This research platform will promote strong collaboration among hydraulic, structural, and geotechnical engineers, and is considered a key breakthrough for success in PBTE.

Comments are closed.