In a previous installment of this series, a colleague eloquently stated that education is our best ally for earthquake and tsunami preparedness in that knowledge of past events and understanding the science behind these natural hazards will help us prepare for the next “big one.” While we do not have recordings of earthquake ground motion or tsunami flow height and speed from the last Cascadia Subduction Zone event in 1700, we have recordings from similar, more recent events.
Computer-based simulation is an ally of equal importance for earthquake and tsunami preparedness. With recordings from previous events, we can use simulation to predict future hazard intensities, the response of critical infrastructure to these future hazards, and perhaps most importantly, to assess uncertainty in those predictions. Making decisions in the face of large uncertainties and incomplete information is not a straight line, and it is not a simple binary outcome of safe or unsafe.
The Pacific Earthquake Engineering Research Center, of which OSU is a core member, has advocated for performance-based earthquake engineering (PBEE) for over 20 years. PBEE uses simulation of structural response to earthquake ground motion to account for uncertainties in structural properties and earthquake intensity. Structural performance criteria range from immediate occupancy, where occupants can immediately return and continue to occupy the space, to collapse prevention or life safety, where the objective is to prevent complete collapse of the building even though it will not be occupiable after the hazard.
The PBEE framework allows stakeholders to make decisions about planning and mitigation for their buildings or communities through investments in retrofits, ordinances, and zoning. This framework provides a science-based approach to answering the ultimate question: should we invest in seismic retrofit now to mitigate losses later?
Simulation played a critical role in the recent development of design guidelines for tsunami loading on bridges. This project was led by the Oregon Department of Transportation (ODOT) with contributions from DOTs in the other four Pacific states (California, Washington, Alaska and Hawaii) and researchers at OSU and other universities.
After validating computer-based models against experiments of wave loading on bridges conducted in Japan and at OSU’s O.H. Hinsdale Wave Laboratory, researchers developed loading equations based on tsunami height and speed determined from probabilistic tsunami hazard analysis (PTHA). The PTHA considered many triggering events around the Pacific rim giving maximum credible tsunami height and speed at coastal bridge locations, for example, at Seaside on the Oregon Coast, due to a magnitude 9.0 earthquake in Alaska’s Aleutian Islands. While an obvious solution is to build bridges at higher elevations, this is not always feasible and computer-based simulation allows bridge engineers to estimate the forces that structural connections must resist to keep a bridge deck on its piers during a tsunami event.
Researchers at OSU also use simulation to assess evacuation routes along the Oregon Coast in the event of a tsunami. In the next Cascadia Subduction Zone event, there may be only 15 to 20 minutes after ground shaking to evacuate from the tsunami inundation zone. What routes should people take? Is it faster to run to higher ground than to drive? What if bridges are severely damaged due to ground shaking? Is there a sufficient number of alternative routes in the transportation network? Simulation tools such as agent-based modeling can address these questions and help coastal municipalities prepare.
In addition to the performance of buildings and bridges, the functionality of water, power, telecommunications, and socio-economic networks also influences recovery after an earthquake or tsunami. The connections between these networks are incredibly complex and require high-performance computing to assess their performance after an event. OSU researchers are working with researchers at the National Institute of Standards and Technology (NIST), the National Center for Supercomputing Applications (NCSA), and other universities in the Center for Risk-Based Community Resilience Planning to build a computational framework for decision-making considering these interconnected networks when planning for and recovering from disasters at local, state, and regional levels.
Everyone would love to push a button and get an answer for the best preparedness and recovery strategies. However, like my colleague said in that previous installment, “there’s no free lunch”. The development, validation, and use of decision-making software requires education in programming as well as the underlying structural, geotechnical, and coastal engineering—programming is not just a computer science issue.
Although courses in C++ and Fortran have been cut from many engineering curricula across the country, computational thinking has made a comeback through the integration of Python in many graduate and undergraduate engineering courses. OSU student organizations such as the Hackathon Club and Girls Who Code help build community by hosting coding events on campus and online. In addition to earthquake and tsunami hazard preparation, programming skills and computer-based simulation will be important in developing decision-making tools for other natural and man-made hazards we face — pandemics, blasts, and wildfires.
Michael H. Scott is a professor of structural engineering in Oregon State University's School of Civil and Construction Engineering.
Get local news delivered to your inbox!
Subscribe to our Daily Headlines newsletter.