ACADEMIA

By Kathleen Ricker, NCSA -- A new earthquake engineering cyberenvironment could change the way communities prepare for the worst -- and help translate science into practical results more quickly. In the winter of 1811-1812, the New Madrid seismic zone -- a rift in the earth's crust that stretches across parts of Illinois, Missouri, Arkansas, Tennessee, and Kentucky -- generated a series of earthquakes now estimated to have been between 7.8 and 8.2 on the Richter scale. The massive temblors reversed the flow of the Mississippi River for several days, created lakes in Missouri and Tennessee, shook the White House in Washington, D.C., and rang church bells in Boston, Massachusetts. MAEviz visualization showing building and bridge damage in Memphis, Tennessee, after a hypothetical earthquake. Buildings are represented by geometric shapes (red crosses for hospitals, purple squares for schools, green circles for fire stations, etc.); bridges are represented by white circles. All include vertical bars indicating the amount of damage to the structure, with red bars indicating probability of collapse and white bars indicating probability of no damage. Information about the MLGW gas network repair rate at various locations is also included. Users may superimpose as many layers of detail as desired. Almost two centuries later, the New Madrid seismic zone contains large population centers that are crucial economic and industrial hubs. A major earthquake could result in tens of thousands of human casualties, with thousands dead in southern Illinois alone. Economic loss, both regionally and nationally, could be anywhere between $100-200 billion. And the population displacement caused by a temblor of the size of the 1811-1812 quakes could, by some estimates, be greater than that caused by Hurricane Katrina. "The amount of energy released by these earthquakes is beyond normal imagination," says Amr Elnashai, director of the Mid-America Earthquake Center at the University of Illinois at Urbana-Champaign, where research is focused mainly on the New Madrid seismic zone and its hazard. "We would be looking at roughly 10 times the area affected by a similar magnitude earthquake on the West Coast." Metropolitan centers like Memphis and St. Louis would be severely damaged, as would the surrounding counties in Illinois, Kentucky, Arkansas, Ohio, and Indiana; even Chicago high-rises would be affected. Closer to ground zero, everything would be severely damaged, including electrical grids, water purification plants, nuclear reactors, and sewage systems. Worst-case scenario The New Madrid seismic zone has more earthquake activity than any other region east of the Rocky Mountains, although the quakes are minor and cause little damage. However, federal, state, and local governments, as well as utility companies and businesses, are responding to the possibility of another major quake by retrofitting bridges, buildings, pipelines, railways, highways, and other crucial infrastructure such as nuclear power plants and electrical power stations. But with continual urban growth and redevelopment, deciding where best to apply limited resources is difficult, especially when taking emergency response into account. Preparation and crisis management require access to the latest science and other kinds of information that can change rapidly and frequently. Moreover, because seismologists are still exploring the unique characteristics of the New Madrid fault itself, such as the ways in which the thick Mississippi River sediments amplify and liquefy ground motion, new findings can have a major impact on disaster planning. To provide a conduit between researchers and practitioners, the Mid-America Earthquake Center (MAE) and NCSA are jointly developing the MAEviz loss assessment system, a cyberenvironment that gives decision makers access to tools aimed at helping them assess risk and determine how to allocate resources for mitigating risk. Based on technology developed at NCSA and the University of Michigan, MAEviz has a modular structure and includes analytical tools and data repositories for modeling earthquake damage. One module simulates the hazards posed by earthquakes of different intensities and durations. Others provide detailed inventories of bridges, pipelines, networks, railways, streets and highways, and buildings, including essential facilities like hospitals and schools, which allow users to assess potential loss. Combining these modules into a workflow with a third type that encodes knowledge about fragility (the probability of damage to facilities and networks) produces loss estimates that still other modules translate into terms of human and economic loss, demonstrating where and how reducing fragility can best reduce those losses, thus helping decision makers determine the most cost-effective retrofits. MAEviz's modular architecture not only makes it more flexible than other loss assessment software, but also more dynamic. As researchers produce algorithms and data that can provide more accurate estimates, the new modules are swapped in so decisions can always be based on the best science available. And because MAEviz is a network-aware environment, new software and data can be downloaded, and new results can be uploaded and shared between collaborators through a secure web portal. If disaster strikes, plans can be retrieved from the network and damage assessments can be recomputed using the latest information. Putting science into practice For Jim Myers, chief architect of MAEviz and leader of NCSA's Cyberenvironments and Technologies Directorate, MAEviz is an example of how cyberenvironments can accelerate the interaction between scientific research and its application in the field. "Five years ago you would have built a MAEviz as an application, and you would have just accepted that the scientific knowledge behind it would grow increasingly outdated," Myers says. "While people might have heard about better methods or newer data at conferences and read about them in publications, they wouldn't have been able to do more than idly speculate about how the new information might affect planning. Why shouldn't you be able to just plug in a new module and immediately see the potential impact of the latest research?" Myers and the MAE researchers envision MAEviz as a collaborative bridge between researchers, engineers, and planners. Researchers create new analysis modules and data sets, engineers create and post scientifically rigorous scenarios based on the new information, and planners download appropriate scenarios to assess the impact on strategy. If the potential impact is significant, all three groups can coordinate to speed additional research and validation efforts, resulting in informed decisions. MAE conducted a pilot collaboration with a regional utility, Memphis Light, Gas, and Water (MLGW). A customized version of the software, MAEviz-MLGWgas, was used to assess potential earthquake damage to the company's existing cast-iron piping network -- a small portion of their gas network, but substantial enough to demonstrate MAEviz's potential as a loss assessment tool and an asset management environment for MLGW. Since then, MAE has also collaborated with other partners to provide modules customized for different areas and needs, including the transportation network for Charleston, South Carolina. MAE is currently working with FEMA to create a comprehensive loss assessment model in the event of a New Madrid earthquake. That model uses MAEviz in combination with FEMA's broader, more general loss assessment tool, HAZUS. MAEviz is attracting interest from other parts of the world as well, especially earthquake-ravaged areas such as Turkey, Pakistan, and Indonesia. The MAE Center is currently partnering with Istanbul Technical University and the Municipality of Istanbul to build a similar framework with a Turkish-language interface for that city. A similar project is underway for the city of Islamabad, Pakistan. Under the direction of Jerome Hajjar and Arif Masud, both at MAE, and funded by the Pakistani government's Earthquake Reconstruction and Recovery Agency and conducted jointly with the National University of Science and Technology in Islamabad, the goal is to estimate losses from earthquakes like the one in October 2005 that killed almost 75,000 people and left 2.8 million homeless in Pakistan. "MAEviz is enabling collaboration between seismologists, social scientists, structural engineers, technical engineers, loss assessors, and emergency managers," says Elnashai. "It really opens things up in a big way." The development of MAEviz is funded by the National Science Foundation, under grant EEC-9701785 through support of the Mid-America Earthquake Center, an NSF Research Center. For further information: http://maeviz.cee.uiuc.edu/ http://mae.ce.uiuc.edu/software_and_tools/maeviz.html