Loading...
HomeMy WebLinkAboutCC AG PKT 2008-07-14 #R• AGENDA STAFF REPORT DATE: July 14, 2008 TO: Honorable Mayor and City Council THRU: David Carmany, City Manager FROM: Lee Whittenberg, Director of Development Services SUBJECT: RECEIVE AND FILE - INFORMATION ON "THE SHAKEOUT EARTHQUAKE SCENARIO" AND THE "CALIFORNIA EARTHQUAKE RUPTURE FORECAST" SUMMARY OF REQUEST: Receive and File Staff Report. Direct Staff to provide to the Planning Commission and Environmental Quality Control Board for information and to return with additional information and any recommended actions as appropriate. • DISCUSSION: Overview of "Shakeout Earthquake Scenario" Preparedness Exercise: On November 13, 2008 City Staff will participate in the state -wide "Golden Guardian `08" earthquake preparedness exercise. The purpose of this state -wide emergency response exercise is to rest the ability of first responders to deal with the impact of a magnitude 7.8 earthquake on the San Andreas Fault in Southern California. The exercise is being jointly organized by the Governor's Office of Emergency Services and the California Office of Homeland Security. "The planned emergency drill is underpinned by the most comprehensive analysis ever of what a major Southern California earthquake would mean on the ground," said Dr. Lucile Jones, chief scientist for U.S. Geological Surrey's Southern California Multi - Hazards Demonstration Project. "We know this science will help state and local agencies develop comprehensive emergency - response plans that will help us avoid the worst impacts of a major quake." • Agenda Item R Z:\07 -14 -08 Council Meeting - Agenda Items\DS - Staff Report - R - Uniform Earthquake Forecast Report.doc\LIM06 -05 -08 Receive and File — Information on `The Shakeout Earthquake Scenario" • and the `Califomia Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 Overview of "The Shakeout Earthquake Scenario": In preparation for the Golden Guardian `OS' preparedness exercise a hypothetical Scenario describing how a magnitude 7.8 earthquake on the San Andreas Fault in Southern California, similar to the recent earthquake in China, would impact the region. In the Scenario, the earthquake would: ❑ kill 1800 persons; ❑ injure 50,000 persons; ❑ cause $200 billion in damage; and ❑ have long- lasting social and economic consequences. The Scenario outlines a hypothetical earthquake in which: ❑ The strongest shaking and greatest damage is near the stretch of the San Andreas Fault that extends through the fastest growing areas of Southern California, including the Coachella Valley, Inland Empire and Antelope Valley. ❑ At least 10 million people will be exposed to heavy shaking. California's efforts • at mitigation have concentrated on life safety and have been largely successful. Thus, in spite of the large numbers of people in highly shaken areas, deaths are estimated at only 1,800. ❑ Building types known to be vulnerable to damage and collapse, do indeed sustain major damage. All un- reinforced masonry buildings within 15 miles of the San Andreas Fault are completely destroyed. Those that are not retrofitted kill many occupants. Many other older building types without retrofitting contribute to over $33 billion in damage to buildings. ❑ The fault offsets all lifelines crossing into Southern California at Cajon Pass (Interstate 15), San Gorgonio Pass (Interstate 10) and along Route 14, including pipelines, power lines, roads, railways, telecommunications and aqueducts. ❑ Strong shaking continues in downtown Los Angeles for 55 seconds - nearly 8 times longer than in the Northridge Earthquake ❑ The prolonged, strong shaking heavily damages and sometimes collapses hundreds of old brick buildings, thousands of older commercial and industrial concrete buildings, many wood -frame buildings, and even a few, high -rise steel buildings. Over 600,000 buildings suffer at least some damage that causes tens of thousands of injuries and hundreds of deaths, and leaves many thousands of people without homes or jobs. • DS - Staff Report - R - Uniform Earthquake Forecast Report 2 • Receive and File — Information on "The Shakeout Earthquake Scenario" and the °Califomia Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 ❑ Fire doubles the fatalities and economic losses. Around Southern there will be 1,600 fires started large enough to warrant fires merge into conflagrations that burn hundreds of city Santa Ana winds, the models still indicate a further $65 and $22 billion in indirect losses fromithe fires. a 911 call, California, and some blocks. Assuming no billion in direct losses ❑ Nearly two thirds of the hospital beds are non - functional in Los Angeles, Orange, Riverside, and San Bernardino counties. At the same time, 50,000 people will seek treatment at emergency rooms. ❑ Thanks to a $6 billion investment in seismic safety, the State highway system fares well. However, although collapse is avoided, some bridges are non- functional so that much of the highway is not passable on the day of the event. The long duration of shaking takes a greater toll on bridges and overpasses under the jurisdiction of cities and counties where the retrofitting processes are not complete or have not begun. ❑ The largest long -term economic disruption comes from damage to the water distribution system. Damage to this system will be so extensive that some areas will have to replace the whole system, and some buildings will be without water for as long as 6 months. The direct and indirect business interruption costs attributed to the lack of water will be $50 billion. ❑ Most of the damage is predictable and much is preventable. Individuals can protect themselves and help their community by: ❑ Storing more water than they already have ❑ Keeping a fire extinguisher and knowing how to use it. ❑ Securing their space. This means securing building contents from flying around and reinforcing a building they own to the most current standards. Provided as Attachment 1 for the information of the City Council and interested citizens is the following documents regarding "The Shakeout Earthquake Scenario ": ❑ "Disaster Earthquake Scenario Unveiled for Southem California" web page, http: / /usgs.gov /newsroom /article asp? ID =1947, downloaded June 4, 2008; ❑ "The Shakeout Earthquake Scenario — A Story That Southem Califomians Are Wilting" web page, http: / /r)ubs.usgs.org /circ /l324, downloaded June 4, 2008; ❑ "The Shakeout Earthquake Scenario — A Story That Southem Califomians Are Writing', U.S. Department of the Interior and U.S. Geological Survey, Circular 1324 and jointly published as California Geological Survey Special Report 207; and Uniform Earthquake Forecast= Status Report 3 Receive and File — Information on °The Shakeout Earthquake Scenario" • and the "Califomia Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 ❑ "The Shakeout Scenario ", U.S. Department of the Interior, U.S. Geological Survey, and California Geological Survey — USGS Open File Report 2008 -1150 and CGS Preliminary Report 25, Version 1.0 — Table of Contents and Chapter 1: Executive Summary. (Complete report available at the Department of Development Services to review) Overview of "Uniform California Earthquake Rupture Forecast": The subject report was recently released by the U.S. Geological Survey, the California Geological Survey, and the Southern California Earthquake Center and describes a new earthquake rupture forecast for California developed by the 2007 Working Group on California Earthquake Probabilities (WGCEP). The study determines that the chances of having one or more magnitude 6.7 or larger earthquakes in California over the next 30 years is greater than 99 %. Such quakes can be deadly, as shown by the 1989 magnitude 6.9 Loma Prieta and the 1994 magnitude 6.7 Northridge earthquakes. The likelihood of at least one or more powerful quake of magnitude 7.5 or greater in the next 30 years is 46% - such a quake is most likely to occur in the southern half of the State. • Provided as Attachment 2 for the information of the City Council and interested citizens are the following documents: ❑ "Forecasting California's Earthquakes — What Can We Expect in the Next 30 Years?", USGS Fact Sheet 2008 -3027; ❑ "The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF)" web page, http : / /r)ubs.usas.gov /of/2007/1437, downloaded April 15, 2008; and ❑ "The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF) ", dated 2008 - Contents, List of Appendices, and Executive Summary. (Complete report available at the Department of Development Services to review) FINANCIAL IMPACT: None. Costs of staff preparation for and participation in the preparedness exercise are ongoing administrative functions of the participating Departments. Uniform Earthquake Forecest.CC Status Report 4 C7 C7 Receive and File — Information on "The Shakeout Earthquake Scenario" and the "California Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 RECOMMENDATION: Receive and File Staff Report. Direct Staff to provide to the Planning Commission and Environmental Quality Control Board for information and to return with additional information and any recommended actions as appropriate. SUBMITTED BY: NOTED AND APPROVED: r xe Whittenberg, erector David Carmany Development Services Depart ent City Manager Attachments: (2) Attachment 1: Documents regarding "The Shakeout Earthquake Scenario ": ❑ "Disaster Earthquake Scenario Unveiled for Southern California" web page, hftp://usgs.gov/newsroom/article.asr)?ID=1-947,, downloaded June 4, 2008; ❑ "The Shakeout Earthquake Scenario — A Story That Southern Californians Are Writing' web page, http://pubs.us9s.org /circ11324, downloaded June 4, 2008; ❑ "The Shakeout Earthquake Scenario — A Story That Southern Californians Are Writing', U.S. Department of the Interior and U.S. Geological Survey, Circular 1324 and jointly published as California Geological Survey Special Report 207; and ❑ "The Shakeout Scenario ", U.S. Department of the Interior, U.S. Geological Survey, and California Geological Survey — USGS Open File Report 2008- 1150 and CGS Preliminary Report 25, Version 1.0 — Table of Contents and Chapter 1: Executive Summary. (Complete report available at the Department of Development Services to review) Uniform Earthquake Forecast.CC Status Report 5 Receive and File — Information on °The Shakeout Earthquake Scenario" • and the "Califomia Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 Attachment 2: Documents regarding the "Uniform California Earthquake Rupture Forecast": ❑ "Forecasting California's Earthquakes — What Can We Expect in the Next 30 Years? ", USGS Fact Sheet 2008 -3027; ❑ "The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF)" web page, http : / /Dubs.usgs.4ov /of/2007/1437, downloaded April 15, 2008; and ❑ "The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF) ", dated 2008 - Contents, List of Appendices, and Executive Summary. (Complete report available at the Department of Development Services to review) 0 Uniform Earthquake Forecast= Status Report 6 • Receive and File — Information on "The Shakeout Earthquake Scenario" and the Tallfomia Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 ATTACHMENT 1 DOCUMENTS REGARDING THE SHAKEOUT EARTHQUAKE SCENARIO": ❑ "DISASTER EARTHQUAKE SCENARIO UNVEILED FOR SOUTHERN CALIFORNIA" WEB PAGE, HTTP: // USGS .GOWNEWSROOM /ARTICLE.ASP ?ID =19 47, DOWNLOADED JUNE 4, 2008; ❑ "THE SHAKEOUT EARTHQUAKE SCENARIO - A STORY THAT SOUTHERN CALIFORNIANS ARE • WRITING" WEB PAGE, HTTP: / /PUBS.USGS.ORG /CIRC /1324, DOWNLOADED JUNE 4, 2008; n U ❑ "THE SHAKEOUT EARTHQUAKE SCENARIO - A STORY THAT SOUTHERN CALIFORNIANS ARE WRITING", U.S. DEPARTMENT OF THE INTERIOR AND U.S. GEOLOGICAL SURVEY, CIRCULAR 1324 AND JOINTLY PUBLISHED AS CALIFORNIA GEOLOGICAL SURVEY SPECIAL REPORT 207; AND ❑ "THE SHAKEOUT SCENARIO ", USGS OPEN FILE REPORT 2008 -1150 AND CGS PRELIMINARY REPORT 25, VERSION 1.0, U.S. DEPARTMENT OF THE INTERIOR, U.S. GEOLOGICAL SURVEY, AND CALIFORNIA GEOLOGICAL SURVEY - TABLE OF CONTENTS AND CHAPTER 1: EXECUTIVE SUMMARY. (COMPLETE REPORT AVAILABLE AT THE DEPARTMENT OF DEVELOPMENT SERVICES TO REVIEW) Uniform Earthquake Forecast.CC Status Report 7 USGS Release: Disaster Earthquake Scenario Unveiled for Southern California (5/22/2008 10:00:00 AM) aQUSGS science for a Changing warlJ - • .. Disaster Earthquake Scenario Unveiled for Southern California Released: 5/22/2008 10:00:00 AM Contact Information: U.S. Department of the Interior, U.S. Geological Survey Office of Communication 119 National Center Reston, VA 20192 Clarice Nassif Ransom USGS Phone: 703-648-4299 Jim Nickles. USGS Phone: 916 - 278 -3016 Don Drysdale, CGS Phone: 916-323-1886 Pagel of 3 Editors' notes: A telephone conference call is scheduled for 10:30 a.m. PDT for a brief presentation and opportunity for questions with the principal authors. Call 605- 990 -0100, and enter conference code 1009678# B -roll animations and high - resolution images are available at hU: / /urbanearth.usgs.gov /shakeout. Scientists today unveiled a hypothetical Scenario describing how a magnitude 7.8 Southern California earthquake - similar to the recent earthquake in China- would impact the region, causing loss of lives and massive damage to infrastructure, including critical transportation, power, and water *MS. In the Scenario, the earthquake would kill 1800 people, injure 50,000, cause $200 billion in damage, and have long - lasting social and economic consequences. This is the most comprehensive analysis ever of what a major Southern California earthquake would mean, and is the scientific framework for what will be the largest earthquake preparedness drill in California history, scheduled for November 13, 2008. The November preparedness exercise, "Golden Guardian '08," will test the ability of emergency responders to deal with the impact of a magnitude 7.8 earthquake on the San Andreas Fault in Southem California, and is being jointly organized by the Governor's Office of Emergency Services and the California Office of Homeland Security. The Golden Guardian exercise will occur during a week -long series of public events planned for the "Great Southern California ShakeOut." A June 41' kick -off event is planned for the "Shakeout" to help communities plan to respond to the risks highlighted in the Scenario. The scientific report describing the ShakeOut Scenario, jointly published by the U.S. Geological Survey (USGS) and the Califomia Geological Survey (CGS), will be released today during a Congressional hearing in Washington, D.C. The House Committee on Natural Resources, Subcommittee on Energy and Mineral Resources, led by Chairman Jim Costa (D -CA), will hold an oversight hearing on USGS efforts to prepare for future earthquakes, at 10:00 a.m. EDT in Room 1324, Longworth House Office Building Although imaginary, the Shakeout Scenario is based on scientists' best predictions of what would actually occur during and after a major earthquake on the San Andreas Fault. The Scenario outlines a hypothetical earthquake in which: • The strongest shaking and greatest damage is near the stretch of the San Andreas Fault that extends through the fastest growing areas of Southern California, including the Coachella Valley, Inland Empire and Antelope Valley. • At least 10 million people will be exposed to heavy shaking. California's efforts at mitigation have concentrated on life safety and have •been largely successful. Thus, in spite of the large numbers of people in highly shaken areas, deaths are estimated at only 1,800. Building types known to be vulnerable to damage and collapse, do indeed sustain major damage. All un- reinforced masonry buildings within 15 miles of the San Andreas Fault are completely destroyed. Those that are not retrofitted kill many occupants. Many other older building types without retrofitting contribute to over $33 billion in damage to buildings. • The fault offsets all lifelines crossing into Southern California at Cajon Pass (Interstate 15), San Gorgonio Pass (Interstate 10) and along Route 14, including pipelines, power lines, roads, railways, telecommunications and aqueducts. • Strong shaking continues in downtown Los Angeles for 55 seconds - nearly 8 times longer than in the Northridge Earthquake h4:// www. usgs.gov /newsroom/article.asp ?ID=1947 6/4/2001 USGS Release: Disaster Earthquake Scenario Unveiled for Southern California (5/22/2008 10:00:00 AM) Page 2 of 3 • The prolonged, strong shaking heavily damages and sometimes collapses hundreds of old brick buildings, thousands of older commercial and industrial concrete buildings, many wood -frame buildings, and even a few, high -rise steel buildings. Over 600,000 buildings suffer at least some damage that causes tens of thousands of injuries and hundreds of deaths, and leaves many thousands of people without homes of jobs. • Fire doubles the fatalities and economic losses. Around Southern California, there will be 1,600 fires started large enough to warrant call, and some fires merge into conflagrations that bum hundreds of city blocks. Assuming no Santa Ana winds, the models still indic further $65 billion in direct losses and $22 billion in indirect losses from the fires. • Nearly two thirds of the hospital beds are non - functional in Los Angeles, Orange, Riverside, and San Bernardino counties. At the same time, 50,000 people will seek treatment at emergency rooms. • Thanks to a $6 billion investment in seismic safety, the State highway system fares well. However, although collapse is avoided, some bridges are non - functional so that much of the highway is not passable on the day of the event. The long duration of shaking takes a greater toll on bridges and overpasses under the jurisdiction of cities and counties where the retrofitting processes are not complete or have not begun. • The largest long -term economic disruption comes from damage to the water distribution system. Damage to this system will be so extensive that some areas will have to replace the whole system, and some buildings will be without water for as long as 6 months. The direct and indirect business interruption costs attributed to the lack of water will be $50 billion. • Most of the damage is predictable and much is preventable. Individuals can protect themselves and help their community by: - Storing more water than they already have - Keeping a fire extinguisher and knowing how to use it. - Securing their space. This means securing building contents from flying around and reinforcing a building they own to the most current standards. "The planned emergency drill is underpinned by the most comprehensive analysis ever of what a major Southern California earthquake would mean on the ground," said Dr. Lucile Jones, chief scientist for U.S. Geological Survey's Southern California Multi- Hazards Demonstration Project. "We know this science will help state and local agencies develop comprehensive emergency- response plans that will help us avoid the worst impacts of a major quake." The ShakeOut Scenario is the product of an interdisciplinary collaboration of over 300 scientists, engineers, and other experts from several agencies, including the USGS, the California Geological Survey, Southern California Earthquake Center, California Office of Emergency Se* and Seismic Safety Commission. To create the Scenario, geologists determined the amount of potential motion on the part of the San Andreas Fault with the greatest risk of imminent rupture, a 200 -mile long section from the Salton Sea in the Coachella Valley to just south of Gorman. From this, seismologists and computer scientists modeled the ground shaking. Engineers and building professionals used the models of ground shaking to estimate damage to the built environment. And from these damages, social scientists evaluated emergency response, casualties, and the impact on our economy and society. The following scientists and engineers led the development of individual sections of the ShakeOut Scenario. They were responsible for bringing together the appropriate team of experts to analyze that aspect of the earthquake, leading the investigations and ensuring that the final document was written. Each of them can speak to the goals of the Scenario, the main results and the details of their expertise. They can be reached by contacting the USGS. Dr. Lucy Jones, Seismologist, USGS, Chief Scientist Dr. Kenneth Hudnut, Geologist, USGS, Geologic setting and ground motion prediction Dr. Keith Porter. Engineer, University of Colorado, Physical Damages Dr. Daniel Pond Geologist, USGS, Secondary Hazards Ms. Hope Seligson, Engineer, MMI Engineering, HAZUS and loss estimation Dr. Kimberley Shoaf, Public Health Scientist, UCLA School of Public Health, Mortality and morbidity • Dr. Michael Reichle, Chief Seismologist (ret.), California Geological Survey, Lifelines Dr. Dennis S. Mileti, Sociologist, California Seismic Safety Commission, Emergency Response Dr. James oltz Social Psychologist, California Office of Emergency Services, Emergency Response Dr. Richard Bemknopf, Economist, USGS, h":// www. usgs .gov /newsroom/article.aSD ?ID =1947 6i4i1)nnR USGS Release: Disaster Earthquake Scenario Unveiled for Southern California (5/22/2008 10:00:00 A11O Page 3 of 3 Economics Dr. Anne Wein, Decision Scientist, USGS, Economics *Dale A. Cox, Project Manager, USGS A copy of the full technical report, The Shakeout Scenario, is available online at http:// uo bs.usgs.gov /of/2008 /1150. A non - technical summary narrative of the Scenario is online at ho: // ubs.usgs.gov /circ/1324/. Paper copies of the narrative are available by request. High resolution images, and a computer animation showing the earthquake rupture and its waves of energy spreading across Southern California are online at http• / /urbaneart usgs gov /shakeout. USGS provides science for a changing world. For more information, visit www.usgs.gov. Subscribe to USGS News Releases via our electronic mailing list or RSS feed. * * ** www.usgs.gov * * ** Links and contacts within this release are valid at the time of publication. Department of the Interior I U.S. Geological Survey URL: http: / /www.usgs.gov /newsroom /article.asp ?ID =1947 Page Contact Information: Ask USGS Page Last Modified: 5/27/2008 9:44 :26 AM • http: / /www.usgs .gov /newsroom/article.asp ?ID =1947 6/4/200f 'he ShakeOut Earthquake Scenario - -A Story That Southern Californians Are Writing Z'91E thquake Hazards Program 3'repared in cooperation with the :allfornia Geological Survey J.S. Geological Survey Circular 1324 :alifornia Geological Survey Special Report 207 rersion 1.0 Pagel of 4 USGS Home Contact USGS Search USGS The ShakeOut Earthquake Scenario —A Story That Southern Californians Are Writing 3y Suzanne Perry, Dale Cox, Lucile )ones, Richard Bernknopf, James Goltz, Kenneth Hudnut, Dennis Mileti, Daniel Pont!, Keith Porter, Michael Reichle, Hope Seligson, Kimberley Shoaf, Jerry Treiman, and Anne Wein 68 0 'lttp: / /pubs.usgs.gov /circ /1324/ 6/4/2008 The ShakeOut Earthquake Scenario - -A Story That Southern Californians Are Writing Page 2 of 4 • Ground motion for the ShakeOut Scenario earthquake 60 seconds after the southern San Andreas Fault first begins rupturing. Yellow shows the highest amplitudes of ground motion. (Simulation by Rob Graves of URS Corporation for the Southern California Earthquake Center on high- performance computers at the University of Southern California; image courtesy of Geoff Ely, University of California San Diego /San Diego Supercomputer Center.) Introduction The question is not if but when southern California will be hit by a major earthquake —one so damaging that it will permanently change lives and livelihoods in the region. How severe the changes will be depends on the actions that individuals, schools, businesses, organizations, communities, and governments take to get ready. To help prepare for this event, scientis of the U.S. Geological Survey (USGS) have changed the way that earthquakW scenarios are done, uniting a multidisciplinary team that spans an unprecedented number of specialties. The team includes the California Geological Survey, Southern California Earthquake Center, and nearly 200 sttp: / /pubs.usgs.gov /circ /1324/ f/4/')nnR Che ShakeOut Earthquake Scenario - -A Story That Southern Californians Are Writing Page 3 of 4 3ther partners in government, academia, emergency response, and industry, working to understand the long -term impacts of an enormous earthquake on the complicated social and economic interactions that sustain southern ifornia society. This project, the ShakeOut Scenario, has applied the best rrent scientific understanding to identify what can be done now to avoid an Earthquake catastrophe. More information on the science behind this project will be available in The Shakeout Scenario (USGS Open -File Report 2008- 1150; http: / /pubs.usgs.ciov /of /2008/1150/). The "what if ?" earthquake modeled in the ShakeOut Scenario is a magnitude 7.8 on the southern San Andreas Fault. Geologists selected the details of this hypothetical earthquake by considering the amount of stored strain on that part of the fault with the greatest risk of imminent rupture. From this, seismologists and computer scientists modeled the ground shaking that would occur in this earthquake. Engineers and other professionals used the shaking to produce a realistic picture of this earthquake's damage to buildings, roads, pipelines, and other infrastructure. From these damages, social scientists projected casualties, emergency response, and the impact of the, scenario earthquake on southern California &rwquo;s economy and society. The earthquake, its damages, and resulting losses are one realistic outcome, deliberately not a worst -case scenario, rather one worth preparing F and mitigating against. Decades of improving the life- safety requirements in building codes have greatly reduced the risk of death in earthquakes, yet southern California's aconomic and social systems are still vulnerable to large -scale disruptions. Because of this, the ShakeOut Scenario earthquake would dramatically alter the nature of the southern California community. Fortunately, steps can be taken now that can change that outcome and repay any costs many times aver. The Shakeout Scenario is the first public product of the USGS Multi - Hazards Demonstration Project, created to show how hazards science can increase a community's resiliency to natural disasters through improved planning, mitigation, and response. Download this report as a 24 -page PDF file (c1324.pdf; 15.7 MB). For questions about the content of this report, contact Suzanne Perry k1so of Interest : wa avucil -rime rceoort luuS -1150 /CGS Preliminary Report 25 The Shakeout scenario by Lucile M. Jones, Richard Bernknopf, Dale Cox, James Goltz, (enneth Hudnut' Dennis Mileti, Suzanne Perry, Daniel Pont!, Keith Porter, Hlchael Relchle, Hope Seligson, Kimberley Shoaf, Jerry Treiman, and Anne .ttP: / /Pubs.usgs.gov /circ /i 324/ 6/4/2008 The ShakeOut Earthquake Scenario - -A Story That Southern Californians Are Writing Wein Page 4 of 4 Suggested citation and version history/ 0 Download a free copy of the current version of Adobe Reader. Help I PDF help I Publications main page I Western Circulars Geology I Earthquake Hazards Program I This report is also available in print from: USGS Information Services, Box 25286, Federal Center, Denver, CO 80225 telephone: 888 ASK -USGS; e-mail: infoservices @usgs.gov and the California Geological Survey. Accessibility FOIA Privacy Policies and Notices • U.S. Department of the Interior I U.S. Geological Survey # � U RL: htt u bs. us s. ov ci rc 13 24 -� - — P //P 9 9 / / / Page Contact Information: Michael Diggles R«�" 0 ittn: / /nubs.usgs.sov /circ /1324/ AiaiI)nnQ t M1. ti "` • • The ShakeOut Earthquake Scenario —A Story That Southern Californians Are Writing By Suzanne Perry, Dale Cox, Lucile Jones, Richard Bernknopf, James Goltz, Kenneth Hudnut, Dennis Mileti, Daniel Ponti, Keith Porter, Michael Reichle, Hope Seligson, Kimberley Shoaf, Jerry Treiman, and Anne Wain Circular 1324 Jointly published as California Geological Survey Special Report 207 • U.S. Department of the Interior U.S. Geological Survey U.S. Department of the Interior DIRK KEMPTHORNE, Secretary U.S. Geological Survey Mark D. Myers, Director State of California ARNOLD SCHWARZENEGGER, Governor The Resources Agency MIKE CHRISMAN, Secretary for Resources Department of Conservation Bridgett Luther, Director California Geological Survey John G. Parrish, Ph.D., State Geologist U.S. Geological , Survey, Reston Virginia: 2008 • Y 9 This report and any updates to it are available online at httP://pubs.usgs.gov /circ /1324/ For product and ordering information: World Wide Web: http: / /www.usgs.gov /pubprDd/, Telephone: 1-886- ASK -USGS For more information on the USGS —the Federal source for science about the Earth, its natural and living resources, natural hazards, and the environment World Wide Web: http:/ /www.usgs.gov /, Telephone: 1 -888-ASK-USGS (1- 888 - 275 -8747) Any use of trade, product or fine names is for descriptive purposes only and does not imply endorsement by the U.S. Government Although this report is in the public domain, permission must be secured from the individual copyright owners to reproduce any copyrighted material contained within this report. Cataloging -in- Publication data are on file with the Library of Congress (URL httpV/www.loc.govA. Produced in the Western Region, Menlo Park, California Manuscript approved for publication, April 21, 2008 Text edited by James W. Handley II and Peter Stauffer Layout and design by Judy Weathers FRONT COVER- -Ground motion for the Shakeout Scenario earthquake 60 seconds after the southern San Andreas Fault first begins rupturing. Yellow shows the highest amplitudes of ground motion. (Simulation by Rob Graves of URS Corporation for the Southern California Earthquake Center on high - performance computers at the University of Southern California; image courtesy of Geoff Ely, University of California San Diego /San Diego Supercomputer Center.) BACK COVER —Within the ShakeOut Scenario study area are many neighborhoods like this one, at severe risk from fire following earthquake. Here, tightly • packed wood buildings will enable small fires to spread and merge into conflagrations that can burn dozens of blocks. Fire following earthquake can have devastating consequences, as tragically seen after the 1906 San Francisco, 1923 Tokyo, and 1995 Kobe earthquakes. It is a significant threat in urban areas of California and doubles the fatalities and economic losses in the hypothetical ShakeOut earthquake. (Google Earth image.) fffl Contents Introduction....................................................................................................................... ..............................1 The Shakeout Scenario Is Not a Prediction ............................................................... ..............................1 The ShakeOut Scenario Narrative ................................................................................ ..............................5 HowDo We Write the Ending? ..................................................................................... .............................15 AdditionalInformation .................................................................................................... .............................16 ShakeOut Scenario Coordinators: Lucile Jones, USGS: Chief Scientist Dale Cox, USGS: Project Manager Suzanne Perry, USGS: Staff Scientist/Writer Richard Bemknopf, USGS: Economics James Goltz, Governors Office of Emergency Services: Emergency Response • Kenneth Hudnut, USGS: Earthquake Design Dennis Mileti, California Seismic Safety Commission: Emergency Response Daniel Ponti, USGS: Secondary Hazards Keith Porter, University of Colorado: Physical Damages Michael Reichle, California Geological Survey: Lifelines Special Studies Hope Seligson, MMI Engineering: HAZUS Loss Estimations Kimberley Shoaf, University of California Los Angeles: Health, Safety, Social, Cultural, and Institutional Jerry Treiman, California Geological Survey: Lifelines Special Studies Anne Wain, USGS: Economics • • • The ShakeOut Earthquake Scenario —A Story That Southern Californians Are Writing BySuzanne Perry, Dale Cox, Lucile Jones, Richard Bernknopf, James Goltz, Kenneth Hudnut, Dennis Mileti, Daniel Ponti, Keith Porter, Michael Reichle, Hope Seligson, Kimberley Shoaf, Jerry Treiman, and Anne Wein Introduction The question is not if but when southern California will be hit by a major earthquake — one so damaging that it will permanently change lives and livelihoods in the region. How severe the changes will be depends on the actions that individuals, schools, businesses, organizations, communities, and governments take to get ready. To help prepare for this event, scientists of the U.S. Geological Survey (USGS) have changed the way that earthquake scenarios are done, uniting a multidisciplinary team that spans an unprec- edented number of specialties. The team includes the California Geological Survey, Southern Cali- fornia Earthquake Center, and nearly 200 other partners in government, academia, emergency response, and industry, working to understand the long -term impacts of an enormous earthquake on the complicated social and economic interactions that sustain southern California society. This proj- ect, the ShakeOut Scenario, has applied the best current scientific understanding to identify what can be done now to avoid an earthquake catastro- phe. More information on the science behind this project will be available in The ShakeOut Sce- nario (USGS Open -File Report 2008 -1150; http: / /pubs.usgs.gov /of /2008/1150/). The "what if?" earthquake modeled in the ShakeOut Scenario is a magnitude 7.8 on the southern San Andreas Fault. Geologists selected the details of this hypothetical earthquake by considering the amount of stored strain on that part of the fault with the greatest risk of immi- nent rupture. From this, seismologists and com- puter scientists modeled the ground shaking that would occur in this earthquake. Engineers and other professionals used the shaking to produce a realistic picture of this earthquake's damage to buildings, roads, pipelines, and other infra- structure. From these damages, social scientists projected casualties, emergency response, and the impact of the scenario earthquake on southern California's economy and society. The earth- quake, its damages, and resulting losses are one realistic outcome, deliberately not a worst -case scenario, rather one worth preparing for and miti- gating against. Decades of improving the life -safety require- ments in building codes have greatly reduced the risk of death in earthquakes, yet southern Cali- fornia's economic and social systems are still vulnerable to large -scale disruptions. Because of this, the ShakeOut Scenario earthquake would dramatically alter the nature of the southern Cali- fornia community. Fortunately, steps can be taken now that can change that outcome and repay any costs many times over. The ShakeOut Scenario is the first public product of the USGS Multi- Hazards Demonstration Project, created to show how hazards science can increase a community's resiliency to natural disasters through improved planning, mitigation, and response. The ShakeOut Scenario Is Not a Prediction On November 13, 2008, the ShakeOut Sce- nario earthquake and projected damages will be used as a basis for public drills and emergency 2 The ShakeOut Earthquake Scenario —A Story That Southern Californians Are Writing response exercises, and so this hypothetical earth- quake has been arbitrarily assigned a date, a time, a plausible collection of aftershocks; and even local weather conditions. Because it is a scenario for planning, it provides detailed numbers, includ- ing casualties, collapsed buildings, and business losses. Yet, the Shakeout earthquake is not a prediction. Certainly, a large earthquake involv- ing this part of the San Andreas Fault is highly probable. In fact, scientists have determined that this is the most likely source of a large earthquake in all of California (see Forecasting California's Earthquakes —What Can We Expect in the Next 30 Years ?, U.S. Geological Survey Fact Sheet 2008 -3027; http : / /pubs - usgs.gov /fs/2008/3027/). When the next San Andreas Fault earthquake does happen, some things are inevitable —the fault rupture will break any road, track, or pipe ..L �i s ` that crosses it, and intense shaking will damage or destroy buildings that weren't constructed to withstand it. The next earthquake will be differ- ent in details from the ShakeOut earthquake, and its total damages and losses will differ, because each earthquake produces its own patterns of shaking and damage. However, the widespread, regional effects will be similar, and so will the long -term social and economic impacts. Get- ting prepared for the ShakeOut earthquake will help southern Californians withstand other earthquakes of comparable size. If we take no additional actions for preparedness and mitiga- tion, and the ShakeOut earthquake does occur, it will cause some 2,000 deaths, 50,000 injuries, $200 billion in damage, and severe, long- lasting disruption. These numbers can climb with each damaging aftershock. „ ., `,�n..�.± y �y ^''.- "iy , ;}, -, . >: �Yr�a�t., v t,:� M:,�. .f•• \• The Cajon Pass is one of five major' lifelinezarridors" that ate the veins and arteries. rough- hich', . etaaa it life circulates in southern California: These iifeline camdors.allo .cdmmuters`, "gods; telephone and Internet lines, electricity, water, ga.as,•nd fuelsro,rnove through tfie mountains that- ';• r, y2" surround southem'Cali#aniia. Because mast of the corridors -must crassthe,southern San Andreas: ; ,. -, Fault, future earthquakes are certain to severthe fifelines m one or more of these corridors and cause sign ca' disruption bj� inferripting the•movemeni'of goods from the ports of tos.Angeles:- "- and Long Beach.lUSGS photo 'bytucile Janes ►.' : "Z7 . .. , 1 e .v s.. v3 .. • n .. 6 _ r � , r'%:' r> .�T f r`'Yr i , s .. 'L` ' , \• �. 'l ` +x %.. "ttIFELINES-CROSS;T.HE-FAULT x'' ..L �i s ` that crosses it, and intense shaking will damage or destroy buildings that weren't constructed to withstand it. The next earthquake will be differ- ent in details from the ShakeOut earthquake, and its total damages and losses will differ, because each earthquake produces its own patterns of shaking and damage. However, the widespread, regional effects will be similar, and so will the long -term social and economic impacts. Get- ting prepared for the ShakeOut earthquake will help southern Californians withstand other earthquakes of comparable size. If we take no additional actions for preparedness and mitiga- tion, and the ShakeOut earthquake does occur, it will cause some 2,000 deaths, 50,000 injuries, $200 billion in damage, and severe, long- lasting disruption. These numbers can climb with each damaging aftershock. „ ., `,�n..�.± y �y ^''.- "iy , ;}, -, . >: �Yr�a�t., v t,:� M:,�. .f•• \• The Cajon Pass is one of five major' lifelinezarridors" that ate the veins and arteries. rough- hich', . etaaa it life circulates in southern California: These iifeline camdors.allo .cdmmuters`, "gods; telephone and Internet lines, electricity, water, ga.as,•nd fuelsro,rnove through tfie mountains that- ';• r, y2" surround southem'Cali#aniia. Because mast of the corridors -must crassthe,southern San Andreas: ; ,. -, Fault, future earthquakes are certain to severthe fifelines m one or more of these corridors and cause sign ca' disruption bj� inferripting the•movemeni'of goods from the ports of tos.Angeles:- "- and Long Beach.lUSGS photo 'bytucile Janes ►.' : "Z7 . .. , 1 e .v s.. v3 .. • n .. 6 _ r � , r'%:' r> .�T f r`'Yr i , s .. 'L` ' , \• �. 'l ` +x THE EARTHQUAKE ADVANCES 0 :30 URS /USC o.os f_UU URS /USC o.os 1 2m& 2 :00 URS/USC oos " 2mlb These computer - generated snapshots show ground motions for the Shakeout Scenario earthquake, in meters per second (m /s;1 fi%s is about 3.3 feet per second►, Yellow • indicates higher amplitudes of ground motion. The snapshots show three points in tlm"O (left), 60 (center), and 120 (right) seconds after the sbutherh San Andreas Fault (dashed white line) first begins rupturing at Bombay Beach, on the eastern shore of the Salton Sea. Note that soma areas remaKorangd ltolo fed for mdch longerthail others, indicating extended, intense shaking in some of the sedimentary basins of southern California. (Simulation by Rob Graves of URS Corporation for the Southern' California Earthquake Center on high - performance computers at the University of Southern California; images courtesy of Gobff Ely; University of Califorhia San Diego /Sah biego Superopinputer Canter.) A V A y O a C m m N n m 1 y a m CA O O 3 c� m 8 0 w • • The ShakeOut Scenario Narrative 5 The ShakeOutScenario Narrative Much like a. movie script; the following fictional narrative will guide those participating in ; the Shakeout Earthquake Scenario public drills and emergency response exercises on November, 13, 2008. More effectively than any statistics, this narrative describes what this magnitude 7:8 earthquake would be like in southern California if i10 additional actions are taken for mitigation or preparedness. Noveiniber' ' 13,:,2008 Thiirsda .9:50 a:rri� .. •1D minutes bef r•" �%�� ,4 begins..- o •e a -quake J - By°mid4io nin ybn this'workda` 200 000 commuters have made th i ,> n9 y. a r- way•from� kern; Riverside, and San Bernardino Counties into the Los Angeles area These driv- ; • , : _, ers :trade it lengthy commute for the lower`cost'of housing in fast- growing communi- ties like Victorville•und Lancaster, on thefar side of the Sari Andreas'Fault. Others cross the fault in the opposite direction, to employers in high desert com munities: The commuters have joined 7.b mi Ilion, other = southera•Califor'nians,in workplaces constructed of steel, concrete, brick, or wood. Of the marry millions of <= i homes and workplaces, only a fraction are covered by' earthquake insurance. The life- safety provisions of California's building codes have been improved over the years, and the many fairly new homes in the Coachella Valley suf- fer only minor damage., Yet every item inside these, homes, if not secured, is heading to the floor. Shat-' tered TVs and other home electronics create treach- erous carpets of glass and cords. Many older buildings suffer structural damage: Many older concrete buildings quickly collapse,, trapping, occupants. The rupture front continues its advance to San Gorgonio Pass and dismantles the ten miles of Inter- - state 10 freeway that straddles the San Andreas Fault. The eastern part of River- side County is now cut off from the iuestern part-, „ Thursday 90:01:00 a.m. 'Y . Most people in Los Angeles and Ventura Counties are not yet aware of what is hap- pening as the earthquake pounds the Coachella Valley and heads their way. By now • • • E Tha ShakeOut Scenario Narraft's the first waves have crashed through the Cajon Pass, severing the 115 freeway, bending rail lines, and derailing a train. Roads, prev'iously through_qoing across the fault, now end abruptly and pick up again 15 feet to fhe right.The strong shaking also sends landslides across the rails and roads. ' Pipelinis snap and electrical trans- mission lines fail. Spraying fuel ignites, causing an explosion. Strong shaking begins to reverberate in the sediment -" fi [led basins of the Inland Empire. Old warehouse , - - .,, -, '- districts and historic downtowns are crumbling, and' marry of their old, unretrofitted buildings have trapped or killed the people inside. Many older concrete bulld- ings ' have collapsed, and many older miao'dframe.buildirgs have'shifted off their foundations, breaking qas.ino • water lines'In the p I roCAM .. The Coachella Valley is still sha"k'i n Thursday 100 T.30 e, mi: Over geologic +i-O the "I^+;^ 4C 11W , ri %. plates has pushed the mountains of southern California up,•whili PENN I fire, rain; and rivers have.brouglit the mountains down, piece by Piece, filling 6asiris,with'sediments jand'ereating low, flat areas. Like many cities, Los Angeles was built atop sedimenti. '56md'of the seismic waves now reach these sediments and find easy territory in which to move back and forth, shaking vigorously long after the waves fade elsewhere. Strong shaking aking will continue in Los Angeles for 55 seconds, to the shock of residents who remember the strong shaking during the 1994 Northridge earthquake, which lasted only 7 seconds. T I The ShakeOutEarthquake Scenario --A Story That Southero'Californians Are Writing Thursday 10.02 a.m. (...2 minutes after the quake began...} At last, the fault has stopped rupturing, bu +seismic;waves continue to advance into Bakersfield, Oxnard, and Santa Barbara —here the shaking is just beginning. Across southern California, the power is out. Emergency generators that have: D een secured against earthquake shaking are still functional and now kick on.•The shaking', has finati stop ed in the Caach tln VI-11 b t 44. e ' ft h e ems- u a ors ocks are Just beginning. Throughoutsoutherri Californialn the next feii•" > months there -will be tens of thousands of ear #h quake aftershocks large enough to feel. There will be dozens lorge,enough to cause additional damage ,and•to imperil victims aridrescuers. Some of the .' aftershock damage will be to people's psyches. Sig, earthquakes are traumatic, and each new bout of shaking increases stress, "especially! in" children who are cut off from their families. in the areas of strong shaking, many mobile homes ?' `have collapsed off supports, snapping water, gas and sewer lines, and blocking rescue routes. Mobile homes installed snugly in shallow pits, or braced for earth -, quakes, are still intact. .The State highway system has fared well. A $6 bil- . " lion investment in seismic retrofitting has paid off, , and the only highway deaths have been in crashes caused by intense earthquake' shaking. However, the long duration of shaking has taken its toll on bridges and overpasses within local jurisdictions, where the retrofitting process is not com- pleted, or not yet begun. No hospitals have seen complete collapses, but many hospital buildings are nonfunc tional. Some hospital structures survived the shaking but must close due to nonstruc- tural damage such as water pipes that break and f lood. Thursday 10:05 a.m. (..: b minutes after the quake began,...) The'U.S. Geological 5urvey'posts preliminary informa- , Lion about the earthquake. Learning that the magni- . tude is 7.8, the world turns its attention to southern California, Locally, news helicopters take to the.alr to begin spot coverage of the devastation. With power out, residents turn to their radios or talk to those they meet in the streets, searching for any'information. Across the region, phone systems, including cellular and 911, are unusable, overwhelmed by the vast number of attempted calls, Thursday 10.30 a.m. (... 30 minutes after the quake began...) Emergency operations centers are activating, and police, fire, and medical personnel shift into • • The ShakeOut Scenario Narrative emergency response mode, focusing on localized incidents with any means available. They react quickly, according to their training and earthquake plans established in advance. All over the region, a foreseeable tragedy unfolds. Buildings that engineers knew were going to perform badly, have performed badly. These are older buildings; constructed with little earthquake resistance. The experts have names for them — non - ductile rein - forced concrete, tilt-up concrete, unreinforced masonry, soft-stories—and hundreds of these buildings have now - followed their reputations into the dust. Thousands—of, other •sftct4e` s are still.standing,but so gravely dam- aged that they can never be, used again. ' While the earth still.shakes in plod s'Ia17` rom" the earthquake's origin, people'in' the earliest;hit,areas are beginning to •confront damaged buildings and to = help those'who are trapped or hurt.:Lacking gloves, crowbars, land training, some people claw through debris with, bare hands. 'U'Ititnately, 95 percent of those who are;•reicued will be.rescued by other victims, as has •been seen in earthquake disasters worldwide. Air traffic is being diverted from southern California. s 9 ' As people start to assess their situations, millions of them discover they are' cut off from their families, with no way to learn the fate of their loved ones or homes. This realization also hits first responders as they move out to help; they understand that the disaster may seem to be over but is just beginning. Fires are starting in countless ways. Power lines arc -, gas appliance lines snap... chemicals spi Il and mix.- a lamp hits a sofa, unnoticed with the power out and the earth shaking, then the power returns and the sofa starts to smolder... Most of the fires start small, but not all are discovered right away. In any case, the phones don't have dial tones. Even if they did, in a disaster this big and wide- spread, there are not enough emergency personnel, to immediately respond to every call for help. Worse, response is slowed by roads that are impassable due, to damage, building debris, or abandoned cars. Worse still, in many places the. , water system is damaged, leaving inadequate water pressure for fire fighting. Once started, a small fire needs only minutes to engulf a home or workplace. Around southern Califor -. nia on this day, there will be 1,600 fires Iarge'enough to warrant a 911 call. The stronger the shaking „the greater, the number`of fires ignited. In areas With, densely packed', wootdfrome buildings; some of• these ignitions wild combine, spreading into conflagrations` that burn dozens of blocks, As soon as the shaking stops, ex'per'ts race to, inspect dams around the •region: A few are found to be' leaking at the toe- -a sign of potential fail =” ure. Emergency responders are'spread even.,• > 10 The Shakeout Earthquake .Scenario— A.Sio y That Sou&6rn- Californialns Are Writing thinner when they must begin the evacuation of downstream areas. No dam failures will occur in this particular earthquake. Thursday 10:33 a.m. (... 33 minutes after the quake began...) A magnitude 7.0 aftershock begins near the Salton 'Sea:and ruptures to the south. Luckily, this is a relatively unpopulated area. Shak- ; ing and its effects are felt throughout Imperial and San Diego Counties, as well as in.Mexicah, Mexico. , , :.Damage to a darn'in San Diego`County requires ari , evacuation. Teaing of firefighters'from Sdn Diego - . County �had;been.getting ready 'to'•come north to,help with the initiaFearthquake's aftermath, but`are now diverted to'respond to the strong aftershock affect ing their;own county. Thursday :12.00 noon (.. ,7 hours after the quake began..) ; Smaller fires are'merging into larger' fires in parts of the •region where shaking was high and wood buildings. are in close proximity. -World and national news cover- age is focused on urban Los Angeles, especially on a few collapsed buildings. This media focus makes the' damage seem even worse than it is, and also more localized' . It will be several days before a clear picture emerges of damage around the region. Fire departments in Arizona and the San Francisco Bay area start to mobilize, but mutual aid is hindered because so many roads into the.affected region are impass- able. By now, some hospitals are beginning to receive and treat the injured, but with routes and communications disrupted, ambulances struggle to reach victims' and get them to hospitals. November 14, 2008 Friday 03 :17 a.m. (... 17f hours after the quake began A magnitude 7.2 aftershock begins near San Bernardino and ruptures west along the base of the San Gabriel Mountains. This ' earthquake is considerably larger than 1994's . magnitude 6.7 Northridge earthquake, which killed 33 people and cost more than $40 billion. The rupture" stops 18 miles east of Pasadena, near Monrovia: The location and size of this earthquake are devastating to the already - weakened infrastructure and . . overextended emergency response resources. The aftershock triggers damaging aftershocks of its own. Friday 0.7.00 a.m. (.__ 21 "hours after the quake began..) By now, a Presidential Disaster Declaration has been issued, and Federal resources have been committed. - The Federal Emergency Management Agency (FEMA), the California Governor's Office of Emergency Services • • The'ShakeOut Scenerio'Narrative 77 (OES), and Operational Areas for emergency manage6 4*have setup a Joint Incident- : Command Center. Communications remain difficult-Ham radio operators begin to assist official responders, Ys Since first hearing about-the earthquake, people outside southern California have been trying to reach family and friends here. Very few have succeeded. Friday e9 :02 a.m. (.::23+ hours after ithe.quake began...:. / = A magnitude 5.6 aftershock rgtiiles residential areas Jn ' Ran& Cu6amonga.- his is a bit bigger than the 1990' -Upland earthquake that caused more than $10 ,million in' damage. Friday'10.00 a.m. (....24 hours after the quake` ` began..:) Utility companies are working around the clock to restore services, yet most people in the areas;of heaviest shaking'lack electricity, natural gas, and : E water. Utility workers, like transportation crews, medical staff, and emergency responders, push themselves to do their crucial jobs despite concerns about their own families. " Donations'of money, services, and material are arriv- ing from all parts of the United States, and a few Red Cross shelters have opened at public schools and undamaged recreation centers, where food, water, ; and personnel are available. Most people, particularly in heavily damaged areas, are camped outside. By now, most stranded motorists have been rescued, and some families are at last,reuniting. To get around; emer- gency responders are using helicopters and any other. means of transportation that works, while residents are using bicycles and four -wheel -drive vehicles. November 15, 2008 • i Saturday 11 :32 p.m. (... 2+ days after the quake began...) A magnitude 5.7 aftershock occurs with an epicenter in Rialta. This is as large as the 1991 Sierra Madre earthquake that killed one person and caused $40 million in damage. Fear of looting far exceeds the reality, yet by now many fearful rumors are cir- culating. Despite official assurances to the contrary, concern grows that if dead bodies are. not recovered and transported away, they will'cause disease outbreaks. • Some are convinced that earthquake scientists are hiding knowledge that an even bigger earthquake is imminent. 12 The ShakeOut Earthquake Scenari" Story That Southern Californian Are Writing November 16, 2008 3 days after the quake began...) It is getting easier for people outside southern California to make contact. • •' with friends and family here. Urban search and rescue teams continue to make rescues, but at a declining rate. Firefighters have extinguished most of the major fires, except where some conflagrations have merged into super- conflagrations— monster fires that consume everything for hundreds of blocks. Many medical staff members have worked with= out sleep since they first began responding to the , disaster Seriously damaged hospitals'have been; evacuated, and open -air trauma centers have been ' set up •i n adjacent areas: 7T er a e :Is• .very short sup- E ply of medical equipment su`cki as'kidney dialysis; ; A' Mac' hines. 5ome,patients'are'being med- evacuated outside the region, to hospitals in'Nevada,'Arizona, and other parts of California Undamaged hospitals have•an influx of earthquake victims with crush . injuries, broken bones, and trauma. This increased , patient load is not distributed evenly, and some ; undamaged hospitals are dramatically overloaded, while others receive few patients. By now, Red Cross shelters have been setup throughout the accessible parts of - r the affected areas. The donation and distribution of money, services, and material have intensified, yet unmet needs are widespread. A coordinated effort among local, State, and Federal :. government agencies is starting to bring water and, food into the region, Our "just in time" economy does • �,;; ,$;, .•;;; ,x not stockpile goods in warehouses. ,; t' '' QIUM ��:r• The National Guard has been mobilized to handle spe= cialized, earthquake- related law enforcement duties, .- allowing local law enforcement to return to regular duties. The fear of looting, intensified by media emphasis, begins to abate. Police and.secilrity person- nel maintain cordons around sites of building collapse, but tenants and owners are allowed to reenter certain other damaged buildings on a very limited basis, 1, 11111,1111111M W III Mid- December, 2008 . <: ..Abouta month later...) .' By now, most gas and electric services have been restored, even in the heavily damaged areas closest to the San Andreas Fault, Landfills store millions of tons of fresh debris. Most of the major roads have reopened, but they are Iiried with heaps of debris. Small bridges that went down or remain damaged are keeping marry local roads dosed. Some freeways also remain closed for repairs, where bridge retrofitting prevented collapse but not all damage. Many commuters who work far from home are unable (and some are � The ShakeOut Scenario By Lucile M. Jones, Richard Bernknopf, Dale Cox, James Goltz, Kenneth Hudnut, Dennis Mileti, Suzanne Perry, Daniel Ponti, Keith Porter, Michael Reichle, Hope Seligson, Kimberley Shoaf, Jerry Treiman, and Anne Wein USGS Open File Report 2008 -1150 CGS Preliminary Report 25 Version 1.0 11: U.S. Department of the Interior U.S. Geological Survey California Department of Conservation (• California Geological Survey U.S. Department of the Interior (• DIRK KEMPTHORNE, Secretary U.S. Geological Survey Mark D. Myers, Director State of California ARNOLD SCHWARZENEGGER, Governor The Resources Agency MIKE CHRISMAN, Secretary for Resources Department of Conservation Bridgett Luther, Director California Geological Survey John G. Parrish, Ph.D., State Geologist U.S. Geological Survey, Reston, Virginia 2008 (6 For product and ordering information: World Wide Web: httpV /www.usgs.gov /pubprod Telephone: 1-888-ASK-USG S For more information on the USGS —the Federal source for science aboutthe Earth, its natural and living resources, natural hazards, and the environment World Wide Web: httpl/www.usgs.gov Telephone: 1 -888- ASK -USGS Suggested citation: Jones, Lucile M., Bernknopf, Richard, Cox, Dale, Goltz, James, Hudnut, Kenneth, Mileti, Dennis, Perry, Suzanne, Ponti, Daniel, Porter, Keith, Reichle, Michael, Seligson, Hope, Shoaf, Kimberley, Treiman, Jerry, and Wain, Anne, 2008, The Shakeout Scenario: U.S. Geological Survey Open - File Report 2008 -1150 and California Geological Survey Preliminary Report 25 [ http :/ /pubs.usgs.gov /of/2008/1150A. Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S. Government Although this report is in the public domain, permission must be secured from the individual • copyright owners to reproduce any copyrighted material contained within this report ii • The ShakeOut Scenario By Lucile M. Jones', Richard Bernknopf , Dale Cox', James Goltz2, Kenneth Hudnut', Dennis Mileti3, Suzanne Perry', Daniel Ponti', Keith Porte?, Michael Reichles, Hope Seligson, Kimberley Shoaf, Jerry Treimans, and Anne Wein' ' USGS 2 Governor's Office of Emergency Services 'California Seismic Safety Commission 4 University of Colorado 'California Geological Survey 6 MMI Engineering University of California, Los Angeles USGS Open File Report 2008 -1150 CGS Preliminary Report 25 Version 1.0 2008 • U.S. Department of the Interior U.S. Geological Survey California Department of Conservation California Geological Survey iii Contents 0 Chapter I. Executive Summary ............................................................. ..............................1 Chapter2. Introduction .............................................................. ............................... ....... 13 Chapter 3. Constructing the Scenario Event .......................... ............................... . ...25 Chapter 4. Physical Damages by Keith Porter ...................................... ..........................93 Chapter 5. Emergency Response and Communications by Dennis Miled and James Goltz ................. ............................... ............................161 Chapter 6. Casualties by Kimberley Shoaf ........... ............................... ............................199 Chapter 7. Regional Economic Consequences by Anne Wein and Adam Rose .............209 Chapter8. Conclusions ......................................... ............................... ............................291 References............................................................. ............................... ............................295 V] iv • The ShakeOut Scenario CHAPTER 1. EXECUTIVE SUMMARY ................................................ ..............................2 Overview .. ................ .. ............ _ ... » ............ ... ............. .. ............ .. ............. .. ... .» ........ .. ..... ....» ..................... ... ... ..... ..... ... 2 Earth Science in the ShakeOut Scenario .. ........................... .. .................... .. ... .. ......... ...... ............ _. ..............................2 TheEarthquake Source ................................................................................. ............................... ..........2 ........................ Ground Motions .............................................................................................................................. ............................... 3 FaultOffsets ............................................... ............................... SecondaryHazards .......................................................................................................................... ............................... 5 Aftershocks..................................................................................................................................... ............................... 5 Engineering in the ShakeOut Scenario ..........................................................»............................... ............................... 6 Buildings......................................................................................................................................... ............................... 6 Non - structural and contents damage ................................... ............................... ,__.....7 ..................... ............................... Utilities, Lifelines, and Infrastructure .................. ............................... ... 7 ......................................... ............................... Fire Following Earthquake ................................................................ ............................... ........................ 8 ................... Social Science in the ShakeOut Scenario .............................. ............................... . g • Emergency Services—. ..................................................................................................... ............................... ...�... Mortality and Morbidity . ' 8 Business Interruption .......... ............................... Movementof Goods .................................................................................................... ............................... ...........10 Conclusions ....... » ...........................................................................................................».................. .............................10 • Chapter 1. Executive Summary Overview This is the initial publication of the results of a cooperative project to examine the implications of a major earthquake in southern California. The study comprised eight counties: Imperial, Kern, Los Angeles, Orange, Riverside, San Bernardino, San Diego, and Ventura. Its results will be used as the basis of an emergency response and preparedness exercise, the Great Southern California ShakeOut, and for this purpose we defined our earthquake as occurring at 10:00 a.m. on November 13, 2008. As members of the southern California community use the Shakeout Scenario to plan and execute the exercise, we anticipate discussion and feedback. This community input will be used to refine our assessment and will lead to a formal publication in early 2009. Our goal in the ShakeOut Scenario is to identify the physical, social and economic consequences of a major earthquake in southern California and in so doing, enable the users of our results to identify what they can change now — before the earthquake —to avoid catastrophic impact after the inevitable earthquake occurs. To do so, we had to determine the physical damages (casualties and losses) caused by the earthquake and the impact of those damages on the region's social and economic systems. To do this, we needed to know about the earthquake ground shaking and fault rupture.' So we first constructed an earthquake, taking all available earthquake research information, from trenching and exposed evidence of prehistoric earthquakes, to analysis of instrumental recordings of large earthquakes and the latest theory in earthquake source physics. We modeled a magnitude (M) 7.8 earthquake on the southern San Andreas Fault, a plausible event on the fault most likely to produce a major earthquake. This information was then fed forward into the Figure 1 -1. ShakeOut Scenario flow -chart. Earth Science in the ShakeOut Scenario The Earthquake Source The ShakeOut Scenario earthquake is a magnitude 7.8 earthquake on the southernmost 300 km (200 mi) of the San Andreas Fault, between the Salton Sea and Lake Hughes. The southern San Andreas Fault was identified in the most recent assessment of seismic risk as most likely source of a very large earthquake in California. A magnitude 7.8 is not the largest earthquake that the southern San Andreas Fault can produce, nor is the San Andreas the only fault to threaten the populated areas of southern California with very large earthquakes. However, those other faults have recurrence intervals (an estimate of the average time) between larger earthquakes that are considerably longer, measured in thousands of years. By contrast, the southern San Andreas Fault has generated earthquakes of ShakeOut size on average every 150 years —and on a portion of the fault that ruptures in the ShakeOut Scenario, the last earthquake happened more than 300 years 0 • A • ago. The extent of the fault rupture in this earthquake was determined from geologic characteristics, after considerable discussion among geologic experts. The most likely rupture initiation point is one of the endpoints of the fault. We started at the southern end of the San Andreas Fault, and ruptured the fault to the northwest. We assumed that the average amount of slip to be released anywhere along the fault would be the amount accumulated since the last event on that portion of the fault, ranging from 2 to 7 meters (6 to 23 ft). We then added a randomized variation of the average slip within each 30 km section of fault. The maximum amount of slip is at the southern end of the rupture near the Salton Sea, where it has been more than 300 years since the last earthquake. Ground Motions The sudden rupture of a fault produces shaking as one of its effects. This shaking moves the ground, and it is these ground motions that we feel and that cause most of the damage in an earthquake. We estimated these ground motions with physics -based computer simulations of the earthquake with computer systems developed by the Southern California Earthquake Center information technology research program. For the past 30 years, before recent advances in information technology that have enabled scientists to obtain meaningful results from physics -based computer simulations, ground motion predictions have typically been made using attenuation relations, which forecast the expected shaking at a site from the magnitude and distance from the fault. However, in any earthquake there are pockets of shaking that are considerably higher or lower because of other factors that affect shaking, including site effects, directivity, and radiation pattern. Our physics -based simulations • modeled all of these factors, primary and secondary, that affect ground shaking, using two inputs: (1) the ShakeOut kinematic rupture description and (2) a velocity model that describes the seismic characteristics of the southern California rocks through which the waves propagate. The results are shown to be consistent with the newest attenuation relations from the Next Generation Attenuation (NGA) relations. We validated our modeling results through comparison of multiple methods, use of distinct velocity models, and comparison with empirically based attenuation relations. In all, four teams were engaged to make independent models of the ground motions. Several features of the ShakeOut earthquake ground motions are consistent across all the models including: • Very strong shaking (approaching 3 m/sec) near the fault; • Strong shaking with medium to long durations (20 -45 sec) in the basins near the fault, including the Coachella, San Bernardino, and Antelope Valleys; • Damaging shaking (at least 0.5 m/sec) overlarge areas (10,000 kn?) of Los Angeles, San Bernardino, and Riverside counties; • Pockets of very strong shaking (>1.5 m/sec) with long durations (45 -60 sec) in areas of the San Gabriel Valley and East Los Angeles. Duration of strong shaking will be an important contributor to damage in any earthquake as large as the ShakeOut Scenario earthquake. Shaking lasts a long time because it takes about 100 seconds for a fault this long to rupture and because some of the waves get trapped and reverberate in sedimentary basins. In the ShakeOut Scenario earthquake (fig. 1 -2), the San Bernardino Valley is shaken extremely strongly but for a relatively short duration, as are Wrightwood and Palmdale, while the Coachella Valley has strong shaking with a long duration. Lower amplitude, but much longer duration ground motions occur in the Los Angeles and Ventura sedimentary basins. ar W 71 U8W 8vda■M§P :8h■ §0AM7.8 8oea Wb V2 Nwismimaim Y ri mm inia f Qde TAM N Eniodk ! M 4W •118• •118' 414 ft.+M1W9ft ..WrWNw-ftMoraWWr.asrx� Ws Iorei■, 4M W ■MI I r� ■lei IN�M 111 r +�. wed � w.r NUA N d Ist: �.0Lu waN sw asp nay •w a� amp w� "JEN d1 91-L1 .. U44 a , aw! ON !!i Wds ■1ri 0 Figure 1 -2. This "ShakeMap" is a representation of the shaking produced by the Shakeout Scenario earthquake. The colors represent the Modified Mercalli Intensity with the warmer colors representing areas of greater damage. To estimate damages from Shakeout ground motion, the ShakeOut Scenario next calculated ground motion parameters used by engineers to estimate damage to structures. Ground motion parameters describe how the ground moves due to different measures of earthquake waves, and are needed because different kinds of structures are damaged by different kinds of waves. The ShakeOut Scenario created all the standard ground motion parameters: peak ground acceleration (PGA), peak ground velocity (PGV), Modified Mercalli Intensity (MW, and spectral accelerations at 0.3, 1.0, and 3 seconds. Fault Offsets Fault offsets occur where the fault that moves in the earthquake is exposed at the Earth's surface. The ShakeOut fault rupture is on the San Andreas Fault and will be dominated by strike - slip, or horizontal displacement, causing structures and lifelines that straddle the fault to be sheared and offset as much as 9 meters (30 feet). Fortunately, there are few structures at risk of direct fault damage from the ShakeOut earthquake, due to the rural setting of the southern San Andreas Fault zone, and to the Alquist -Pdolo Earthquake Fault Zoning Act of 1972, which prevents the construction of buildings used for human occupancy on the surface trace of active faults. Damage from ShakeOut surface rupture is most serious where lifelines (roads, railroads, and utilities) cross the fault. Many of these crossings are concentrated within a few mountain 10 4 • passes and the disruption to these lifeline corridors has a major economic impact. Roads cross the fault at 966 places; the most critical damage occurs to Interstate 10 in the Coachella Valley and in San Gorgonio Pass, Interstate 15 in Cajon Pass, CA -14, CA -111, CA -62, Box Canyon Road, and Big Pines Highway. Other disrupted lifelines include fiber optic cables (90 crossings), petroleum and natural gas pipelines (39 crossings), railroads (21 crossings), aqueducts (32 crossings), and overhead electric power transmission lines (141 crossings). Secondary Hazards We investigated secondary hazards that can be triggered by large earthquakes in southern California including liquefaction, landslides, tsunamis, and seiches. All of these have caused significant additional damage in many big earthquakes, but only landslides and liquefaction will produce significant impacts in the ShakeOut Scenario. The Shakeout Scenario earthquake will produce between 10,000 and 100,000 individual landslides, the vast majority of which will consist of rock falls, rock slides, rock avalanches, soil falls, disrupted soil slides and soil avalanches. Most of these will occur on steep slopes within the Transverse Ranges, primarily in the eastern San Gabriel Mountains. Conditions that can lead to liquefaction are potentially widespread in parts of the eight - county area impacted by the ShakeOut Scenario earthquake, particularly the Santa Clara River /Oxnard Plain areas of Ventura County, parts of the San Fernando and San Gabriel Valleys, portions of the coastal basin or flatland areas of Los Angeles and Orange Counties, the Santa Ana River corridor, the Imperial Valley, the southern Coachella Valley, and coastal areas of San Diego County. However, liquefaction requires both strong shaking and a high ground -water table. Strong ground motions from the ShakeOut Scenario earthquake mostly occur within the inland desert and • mountain regions of southern California where ground water levels are typically low year - round. As a result, only the southern Coachella Valley will suffer significant liquefaction impacts in the ShakeOut Scenario earthquake, with localized liquefaction otherwise confined mostly to areas adjacent to perennial stream and river channels, such as in the upper Santa Ana and Santa Clara river basins. Because of the large distance from the earthquake to the coast, tsunamis are not a significant risk. Aftershocks Aftershocks are earthquakes and cause shaking and damage just like any other earthquake. Their additional shaking can damage weakened structures, necessitate evacuations, endanger rescue workers, and undo efforts to restore and rebuild. Based on experience in numerous earthquakes worldwide, after a mainshock earthquake as large as the ShakeOut Scenario earthquake, damaging aftershocks can occur for decades in a broad region around southern California, and any given region may experience more severe shaking from a close aftershock than from the original mainshock. Aftershock behavior in the aggregate can be well described by some simple, empirical laws, and these can be used to simulate sequences of aftershocks that realistically mimic actual aftershock sequences. For the ShakeOut Scenario, we generated ten random realizations of aftershocks for the first week following our mainshock. In reality, large, damaging aftershocks may occur months or years after the initial event. We picked one of the simulations to be the aftershocks for the ShakeOut drills. This sequence includes two magnitude (M) 7 aftershocks. A M7.0 aftershock occurs 33 minutes after the mainshock, beginning at the southern end of the mainshock, near the Salton Sea, and rupturing south toward Mexico. It causes damage in Imperial and eastern San Diego Counties as well as in • Mexicali, Mexico. A M7.2 event occurs 17 hours after mainshock on the Cucamonga Fault, rupturing along the front of the San Gabriel Mountains from Cajon Pass to Monrovia. The aftershocks in this sequence would cause substantial additional damage, but neither large aftershock has been evaluated in detail. Engineering in the ShakeOut Scenario The damage and impacts of the ShakeOut Scenario earthquake were estimated through a three -step process. First, FEMA's loss estimation program, HAWS, was run using the physics - based ground motion model. For Los Angeles County, HAWS used a refined database of structures created from tax assessor's data. For the other counties, this was not available and the default HAWS database was used. In addition, HAZUS default mapping schemes (the relationships between basic inventory data and the assumed structural characteristics) were modified to reflect available information on unreinforced masonry buildings tabulated by the California Seismic Safety Commission, building density concentrations in urban core areas, and construction pattern changes over time throughout the eight counties. In the second step, expert opinion was collected through 13 special studies and 6 expert panels. Panels generally estimated impacts to public utilities, especially where multiple utility companies provide a public service such as water supply or electricity. Engineers and operators were invited to attend the half -day panel discussions, and were presented the results of prior Earth science studies (shaking, faulting, etc.), as well as damage to other interacting lifelines that had already been assessed. They were then asked to posit a realistic scenario of damage, service interruption, restoration, and to suggest promising mitigation options. To complement the panels, special studies were used for buildings and for lifelines when the panel process was impractical, such as private utilities or utilities (such as highways) where in -depth analysis was desired. In these cases, contributors were selected for their specialized expertise. They too were presented with all previously estimated Earth- science and relevant utility impacts, and asked to summarize assets exposed to damage, evidence of past seismic vulnerability, and to posit a realistic scenario of damage, loss of function, restoration, and promising mitigation measures. Crucial special studies were reviewed by panels of highly qualified experts. In the third step, the expert evaluations were merged with the HAZUS results to create the final estimates of probable damages. The major losses for this earthquake fall into four categories: building damages, non- structural damages, damage to lifelines and infrastructure, and fire losses. Within each category, the analysis found types of losses that are well understood —that have been seen in previous earthquakes and the vulnerabilities recognized but not removed —and types of losses that had been less obvious – where the type of failure is only recently understood or the extent of the problem not yet fully recognized. The study also found numerous areas where mitigation conducted over the last few decades by state agencies, utilities and private owners, has greatly reduced the vulnerability. Because of these mitigation measures, the total financial impact of this earthquake is estimated to be "only" about $200 billion with approximately 1,800 fatalities. However, these are still big numbers Buildings Total losses to buildings are estimated at $33 billion. The two classes of older, known, poor performers -- unreinforced masonry (where bricks or stone blocks with mortar form the bearing walls, called "URM") and non - ductile reinforced concrete buildings - -pose the greatest risk to life safety. These types of buildings are no longer allowed to be built, but many of these buildings still exist and are not retrofitted. These types of buildings will be heavily damaged or destroyed near the fault, but in general will suffer less damage in the Los Angeles area. All URM buildings in the City of Los Angeles have been evaluated, and most have been strengthened to reduce loss of life. The 6 • strong shaking in Los Angeles will have very long periods (the waves will be big but slow) and these smaller buildings will in many cases ride out the shaking with less damage. Woodframe construction generally fares well in earthquake shaking and woodframe buildings are less likely than other types of buildings to be damaged. However, because woodframe construction is so prevalent in California, substantial losses will still occur. Woodframe building damage is most likely: • in older homes where the house is not bolted to the foundation or the cripple wall is not reinforced. • in buildings with a "soft first story" — a large opening such as garage door or display windows on the first floor and without compensating reinforcement. • in buildings where building codes were not rigorously followed - -a condition difficult to recognized until after the earthquake. Steel moment frame buildings built before 1994 were found to form cracks in their connections during the 1994 Northridge earthquake. Similar damage occurred in the 1995 Kobe earthquake and some buildings collapsed. Special study was conducted to analyze the behavior of steel frame high -rise buildings in the ground motions modeled for this earthquake. This event shows amplified long period motions caused by resonance in the sedimentary basins, particularly the very deep Los Angeles Basin. A special panel of structural engineers evaluated the analytical study and concluded "Given these ground motions, the collapse of some pre-1994 welded -steel moment -frame buildings is a credible scenario." Because this result comes from the long period • ground motions, the area where this type of damage is possible is relatively large and includes much of the urbanized areas of Los Angeles, Orange, Riverside and San Bernardino Counties. It is impossible to determine how many and which buildings are the most susceptible without detailed structural analysis which is beyond the scope of this study. For the purposes of the ShakeOut emergency drills, we posit that 5 steel moment -frame high -rise buildings will collapse and that 10 more will be "red- tagged." Non - structural and contents damage Non - structural and contents damage is damage to the parts of a building other than what is holding it up, including interior walls, water pipes, air conditioning systems, and all moveable property such as electronics, and dishes. As building codes improve and buildings remain standing during earthquakes, the relative importance of non - structural damage increases. In recent earthquakes, the non - structural and contents losses have typically been comparable to the structural losses. Non - structural damages and mitigation have not been regulated in any way. Many of these losses are simple to prevent through securing contents and non - structural elements of the buildings. This is one of the most important ways that individuals can reduce the losses. Utilities, Lifelines, and Infrastructure California's investments in mitigation have paid off most obviously in increased robustness and resiliency of the region's lifelines. The retrofitting of highway bridges, conversion of ceramic insulators in the electric grid to polymers, and replacement of cast iron pipes mean that many utilities will be able to restore function much more quickly after the earthquake. Significant vulnerabilities remain in the water conveyance system and in the lifelines that • cross the San Andreas Fault. Pipes of concrete and iron are brittle and break in many places in an l earthquake. The number of pipe breaks will be large enough that recreating the water system will be necessary in the hardest hit areas. Because this earthquake affects such a large area, there will not be enough pipe and connectors or trained manpower to repair all the breaks quickly. The worst hit areas may not have water in the taps for 6 months. This damage to the water system will also greatly increase the problems in fighting the fires that will follow the earthquake. The cost to repair water and sewer lines will be $1 billion. The lifelines that cross the fault will all break when the fault moves. This will disrupt the movement of water, petroleum products, telecommunications, and general transportation. Repair of the lifelines will be slowed because the lifelines all cross the fault at just a few passes in the mountains and therefore interact with each other. For instance, repairing pipelines broken at Cajon Pass will require access that depends upon repair to Interstate 15. That in turn could be delayed if a wildfire starts after damage to the electric lines in the same location. Many roads and highways will be impassable in the first few days after the earthquake because of debris on the roads, damage to bridges, and lack of power for the traffic signals. This will have a significant negative impact on the emergency response. Because of the major highway bridge retrofit program of the last 20 years, highway bridges are not expected to completely collapse, but some will not be passable. Many bridges on local roads have not been retrofitted and more damage is expected on those. The continuing impairment of the roads for months after the earthquake until everything can be repaired has a significant economic cost, estimated at $5 billion over one year. Fire Following Earthquake Southern California is unfortunately well situated for major fires to be generated following • earthquakes. The number of ignitions that will create fires large enough to call the fire department can be extrapolated from previous earthquakes and depends upon the number of households at different levels of seismic shaking. This leads to an estimate of 1,600 ignitions of which 1,200 will be too large to be controlled by one fire engine company. In areas of dense woodframe construction, these fires if not controlled will grew quickly to involve tens or hundreds of city blocks. The fire risk is increased by the damage to the water distribution system and by the traffic gridlock that will result from the Shakeout earthquake. The final level of fire damage is difficult to assess because it depends upon several unpredictable factors, especially the degree to which fires spread when the fire protection services lose water and are overwhelmed. We use the minimum value from the fire estimates at $40 billion in damage to buildings and $25 billion in damage to building contents. Social Science in the ShakeOut Scenario The ShakeOut Scenario earthquake causes damage to the built environment that then ripples through and damages the social systems of the study region. This study has investigated the impacts of the earthquake on emergency services, human health, the regional economy, and trade operations from the Ports of Los Angeles and Long Beach. Emergency Services An emergency response matrix has been developed to help understand what the demands for emergency services will be like. Seventeen functions of emergency services are grouped into seven general classes of activities, including crisis information (public information and responder communications), search and rescue, victim services (shelter, provision of food and water and the • management and distribution of donated goods and services), access management and law enforcement (control and security and traffic control), the staffing and functioning of emergency operations centers, fire suppression, medical emergency response, and service restoration, (repair of utilities, route recovery and debris removal). Research.results and experience in past earthquakes have been analyzed to create this response matrix. Among the findings are that 95% of rescues from downed buildings are carried out by fellow victims. Training ordinary citizens how to search safely could greatly reduce injuries. • Many Emergency Operations Centers have not considered the impact of earthquakes on the contents of their Centers. Securing computers and desks and other non - structural mitigation activities would have large payoffs at low cost. Mortality and Morbidity Shaking in the ShakeOut Scenario earthquake will kill and injure many people, by causing buildings to collapse, creating falling debris and flying objects, and increasing traffic accidents when drivers lose control of automobiles. Many additional deaths and injuries will result in fires that follow the shaking. Estimating the total number of injuries and deaths is very uncertain particularly because the Scenario posits types of building failures that have not yet been observed. Because of strong life- safety building codes over the years, the Shakeout Scenario estimates only approximately 1,800 deaths, of which about half occur because of the fires following the earthquake. There will also be about 750 people with very severe injuries who will require rapid, advanced medical care to survive. Approximately 50,000 people will have injuries that need emergency room care. The final mortality could increase if hospitals cannot function because of • damage or if the transportation disruptions prevent people getting to emergency rooms. Business Interruption The economic impact of the earthquake is not limited to the structures and goods broken or burnt in the event. Much of the economic activity of the southern California region will be interrupted by the damage to structures and infrastructure. In particular, beyond their direct losses in stock (such as buildings, machines, and inventory), businesses will be unable to function because of loss of electricity, gas, water, and a transportation system. Some of the losses can be recaptured when the business resumes but the amount recaptured decreases with time as customers and suppliers find alternatives. Because the duration of outage is so long, the lack of water conveyance becomes the largest factor in business interruption losses for the ShakeOut earthquake, resulting in $50 billion in lost economic activity. • Northwest Trade va ue: $3.213 JgtzL39,900 aae value: $52.0 Jobs: 1,114,700 Jobs: Great Plains Trade va 9.313Great Lakes Jobs: 243,200 Trade value: $53.7B j J _ bs: 681,800 ---A Southeast I tra Trade value: $37. $32.513 Jobs: 498,900 700 ae value: $25. Jobs: 275,300 Figure 1 -3. National Impact of San Pedro Ports. Source: BST Associates Trade Impact Report; 2007. Movement of Goods The Shakeout Scenario earthquake will be far enough from both Los Angeles International Airport and the Ports of Los Angeles and Long Beach that the damage there will be minimal. This of course is not the case for many other possible earthquakes. The availability of these transportation resources is a significant asset in mobilizing the emergency response. Transportation from the Ports to the rest of the country is carried predominately by rail lines which will be rendered impassable by the fault offsets and ground motions. Significant economic disruption will result and the extent of the damage depends critically on how rapidly the railways and highways can be rebuilt. The modeling estimates that the Ports will not function for the first 3 days after the event because of lack of electricity, general chaos, and the potential for slight damage to large structures such as cranes. For the next 2 weeks, the Ports will operate at 10% of capacity. because there will be limited rail service and limited alternative transportation. They will gradually return to full capacity from 2 weeks to 2 months as rail service is reestablished and highways reopen. We estimate that 85% of the lost business will be recaptured but that 15% will be permanently lost to ship diversions, perished products, cancelled Far East shipments, and declined bookings. Conclusions The magnitude 7.8 ShakeOut earthquake is modeled to cause about 1,800 deaths and $213 billion of economic losses. These numbers are as low as they are because of aggressive retrofitting programs that have increased the seismic resistance of buildings, highways and lifelines, and economic resiliency. These numbers are as large as they are because much more retrofitting could still be done. The sources of the different losses are shown in Table 1 -1. Table 1 -1. Total Regional Economic Impacts of Shake -Out (in billions of 2008 dollars). 10 �e • Indicator Total Impacts Building Damage $32.7 Related Content Damage 10.6 High -Rise Building Damage 2.2 Related Content Damage 0.7 Fire Damage 40.0 Related Content Damage 25.0 Highway Damage 0.4 Pipeline (water, sewer, gas) Damage 1.1 Sub -total Property Damage 112.7 Business Interruption 96.2 Relocation Costs 0.1 Traffic Delay Costs 4.3 Sub -total Additional Costs 4.4 Total $213.3 The earthquake modeled here may never happen. Big earthquakes on the San Andreas Fault are inevitable, and by geologic standards extremely common, but probably will not be exactly like • this one. The next very damaging earthquake could easily be on another fault. However, lessons learned from this particular event apply to many other events and could provide benefits in many possible future disasters. The ShakeOut Scenario has identified five major areas of loss: • Older buildings built to earlier standards. • Non - structural elements and building contents that are generally unregulated. • Infrastructure crossing the San Andreas Fault. • Business interruption from damaged infrastructure, especially water systems. • Fire following the earthquake. The ShakeOut Scenario also found that previous efforts to reduce losses through mitigation before the event have been successful. There are dozens more actions and policies that could be undertaken at the individual and community levels to further reduce these losses. For instance, actions to improve the resiliency of our water delivery system would reduce the loss from business interruption, as well as reduce the risk of catastrophic conflagrations. At an individual and business level, actions to secure non - structural items in buildings and retrofitting of existing structures will greatly reduce individual risk. Planning and preparedness can improve personal and business resiliency. Over the next 6 months, the ShakeOut Scenario will be used to prepare for future earthquakes and exercise in the Great Southern California ShakeOut in November 2008. This process will encourage public discussion of these risks and possible solutions. The risks can be analyzed and described by scientists but the solutions will come from southern California itself. I• 11 • Receive and File — Information on 'The Shakeout Earthquake Scenario" and the °Califomia Earthquake Rupture Forecast" City Council Staff Report July 14, 2008 ATTACHMENT 2 DOCUMENTS REGARDING THE "UNIFORM CALIFORNIA EARTHQUAKE RUPTURE FORECAST': ❑ "FORECASTING CALIFORNIA'S EARTHQUAKES - WHAT CAN WE EXPECT IN THE NEXT 30 YEARS?", USGS FACT SHEET 2008 -3027; ❑ "THE UNIFORM CALIFORNIA EARTHQUAKE • RUPTURE FORECAST, VERSION 2 (UCERF)" WEB PAGE, HTTP :HPUBS.USGS.GOV /OF/2007/1437, DOWNLOADED APRIL 15, 2008; AND • ❑ "THE UNIFORM CALIFORNIA EARTHQUAKE RUPTURE FORECAST, VERSION 2 (UCERF)", DATED 2008 - CONTENTS, LIST OF APPENDICES, AND EXECUTIVE SUMMARY. (COMPLETE REPORT AVAILABLE AT THE DEPARTMENT OF DEVELOPMENT SERVICES TO REVIEW) Uniform Earthquake Forecast.CC Status Report $ • Forecasting California's Earthquakes What Can We Expect in the Next 30 Years? In a naw comprehensive scientists h ft deteftine'd that the chance orhaving one or more piagnt tude 6.7 brlarger earth#pakes in the California area over Me'next 30 years is greater than 99 %. Such quakes -can bey;K, daedty; as shown by the7196fmagnitude 6.9 Loma Prieta and the 1994 negnituda� _` 671Uorthridge earthquakes. the like- _...- =• - -- ,. insurance, and i will be affected What Is an Earthquake Rupture Forecast? Californians know that their State is subject to frequent —and sometimes very destructive — earthquakes, Accurate forecasts of the likelihood of quakes can help people prepare for these inevitable events. Because scientists cannot yet make precise predictions of the date, time, and place of future quakes, forecasts are in the form of the probabilities that quakes of certain sizes will occur during specified periods of time. In our daily lives, we are used to making decisions based on probabilities —from weather forecasts (such as a 30% chance of rain) to the annual chance of being killed by lightning (about 0.0003%). Similarly, earthquake prob- abilities derived by scientists can help us plan and prepare for future quakes. Earthquake forecasts for California have been developed in the past by multidisciplinary groups of scientists and engineers, each known as a "Working Group on Califomia Earthquake Probabilities" ( WGCEP 1988, 1990, 1995, 2003). However, tllose forecasts were limited to particular regions of California Because of this, WGCEP 2007 was commissioned to develop an updated, statewide forecast, the latest result of which is the Uniform California U.S. Department of the Irwar U.S. Geological Survey Earthquake Rupture Forecast, Version 2, or " UCERF" (U.S. Geological Survey (USGS) Open -File Report 2007 -1437, http: //pubs.usgs. gov/of/2007/14370. Organizations sponsoring WGCEP 2007 include the USGS, California Geological Survey, and the Southern California Earthquake Center. The comprehensive new forecast builds on previous studies and also incorporates abundant new data and improved scientific understanding of earthquakes. When an earthquake occurs, two things happen —a fault ruptures (a crack in the Earth's crust gives way and slips under tectonic pres- sure) and seismic waves, caused by this sudden fault motion, radiate out like ripples from a pebble tossed into a pond. The shaking that occurs as seismic waves pass by causes most quake damage. The strength of the waves depends party on the quake's magnitude, which is a function of the size of the fault that moves and the amount of slip. The UCERF study's goal was to determine probabilities for different parts of California of earthquake ruptures of various magnitudes, but not to estimate the likelihood of shaking ("seismic hazard ") that will be caused by these quakes. This distinction is important, because even areas in the State with a low probability of fault rupture can experience shaking and dam- age from distant, powerful quakes. USGS Fact Sheet 2008.3027 2008 California sits on the boundary between two of the Earth's major tectonic plates —the Pacific and North American Plates —which move inexorably past each other at a rate of about 2 inches per year. Much of this motion is accommodated from time to time by sudden slip on faults, producing earthquakes. Although the San Andreas Fault is the main locus of slip, hundreds, if not thousands, of other faults splay out from the plate boundary, spreading the threat of large earthquake ruptures through most of the State. T he new Uniform California Earthquake Rupture Forecast ( UCERF) combines information from geodesy (precise data on the slow relative movement of the Earth's tectonic plates), geology (mapped locations of faults and docu- mented offsets on them), seismology (occurrence patterns of past earth- quakes), and paleossismology (data from trenches across faults documenting the dates and offsets of past earthquakes on them). The firstthree kinds of data are shown here as layers in the diagram. All four kinds of data are combined mathematically to produce the final probability values for future ruptures in the California area, in regions of the State, and on individual faults. Building on several previous studies and decades of data collection, UCERF was developed by a multidisciplinary group of scientists and engineers, known as the 2007 Working Group on California Earthquake Probabilities. Advice and commentwas sought regularly from the broader community of earthquake sci- entists and engineers through open meetings and workshops. Where experts disagreed on aspects of the forecast, alternative options were accounted for in calculations to reflect these uncertainties. The final forecast is a sophisticated integration of scientific fact and expert opinion. Seismograph C Geologic field mapping and aerial photos trace out California's many faults and documentthe accumulated slip in earth- quakes over thousands of years. Color spectrum shows rates of slip, from fast (purple and red) to ve slow (dark blue). TECTONIC MOVEMENT N -I a to Global positioning system (GPS) observations by satellite document how fast various points in California are moving (arrows) in —response to the steady motion of the Pacific and North FAULTS , • 0.01% 0.1% 1% -0 The final forecast results from evaluating and integrating several types of scientific data. Trenching across the Hayward Fault in Fremont is How Likely is a Damaging Quake in the Next 30 Years? California straddles the boundary between two of the Earth's tectonic plates —as a result, it is broken by numerous earthquake faults. Taking into account the earthquake histories and relative rates of motion on these many faults, the UCERF study concludes that there is a probability of more than 99% that in the next 30 years Californians will experience one or more magnitude 6.7 or greater quakes, potentially capable of causing extensive damage and loss of life. For powerful quakes of magnitude 7.5 or greater, there is a 46% chance of one or more in the next 30 years --such a quake is twice as likely to occur (37%) in the southern half of the State than in the northern half (15%). Smaller magnitude earthquakes are more frequent than larger quakes. According to the new forecast, about 3 magnitude 5 or greater quakes will occur in the California region per year, and a magnitude 6 or greater quake about every 1.5 years. These numbers do not include aftershocks that follow larger quakes urclud- ing them would roughly double the expected number of magnitude 5 or greater quakes. �TEWIDE F IU PRO 1 The numbers Fepresent curreht best estimates. As earthquake science progresses, flies- prob- , abilities iiiill change: ,Actual repeat•times'v, J considefably and only rarely will be exactly as listed io the table. „ 30 -year probability of "Average- " Magnitude one or more events' * repeat . greateitlianoe4quai ' ?`.time ; to the magnitude (jrearsl;"- . -7 •X 94%. • ; '7.5 '•w 8 ,R , • ,14% *Not including Cascadia Subduc#6n Zane ,, r For the entire California region, the fault with the highest probability of generating at least one magnitude 6.7 or larger earthquake is the southern San Andreas (59% in the next 30 years). For northern California, the most likely source of such a quake is the Hayward- Rodgers Creek Fault (31 % in next 30 years) --see USGS Fact Sheet 2008 -3019. Quake probabili- ties for many parts of the State are similar to those in previous studies, but the new probabil- ities for the Elsinore and San Jacinto Faults in southern California are about half those previ- ously determined. For the far northwestern part of the State, a major source of quakes is the offshore 750- milelong "Cascadia Subduction Zone," which extends south about 150 miles into California. For the next 30 years there is a 10% probability of a magnitude 8 to 9 quake somewhere along the zone --such quakes occur about every 500 years. ® Printed on recycled paper _. Which Faults Currently Have Elevated Probabilities? The new California earthquake forecast incorporates the con- "• cept that earthquake probabilities change with time. For exam- pie, a large quake may be less likely in the near future on a fault that has recently had on"ut more likely on a fault on which r. # the tectonic stresses have had much time to build back up. The ^?:fVorthem San Andreas} ault forecast also incorporates adjustments for areas that have recently shown a change in the level of earthquake activity" Faults that have elevated current probabilities include the south- '; ern San Andreas and Hayward - Rodgers Creek Faults, though Hayward - Rodgers major quakes on these faults may still be decades away. Creek Fault Calaveras Fa Current Earthquake Probabilities =� Relative to Long -Term Probabilities i� 1.6times Equal 1.6times Garlock Fault smaller greater 'A \". Southern San Andreas Fault San Jacinto Fault Elsinore Fault a 200 MILES, f 1 0 200 KILOMErERS The UCERF forecast was evaluated by an independent scientific review panel, as well as by both the California and National Earth- quake Prediction Evaluation Councils, making it one of the most extensively reviewed earth- quake forecasts ever produced. Uncertainties remain because the new quake probabilities are the result of evaluating and accommodat- ing several earthquake theories. As scientific understanding of quakes improves, the prob- abilities will change. The results of the UCERF study are a reminder that all Californians live in earthquake country and should therefore be prepared (see Putting Down Roots in Earthquake Country at httpJ /www carthquakccountry .info/roots). The Ndivi UAL The, I TIE The• UCERF repdrt` assigns individual probabili=•, x ties to 'specific known major faults: B -low ere:; 30 -year probabilities for seven of the faults for' which scientists,have.the most data. Many other „faults also'have'significent probabilitie"n fact• the next big,quake in Califomia is just as likely,t4_ " occur or one of thdotherfaults iq the State: ; ProSabilityof ooe or- I more magnitude.6.7; ^c or greater quake Soiitherii San Andreas 59966 ' Iiayvvard- Rodgers6reajc 31% I ski Jac htt ' �Ttiitliein S�ati,Andreas • 219G, Efsiiibie Tw ::i 11 %. kV `Gar1Qi is 696'. USGS has already used the UCERF to estimate California's seismic hazard, which in turn will be used to update building codes. Other subsequent studies will add information on the vulnerability of manmade structures to estimate expected losses ("seismic risk"). In these ways, UCERF will help to increase public safety and community resilience to earthquake hazards. Earthquakes cannot be prevented, but the damage they do can be greatly reduced through prudent planning and preparedness. The ongo- ing work of USGS, California Geological Survey, Southern California Earthquake Center, and other scientists in evaluating quake prob- abilities is part of the National Earthquake Haz- ard Reduction Program's efforts to safeguard lives and property from the future quakes that are certain to strike in California and elsewhere: in our Nation. Edward H. Flelit Kevin R. Milner, and the 2007 Working Group on California Earthquake Probabilities Edited by Peter H. Shaeffer and lames W. Hendley H lbw by David R. ]ones COOPERATING ORGANIZATIONS Southern California Earthquake Center (SCEC) California Geological Survey (CGS) California Earthquake Authority For more information contact Earthquake Information Hotline (650) 3294085 U.S. Geological Survey, Mail Stop 977 345 Middlefield Road, Menlo Park, CA 94025 htW.11 atthquake.- SLgov/ http: //www.scec.org This Fact Sheet and any updates to it are available online at httpJ/pubs.usp.gov /f&i=8no27/ • .he Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2) Prepared in cooperation with the '"alifornia Geological Survey and the Southern California Earthquake Center JSGS Open File Report 2007 -1437 :GS Special Report 203 XEC Contribution #1138 Pagel of 3 USGS Home Contact USGS Search USGS The Uniform California Earthquake Rupture Forecast, Version 2 (UCERIF 2) By 2007 Working Group on California Earthquake Probabilities* 2008 and H. Field, Timothy E. Dawson, Karen R. Felzer, Arthur D. Frankel, Vipin Gupta, Thomas H. Jordan, Tom sons, Mark D. Petersen, Ross S. Stein, Ray J. Weldon II, and Chris J. Wills Predicted fte of Ell t 6.5 Ratb of Preftbd Rata of M t 6.5 to Extrapolated Rats ft m W5 ar•o►probabaar ayNrpmbaunyr I N milli: 11111111 Is 0111111 .1111, ;'1'111111;, nNa 1.6m 0.001% tkim 0.t Q9 ?,0 32 6p itlp:/ /pubs.usgs.gov /of/2007/1437/ 4/15/2008 The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2) Page 2 of 3 California's 35 million people live among some of the most active earthquake faults in the United States. Public safety demands credible assessments of the earthquake hazard to maintain appropriate building codes for safe construction and earthquake insurance for loss protection. Seismic hazard analysis begins with an earthquake rupture forecast —a model of probabilities that earthquakes of specified magnitudes, locations, and faulting type will occur during a specified time interval. This report describes a new earthquake rupture forecast for California developed by the 2007 Working Group on California Earthquake Probabilities (WGCEP 2007). Main Text Download the main text of this 104 -page report (of2007- 1437_text.pdf; 11.5 MB). Download iust the Executive Summary as a 7 -page report (of2007- 1437_summary.pdf; 0.7 MB). Download a supplemental Excel workbook. This workbook contains supplementary materials in 13 spreadsheets (of2007- 1437_data.xis; 0.3 MB): • Sheet (1) Rupture Magnitudes & Rates • Sheet 2: Total 30 -year Rupture Probabilities • Sheet 3: Total 5 -year Rupture Probabilities • Sheet 4: M >_6.7 30 -year Rupture Probabilities • Sheet 5: Segment Rates & Mean Recurrence Intervals • Sheet 6: Total 30 -year Segment Probabilities • Sheet 7: Total 5 -year Segment Probabilities • Sheet 8: M>_6.7 30 -year Segment Probabilities • Sheet 9: M>_6.7 Probabilities in WGCEP (2003: Box • Sheet 10: Total 30 -year Probabilities on Faults • Sheet 11: M>_6.7 30 -year Probabilities on Faults • Sheet 12: B -Fault Data • Sheet 13: Region Polygons lkppendixes appendix A: California Fault Parameters for the National Seismic Hazard Maps and Working Group on California arthquake Probabilities 2007, by C.J. Wills, R.J. Weldon II, and W.A. Bryant kppendix B: Recurrence Interval and Event Age Data for Type A Faults, by T.E. Dawson, R.J. Weldon II, and G.P. 3iasi kppendix C: Monte Carlo Method for Determining Earthquake Recurrence Parameters from Short Paleoseismic :atalogs: Example Calculations for California, by T. Parsons kppendix D: Magnitude -Area Relationships, by R.S. Stein • kppendix E: Overview of the Southern San Andreas Fault Model, by R.J. Weldon II, G.P. Biasi, C.J. Wills, and T.E. )awson kppendix F: Summary of Geologic Data and Developments of A- Priori Rupture Models for the Elsinore, San Jacinto, ind Garlock Faults, by T.E. Dawson, T.K. Rockwell, R.J. Weldon II, and C.J. Wills kppendix G: Development of Final A -Fault Rupture Models for WGCEP /NSHMP Earthquake Rate Model 2.3, by E. 7eld, R.J. Weldon II, V. Gupta, T. Parsons, W.J. Wills, T.E. Dawson, R.S. Stein, and M.D. Petersen tppendix H: WGCEP Historical California Earthquake Catalog, by K.R. Felzer and T. Cao ttp : / /pubs.usgs.gov /of/2007/1437/ an SmnnQ The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2) Appendix I: Calculating California Seismicity Rates, by K.R. Felzer Page 3 of ? Appendix J: Spatial Seismicity Rates and Maximum Magnitudes for Background, by M.D. Petersen, C.S. Mueller, 61 Frankel, and Y. Zeng Appendix K: A- Priori Rupture Models for Northern California Type -A Faults, by C.J. Wills, R.J. Weldon II, and E.H. Field Appendix L: Cascadia Subduction Zone, by A.D. Frankel and M.D. Petersen Appendix M: Empirical Estimation of Regional Time Variation in Seismicity, by K.R. Felzer Appendix N: Conditional, Time - Dependent Probabilities for Segmented Type -A Faults in the WGCEP UCERF 2, by E. H. Field and V. Gupta Appendix O: Paleoseismic Investigations of the Walnut Site on the San Jacinto Fault, by T.E. Fuma/ and K.J. Kendrick (three large sheets) Appendix P: Compilation of Surface Creep on California Faults and Comparison of WGCEP 2007 Deformation Model to Pacific -North America Plate Motion, by B.A. Wisely, D.A. Schmidt, and R.J. Weldon II For questions about the content of this report, contact William Ellsworth Suggested citation and version history Iwo of Interest USGS Fact Sheet 2008 -3027 Forecasting California's Earthquakes —What Can We Expect in the Next 30 Years? by Edward H. Field, Kevin R. Milner, and the 2007 Working Group on California Earthquake Probabilities Download a free copy of the latest version of Adobe Reader. Help I PDF help I Publications main page I Western Open -File Reports for 2007 Geology I Earthquake Hazards Program I This report is available in paper form with a CD -ROM in the back from the California Geological Survey. Accessibility FOIA Privacy Policies and Notices U.S. Department of the Interior I U.S. Geological Survey URL: http : / /pubs.usgs.gov /of/20"07/1437/ Page Contact Information: Michael Diggies bae Last Modified: April 14, 2008, 10:00 AM http : / /pubs.usgs.gov /of/2007/1437/ 4/15/2001 SC EC PM'O --ousGs 1W ence>hiia Than- - ra�rid The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2) By 2007 Working Group on California Earthquake Probabilities* USGS Open File Report 2007 -1437 CGS Special Report 203 • SCEC Contribution #1138 2008 *Edward H. Field, Timothy E. Dawson, Karen R. Felker, Arthur D. Frankel, Vipin Gupta, Thomas H. Jordan, Tom Parsons, Mark D. Petersen, Ross S. Stein, Ray J. Weldon II, and Chris J. Wills U.S. Department of the Interior U.S. Geological Survey • California Department of Conservation California Geological Survey U.S. Department of the Interior DIRK KEMPTHORNE, Secretary U.S. Geological Survey Mark D. Myers, Director State of California ARNOLD SCHWARZENEGGER, Governor The Resources Agency MIKE CHRISMAN, Secretary for Resources Department of Conservation Bridgett Luther, Director California Geological Survey John G. Parrish, Ph.-D., State Geologist U.S. Geological Survey, Reston, Virginia 2008 For product and ordering information: World Wide Web: httpl /www.usgs.gov /pubprod Telephone: 1- 888 - ASK -USGS For more information on the USGS —the Federal source for science aboutthe Earth, its natural and living resources, natural hazards, and the environment World Wide Web: httpl /www.usgs.gov Telephone: 1- 888- ASK -USGS Suggested citation: 2007 Working Group on California Earthquake Probabilities, 2008, The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2): U.S. Geological Survey Open -File Report 2007 -1437 and California Geological Survey Special Report 203 [ http : / /pubs.usgs.gov /of/2007/1091h. Any use of trade, product, orfirm names is for descriptive purposes only and does not imply endorsement by the U.S. Government Although this report is in the public domain, permission must be secured from the individual copyright owners to reproduce any copyrighted material contained within this report • Executive Committee (ExCom) Edward A. Field (Chair) Tom Parsons Mark D. Petersen Ross S. Stein Ray J. Weldon II Chris J. Wills Scientific Review Panel (SRP) William L. Ellsworth (Chair) Michael L. Blanpied Lloyd C. Cluff C. Allin Cornell Arthur D. Frankel David D. Jackson David P. Schwartz Steve G. Wesnousky U. S. Geological Survey, Pasadena, Calif: U. S. Geological Survey, Menlo Park, Calif: U. S. Geological Survey, Golden, Colo. U. S. Geological Survey, Menlo Park, Calif: SCEC, University of Oregon, Eugene, Oreg. California Geological Survey U. S. Geological Survey, Menlo Park, Calif. U. S. Geological Survey, Reston, Virg. Pacific Gas and Electric Company Stanford University, Stanford, Calif. U. S. Geological Survey, Golden, Colo. University of California, Los Angeles, Los Angeles, Calif. U. S. Geological Survey, Menlo Park, Calif: University of Nevada, Reno, Reno, Nev. Management Oversight Committee (MOC) Thomas K Jordan (Chair) SCEC Director, University of southern Califomia • Rufus D. Catchings U. S. Geological Survey, Menlo Park, Calif: Jill R. McCarthy U. S. Geological Survey, Golden, Colo. Michael S. Reichle California Geological Survey -i iii Authors and Co-authors of Appendices Glenn P. Biasi William A. Bryant Tianqing Cao Timothy E. Dawson Karen R. Felzer Edward R Field Arthur D. Frankel Thomas E. Fumal Vipin Gupta Katherine J. Kendrick Charles S. Mueller Tom Parsons Mark D. Petersen David A. Schmidt Ross S. Stein Tom K. Rockwell Ray J. Weldon 11 - Chris J. Wills Beth A Wisely Yuehua Zeng University of Nevada, Reno, Reno, Nov. California Geological Survey California Geological Survey U. S. Geological Survey, Menlo Park, Calif U. S. Geological Survey, Pasadena, Calif. U. S. Geological Survey, Pasadena, Calif. U. S. Geological Survey, Golden, Colo. U. S. Geological Survey, Menlo Park, Calif SCEG University of Southern California U. S. Geological Survey, Pasadena, Calif U. S. Geological Survey, Golden, Colo. U. S. Geological Survey, Menlo Park, Calif. U. S. Geological Survey, Golden, Colo. University of Oregon, Eugene, Oreg. U. S. Geological Survey, Menlo Park, Calif. San Diego State University, San Diego, Calif. •SCEC/University of Oregon, Eugene, Oreg. California Geological Survey University of Oregon, Eugene, Oreg. U. S. Geological Survey, Golden, Colo. iv 0 UCERF 2 Report Contents Listof Appendices ........................................................................................................ ............................... A List of Supplementary Materials .. ............................... .. ExecutiveSummary ........................................................................................................ ..............................1 A. Introduction ............................................................................................................... ............................... 9 A.1. Motivation and Structure of WGCEP ( 2007) ..................................................... .............................11 A.2. Key Differences and Updates from Previous WGCEP Models ......................... .............................12 A.3. Modeling Framework ......................................................................................... .............................13 A.4. Representation of Epistemic Uncertainty ........................................................... .............................14 A.S. Review and Consensus - Building Processes ....................................................... .............................16 A.6. Key Products ...................................................................................................... .............................18 A.7. Report Organization ........................................................................................... .............................19 B. Fault Models ............................................................................................................ ............................... 20 B.1. Fault Section Database ..................................................................................... ............................... 20 B.2. Alternative Fault Models .................................................................................. ............................... 22 C. Deformation Models ............................................................................................... ............................... 24 C.1. Preferred and Alternative Deformation Models ............................................... ............................... 24 C.2. Aseismic Slip Factors ....................................................................................... ............................... 27 C.3. Crustal Shear Zones ......................... ............................... . ...... ............................... 29 ......... C.4. Path - Integral Tests of Deformation Models ..... ............................... . 30 ....... ............................... D. Earthquake Rate Models.. .................................................................................................... .. 33 ................ D.1. Earthquake Source Types ................................................................................. ............................... 33 • D.2. Reduction of Moment Rate on Faults ............................................................... ............................... 34 D.3. Type -A Source Rate Models ............................................................................ ............................... 35 DA. Type -B Source Rate Models ............................................................................ ............................... 41 D.S. Type -C Source Rate Models ............................................................................ ............................... 42 D.6. Background Seismicity Rate Models ............................................................... ............................... 43 D.7. The Cascadia Subduction Zone ........................................................................ ............................... 43 D. 8. Tests of the Earthquake Rate Models ............................................................... ............................... 46 E. Probability Models .................................................... ............................... ............ 59 ... ............................... E.1. WGCEP (2003) Probability Models ................................................................. ............................... 59 E.2. Selection of Probability Models ........................................................................ ............................... 61 E.3. Statewide Empirical Model ............................................................................... ............................... 62 E.4. Cascadia Probability Model ........................................... ............................... ............ 64 ....................... F. Results of Probability Calculations .......................................................................... ............................... 65 F.1. Probability of What? ......................................................................................... ............................... 65 F.2. Regional Probabilities ....................................................................................... ............................... 66 F.3. Probabilities for Faults ...................................................................................... ............................... 71 F.4. Probability Sensitivity Analysis ......................................... ............................... ........................ 80 G. Discussion ............................................................................................................... ............................... 84 G.1. Model Limitations and Opportunities for Future Improvements ..................... ............................... 84 G.2. Accomplishments and Key Differences From Previous Studies ...................... ............................... 87 Acknowledgments........................................................................................................ ............................... 89 References.................................................................................................................... ............................... 90 Listof Acronyms ......................................................................................................... ............................... 96 V UCERF 2 Report List of Appendices A. California Fault Parameters for the National Seismic Hazard Maps and Working Group on California Earthquake Probabilities 2007, by C. J. Wills, R. J. Weldon H,, and W. A. Bryant B. Recurrence Interval and Event Age Data for Type A Faults, by T. E. Dawson, R. J. Weldon H,, and G. P. Biasi C. Monte Carlo Method for Determining Earthquake Recurrence Parameters from Short Paleoseismic Catalogs: Example Calculations for California, by T. Parsons D. Magnitude -Area Relationships, by R.S. Stein E. Overview of the Southern San Andreas Fault Model, by R J. Weldon H,, G. P. Biasi, C. J. Wills, and T.E. Dawson F. Summary of Geologic Data and Developments of A Priori Rupture Models for the Elsinore, San Jacinto, and Garlock Faults, by T. E. Dawson, T. K Rockwell, R J. Weldon A and C. J. Wills G. Development of Final A -Fault Rupture Models for WGCEP/NSHMP Earthquake Rate Model 2.3, by E. H. Field R J. Weldon H,, Y. Gupta, T. Parsons, W. J. Wills, T. E. Dawson, R. S. Stein, and M. D. Petersen H. WGCEP Historical California Earthquake Catalog, by K R Felker and T. Cao I. Calculating California Seismicity Rates, by K R. Felker J. Spatial Seismicity Rates and Maximum Magnitudes for Background, by M. D. Petersen, C. S Mueller, A. D. Frankel, and Y. Zeng K. A- Priori Rupture Models for Northern California Type -A Faults, by C. J. Wills, R. J. Weldon H,, and E. H. Field 0 L. Cascadia Subduction Zone, by A. D. Frankel and M.D. Petersen M. Empirical Estimation of Regional Time Variation in Seismicity, by K R. Felker N. Conditional, Time - Dependent Probabilities for Segmented Type -A Faults in the WGCEP UCERF 2, by E. H. Field and Y. Gupta O. Paleoseismic Investigations of the Walnut Site on the San Jacinto Fault, by T. E. Fumal and K J. Kendrick P. Compilation of Surface Creep on California Faults and Comparison of WGCEP 2007 Deformation Model to Pacific -North America Plate Motion, by B. A. Wisely, D. A. Schmidt, and R. J. Weldon H vi UCERF 2 Report • List of Supplementary Materials Supplementary Sheet 1: Sheet 2: Sheet 3: Sheet 4: Sheet 5: Sheet 6: Sheet 7: Sheet 8: Sheet 9: Sheet 10: Sheet 11: Sheet 12: Sheet 13: Excel Spreadsheet Rupture Magnitudes & Rates Total 30 -year Rupture Probabilities Total 5 -year Rupture Probabilities 1vr2:6.7 30 -year Rupture Probabilities Segment Rates & Mean Recurrence Intervals Total 30 -year Segment Probabilities Total 5 -year Segment Probabilities XV 6.7 30 -year Segment Probabilities 1VT2!6.7 Probabilities in WGCEP (2003: Box Total 30 -year Probabilities on Faults M? [7 UCERF 2 Report Executive Summary 11 The Uniform California Earthquake Rupture Forecast, Version 2 (UCERF 2) 2007 Working Group on California Earthquake Probabilities (WGCEP) and the USGS National Seismic Hazard Mapping Program (NSHMP) Executive Summary California's 35 million people live among some of the most active earthquake faults in the United States. Public safety demands credible assessments of the earthquake hazard to maintain appropriate building codes for safe construction and earthquake insurance for loss protection. Seismic hazard analysis begins with an earthquake rupture forecast —a model of probabilities that earthquakes of specified magnitudes, locations, and faulting types will occur during a specified time interval. This report describes a new earthquake rupture forecast for California developed by the 2007 Working Group on California Earthquake Probabilities (WGCEP 2007). 2007 Working Group on California Earthquake Probabilities WGCEP 2007 was organized in September, 2005, by the U. S. Geological Survey (USGS), the California Geological Survey (CGS), and the Southern California Earthquake Center (SCEC). It was charged with two tasks: (1) collaborate with the National Seismic Hazard Mapping Program (NSHMP) in producing a revised, time - independent forecast for California as input to the 2007 revisions of the national seismic hazard maps, 'and (2) create a uniform, statewide, time - dependent model that among other purposes, Colo.uld be used by the California Earthquake Authority (CEA) in setting earthquake insurance rates. The national seismic hazard maps utilize a time- independent forecast in which the probability of each earthquake rupture is completely independent of the timing of all others. Time - dependent models are based on the concept of stress renewal: the probability of a fault rupture drops immediately after a large earthquake releases tectonic stress on the fault and rises again as the stress is regenerated by continuous tectonic loading. However, observations in California and elsewhere show that the earthquake cycle associated with this elastic rebound theory can be highly irregular, owing, for example, to stress interactions among neighboring faults. We do not understand these interactions well enough to model them explicitly; therefore, variations in the earthquake cycle must be calibrated empirically using historical observations of seismicity and geologic data on the dazes and sizes of prehistoric earthquakes (paleoseismology). Time - dependent earthquake rupture forecasts, in which the probabilities of future events are conditioned on the dates of previous earthquakes, have been the focus of five previous Working Groups on California Earthquake Probabilities (WGCEP 1988, 1990, 1995 & 2003). Each of 2 1 Executive Summary these working groups has expanded on its predecessors, improving the data and forecasting methodology, and each has drawn on input from broad cross- sections of the earth science community. Building on this experience, we calculate time- dependent probabilities of large earthquakes on major faults (generally those with the highest rates of slip) where the requisite information is available: the expected mean frequency of earthquakes and the elapsed time since the last earthquake. Where such information is lacking, we use time- independent probabilities, which require only an estimate of earthquake frequency. The WGCEP 2007 study differs from previous WGCEP efforts by: • reporting earthquake probability for the entire state of California instead of subregions; • using uniform methodology across all regions; • using the same earthquake rate model as the 2007 National Seismic Hazard Map Program; • compiling and using updated, uniform, and publicly accessible statewide data; • developing new methods to make models more rigorously adherent to observational data, particularly fault slip rates (moment balanced); • making analysis tools and data available through a readily accessible web -based interface. In general, we have adopted the results from previous working groups where justified and have updated the model only when compelled to by new information or understanding, or by necessity to conform. the analysis to a uniform statewide approach and with the NSHMP assessment. UCERF 2 Report Review and Consensus - Building Processes All UCERF 2 model elements and WGCEP 2007 documents were reviewed by an internal Scientific Review Panel (SRP) comprising experts who were not WGCEP 2007 members. The SRP reported to the Management Oversight Committee (MOC), which coordinated the review and oversaw consensus - building processes. External oversight and review was provided by the National Earthquake Prediction Evaluation Council ( NEPEC) and the California Earthquake Prediction Evaluation Council ( CEPEC), as well as CEA's Multidisciplinary Research Team. CEPEC and NEPEC tracked model development throughout the WGCEP 2007 process and reviewed the final report Advice and comment from the scientific and engineering communities was sought regularly through open meetings and workshops during the several phases of UCERF development. Participants included experts from academia, Private and corporate providers of hazard assessments, Colo.nsulting companies, and government agencies. WGCEP progress was reported at major scientific gatherings such as annual meetings of the American Geophysical Union, the Seismological Society of America, and the Southern California Earthquake Center. Model Framework We have built on previous WGCEP and NSHMP efforts to quantify regional earthquake probabilities in California, using the best available science to develop a new framework for a Uniform California Earthquake Rupture Forecast ( UCERF). The UCERF framework comprises a sequence of four model types: a fault model that gives the physical geometry of the larger, known faults; a deformation model that gives slip rates and • C: UCERF 2 Report aseismicity factors to each fault section; an earthquake rate model that gives the long -term rate of all earthquakes of magnitude five or greater (M 2: 5) throughout the region; and a probability model that gives a probability of occurrence for each earthquake during a specified (future) time interval. This report presents the latest versions of each of these models, including the statewide time- independent earthquake rate model incorporated into the 2007 revisions to the national seismic hazard map (ERM 2.3) and the time - dependent earthquake probability model derived from ERM 2.3 ( UCERF 2). The results are intended for use in forecasting the intensity of ground shaking throughout California. The model incorporates both aleatory uncertainties (arising from natural variability) and epistemic uncertainties (resulting from lack of knowledge). The latter were included by constructing a logic tree with branches representing viable alternative hypotheses. We restricted our consideration to data and methods that have been published, or accepted for publication, in peer - reviewed scientific journals or as U.S. Geological Survey Open File Reports. If relevant published models differed significantly, we applied logic -tree weighting to represent the alternatives. Generally, two alternatives were given equal weight in the absence of any clear evidence to favor one over the other. When there was evidence to favor a given branch, the assignment of relative weights was made though a consensus - building process, which we describe for each case. Earthquake Rate Model The WGCEP 2007 earthquake rate model features a new fault geometry with more accurate values of dip and seismogenic depth, and new compilations of fault slip rates and paleoseismic events. The final version, ERM 2.3, includes two Executive Summary 13 alternative fault models for southern California thrust -fault geometry and three alternatives representing the uncertain slip distribution between the southern San Andreas and San Jacinto faults. A significant logic-tree branching involves the choice of the magnitude -area relationship, which is used to translate from fault slip rates to earthquake rates; the global database of rupture areas and magnitude determinations has significant spread, leaving room for alternative interpretations. Another important model branching incorporates alterative representations of the earthquake rates on major faults. We compiled an a priori earthquake rate model derived by a community consensus of paleoseismic and other geologic . observations. We also _calculated a moment - balanced version of the model, which modifies the earthquake rate to match the observed Iong -term slip -rate data; the resulting rates were constrained to fall within the ranges derived from paleoseismic observations. These two models balance a consensus of geologic and seismologic expert opinion with strict adherence to specific observational data. We tested ERM 2.3 in three different ways: by comparing the predicted magnitude- frequency distributions of earthquakes with a unified historic and instrumental earthquake catalog for California and surrounding regions, by comparing integrating measures of deformation across the plate - boundary zone with the plate rate, and by comparing the distribution of source types in the model with historical data. A maj or issue was overprediction of the rate of M >_ 6.5 earthquakes, known informally as "the bulge ", a problem common to previous WGCEP and NSHNO studies. ERM 2.3 predicts an annual rate for M?! 6.5 earthquakes of 0.32 events/yr, which exceeds the historically observed rate of 0.24 eventslyr by about a third, though it lies within the 4 1 Executive Summary 95% confidence bounds on the observed rate (0.13 -0.35 events/yr). In comparison, the NSHNT 2002 model for California exceeded the observed rate by a factor of two. Time - Dependent Earthquake Probability Model We tightly coordinated the development of the earthquake rate models for California with NSHU P, so that both the 2007 revisions of the national seismic hazard maps and UCERF 2 are based on ERM 2.3. Constructing an earthquake rupture forecast from ERM 2.3 required a probability model that specifies how events are distributed in time, and here we departed from the NSENT 2007 conventions by considering, along with a time - independent (Poisson) forecast, time - dependent forecasts that use stress - renewal UCERF 2 Report assumptions to condition the event probabilities for the most active faults on the date of their last major rupture. Our choice of UCERF 2 model branches was based on a careful review of all available probability models. A particularly influential branching is the "empirical" probability model, which includes a geographically variable estimate of California earthquake rate changes observed during the last 150 years. We lack consensus on the underlying physics that causes broad earthquake rate changes, though there is much promising research involving fault interactions. Rather than applying complex physical models to adjust probability, WGCEP 2007 relies on the simpler empirically -based correction. An important seismic hazard for California is the Cascadia subduction zone, which extends Participation Probabilities 0.00196 0.6196 0.196 146 1096 100% Figure A. Participation probability maps, displaying the mean UCERF 2 probabilities that an individual 0.1° x 0.1° cell in the statewide grid will be involved in a fault rupture of any source type above the specified magnitude threshold during the next 30 years. The magnitude thresholds shown here are M• 5.0, 6.7, ' and 7.7. Probability color scale is logarithmic; i.e. each decrement unit represents a 10 -fold decrease in probability. �U_ 0 A UCERF 2 Report about 1200 km from Vancouver Island in British Columbia to Cape Mendocino in California and is capable of generating an earthquake of M 9 or larger. Because this fault lies mostly outside the state, we treated it as a special case with its own logic tree, which included two rupture scenarios: (1) M 8.8 -9.2 events that rupture the entire Casca.dia subduction zone every 500 years on average, and (2) M 8.0 -8.7 events whose ruptures cover the entire zone over a period of about 500 years. A time- independent model was applied to the M 8.0 -8.7 scenario, and a time- dependent model to the M 8.8 -9.2 scenario. In computing event probabilities, the branches were weighted by expert opinion gathered in open workshops. The UCERF 2 model has been implemented in - a modular (object - oriented), extensible framework using the OpenSHA platform, so that experiments with alternative branch weights can be easily investigated and future updates can be quickly accommodated as new data and methods emerge. The final UCERF 2 logic tree incorporated 480 branches that received nonzero weight, each of which produces a separate set of probabilities for all earthquakes in California. We take the mean and spread of these results to represent the best estimate of earthquake probability and its sensitivity to parameter uncertainty. Results of Probability Calculations According to UCERF 2, a M >_ 6.7 earthquake is virtually assured in California during the next 30 years (99.7% probability of occurrence). Larger events are less likely: the mean 30 -year UCERF 2 estimate gives a 94% chance of a M > 7.0 earthquake, a 46% chance of a M > 7.5 shock, and 4.5% chance of a M >_ 8.0 event. The UCERF 2 range for these latter probabilities is 85 -99 %, 29- 65%, "and 0 -11 %, respectively. In addition, we estimate a 10% probability of a M,2 8.0 Executive Summary 15 earthquake somewhere along the Cascadia subduction zone (perhaps far from California) in the next 30 years. We emphasize that the probabilities calculated for the largest magnitude events should be used with caution, because they depend critically on rupture scenarios that involve fault lengths longer than historically observed ruptures, as well as an extrapolation of scaling relationships, such as the magnitude -area relationships, beyond the limits of the empirical data. Dividing the state into two approximately equal areas, we find the 30 -year probability of a large earthquake to be higher in the southern half: a M >_ 6.7 earthquake has a 97% chance of occurring in southern California in 30- years, Colo.mpared to a 93% probability in northern California, and the odds for a M >_ 7.5 event are doubled (37% vs. 15 %). In addition to state -wide and regional estimates, our report gives probabilities for individual faults and fault segments throughout the state, as well as a geographically variable background rate. The UCERF 2 earthquake rupture forecast can be visualized by mapping the mean probability that an element of area on a statewide grid will include a fault rupture of any source type above a specified magnitude threshold during the next 30 years. Figure A presents these "participation probability" maps for three magnitude thresholds. For events with M >_ 5.0, the areas where the participation probabilities exceed 1% (yellow or warmer in color) include over half the state, reflecting the widespread distribution of California seismicity, much of which is represented in the model as "background." At M>_ 6.7, this same probability level is confined to the major faults, and at M >_ 7.7, it is generally restricted to the longer strike -slip strands of the San Andreas fault system. 6 1 Executive Summary UCERF Z Report Table A. 30 -year probability of M - 6.7 events on the Type-A faults, rounded to the nearest percent. Table A summarizes the mean probabilities for M >_ 6.7 events on the principal strike -slip faults of California, which accommodate most of the motion between the North America and Pacific plates, and it compares our results with those of WGCEP 1995 - -for - southern --California -and WGCEP 2003 for the Bay Area. The most dangerous fault is the southern part of the San Andreas, which has a 59% probability of generating a M >_ 6.7 earthquake in the next 30 years. This compares with 21% for the northern San Andreas fault. We have enough data to calculate time - dependent earthquake probability on the principal strike -slip faults in Table A. These faults exist within a web of faults with lower slip rates that we know less about, which are consequently treated as time- independent sources. In southern California, the contribution to overall regional probability from these lower slip -rate faults, which include the reverse faults of the Transverse Ranges, exceeds that of the principal strike -slip faults. Reliability of Results The larger the area considered and the longer the time considered generally makes a probability calculation more reliable. Thus the statewide 30- year probability values are more reliable estimates than those for individual faults. However, even the most reliable of our calculations are subject to considerable sensitivity to parameters. For example, across the 480 branches of the logic tree we find a minimum 30 -year probability of 29% for a M?: 7.5 earthquake, and a maximum of 65 %. Calculations are quite sensitive to - parameter choices on individual faults; while the mean calculated probability on the southern San Andreas fault is 59 %, we find that the value could reasonably be anywhere between 22% and 94% (see Table A). There are known limitations with the WGCEP 2007 model, which are discussed in detail in the main report. More research time will bring improvements in key topical are For example, new earthquake faults will continue to be discovered. Improvements in our methods for determining maximum magnitudes associated with poorly understood faults are needed. A related major research challenge involves improving our ability to forecast more complex earthquake ruptures that include fault jumps, branching, and segment - breaking ruptures. Comparisons with Previous Studies The 30 -year probability of a M2:6.7 earthquake striking the San Francisco Bay Area is 63% for UCERF 2, which is indistinguishable from the 62% value reported by WGCEP 2003 (P ,• UCERF 2 Report (see Table A). Moreover, the extrema calculated from all of the UCERF 2 branches [0.41 -0.84] approximate the 95% confidence interval of WGCEP 2003 results for the aggregate Bay Area probabilities [0.38 - 0.85]. This agreement indicates that we succeeded in capturing the most important epistemic uncertainties (in part because we were guided by the comprehensive uncertainty analysis of the WGCEP 2003 report). As shown in the table, there are differences between WGCEP 2007 and WGCEP 2003 calculations for individual fault probabilities in the Bay Area. However, none exceed the uncertainty ranges reported by either working group. The differences resulted primarily from inclusion of paleoseismic observations in UCERF 2 and the restricted inventory -of. probability models that could be used for our statewide analysis. The differences in the mean 30 -year probabilities for M >_ 6.7 events between the 1995 and 2007 studies are more significant. The most important arise from new paleoseismic data and analysis, new geodetic data, and an earthquake rate model that allows a greater variety of rupture sizes on faults. One important change is to the San Jacinto fault, where the probability has been halved from 61%, reported by WGCEP 1995, to 31% [14 0/6-54 %] calculated by WGCEP 2007 (see table). Similarly, Elsinore fault probability is halved from 24% to 11% [5916-25 %] because of the increased array of possible earthquake magnitudes allowed in the model. Recommendations The comprehensive nature of the UCERF 2 analysis has identified many opportunities for future model improvements, and we outline in the report specific recommendations for further research. Examples include the relaxation of fault segmentation and the inclusion of fault -to -fault ruptures, which may be in part responsible for the Executive Summary 17 "bulge" problem; the inclusion of earthquake triggering and clustering, as manifested in aftershock sequences; and improved magnitude- area relationships.