1 Can earthquakes be predicted? Karen Felzer U.S. Geological Survey
2 Earthquake predictions that most seismologists agree with
3 Long term earthquake probabilities These kinds of predictions are important for construction codes, insurance rates, and emergency preparedness, but they are a long time to duck and cover! 2008 Working Group on California Earthquake Probabilities >99% chance that a M 6.7 earthquake will occur in CA within 30 years.
4 Aftershock Probabilities Seismologists assign probabilities to how large the aftershocks will be, including the probability that an aftershock will be larger than its mainshock. STEP probability map
5 Whether any more precise prediction is possible has long been debated Since my first attachment to seismology I have had a horror of predictions and predictors. Journalists and the general public rush to any suggestion of earthquake prediction like hogs to a full trough -Charles Richter, 1977 I can tell you scientists who will bet good money that no one will ever be able to [predict earthquakes]. I m not one of them. -Tom Jordan, 2006 Director of the Southern California Earthquake Center
6 Office Survey, USGS Pasadena 50% think prediction is possible given unlimited time and resources
7 Fraction of articles in Bull. Seis. Soc. Am. with earthquake prediction keyword
8 Two types of predictions commonly attempted: Intermediate term: Narrow region and magnitude, time range 5-30 years. Short term: Narrow region and magnitude, time given to the day or month.
9 Intermediate term prediction
10 The seismic gap model stress quake quake quake quake quake time Stress is low after a large earthquake and must rebuild before another large quake can occur
11 The Seismic Gap Model Earthquakes occur periodically or quasiperiodically. A fault that has just ruptured is safe. A fault that has not ruptured for some time is a gap that will be filled soon.
12 Bakun and Lindh (1985) used the model to predict a Parkfield eq Earthquake supposed to occur with 95% certainty Figure from Bakun and Lindh, 1985 M 6 earthquakes predicted at Parkfield every 22 years Like Old Faithful!
13 Quotes from the Media Scientists say the Hayward Fault is overdue for a major earthquake -ABC News Geologist uncovers earth s secrets, says Southern California is overdue for a major earthquake -University News The [San Andreas Fault] is 10 months pregnant! -guardian.co.uk
14 Short term prediction
15 Short term predictions are generally based on anomalies or patterns Photo taken by Shou in Pasadena, 1997 Do lines in the clouds predict earthquakes?
16 Short term prediction publications Accelerating seismicity and stress release before large earthquakes, Bowman et al., 2001 Precursory seismic quiescence, Wyss and Habermann, 1987 Ground water radon anomaly before the Kobe earthquake in Japan [radon observed to increase] Igarashi et al A mechanism for anomalous decline in radon precursory to an earthquake, Kuo et al A drop in mean earthquake magnitude before large earthquakes found by Smith, 1986 An increase in mean earthquake magnitude before large earthquakes found by Nuannin et al., 2007
17 So -- Has anyone made successful predictions?
18 Global seismic gap predictions of Nishenko not fulfilled Random Earthquake model observed quakes Seismic Gap Prediction Figure 1 From Rong et al. (2003)
19 A M 6 at Parkfield did occur -- but not until 2004! add on: What was the probability of occurrence in 2004 And earthquakes may not have been so periodic there to begin with!
20 What s wrong with the seismic gap model?? Earthquake stress drop << fault stress, so stress is always high. Rupture occurs not because of slow stress increase but because of rapid shaking-induced strength decrease. stress quake quake quake time quake quake
21 What s wrong with the seismic gap model?? And/Or: There are so many subparallel faults strands in the crust that some are always ready for failure. Section of the San Andreas, from SCEC website
22 No short term schemes has been verified to date Accelerating seismicity and stress release before large earthquakes, Bowman et al., 2001 Precursory seismic quiescence, Wyss and Habermann, 1987 Ground water radon anomaly before the Kobe earthquake in Japan [radon observed to increase] Igarashi et al A mechanism for anomalous decline in radon precursory to an earthquake, Kuo et al A drop in mean earthquake magnitude before large earthquakes found by Smith, 1986 An increase in mean earthquake magnitude before large earthquakes found by Nuannin et al., 2007
23 Reporter to Richter: Did anyone predict this earthquake? Richter: Not yet!
24 Summary results of serious predictions for SoCal Prediction Predictor Result Major SoCal earthquake M 6.0 at Parkfield, Palmdale Bulge - large San Andreas earthquake, 1970s M 6.4, Jan - Sept. 2005, Eastern Southern California Bailey Willis, Stanford University W. H. Bakun and A. G. Lindh, USGS Menlo Park Robert Castle, USGS Menlo Park Vladimir Keilis-Borok, UCLA no earthquake no earthquake no earthquake no earthquake
25 Is earthquake prediction just really difficult or is it impossible??
26 Why I think earthquake prediction is impossible
27 All earthquakes start at small points, called hypocenters, and propagate over the fault plane at the same velocity fault plane hypocenter small quake Big Quake! Small earthquakes stop after a small rupture; large earthquakes continue to rupture a large area
28 This observation, and many others, inspired the Cascade Model hypocenter Mainshock fault plane Unit earthquake Propagation is stochastic Points have a constant prob. of rupturing if stressed Magnitude is undetermined until earthquake is over
29 Cellular automata approximation of cascade model: 20 random simulations Animation by Mike Harrington Each point has a 75% chance of triggering a neighbor
30 The cascade model is consistent with empirical earthquake statistical laws
31 Magnitude-Frequency stats produced by cascade model: inverse power law
32 Magnitude-frequency statistics seen in data: Inverse power law Global CMT catalog The Ishimoto-Iida/ Gutenberg-Richter law N(A) ~ 1/A
33 In the cascade model the magnitude of each earthquake should be random
34 Test for magnitude randomness Take magnitudes, above the completeness threshold, from the California and Global catalog. Take the difference between consecutive magnitudes. Compare to the differences between a randomly generated set of magnitudes.
35 A Kolmogorov-Smirnoff test indicates that each magnitude is chosen randomly
36 The cascade model is supported by the similar starts of small & large earthquakes seconds from origin The 1st second of two quakes. Which will be larger?
37 10 seconds of record, same earthquakes Magnitude 2 Magnitude 3 seconds from origin
38 Cascade Model Summary In the cascade model earthquake growth is stochastic and self similar. Earthquakes of any magnitude may be generated from the same starting conditions. Earthquake magnitude evolves as the earthquake grows => cannot be predicted beforehand.
39 Conclusions Average earthquake rates and increased rates during aftershock sequences can be empirically forecast. More precise earthquake prediction requires predetermined earthquake magnitude. Observation indicates that earthquake magnitude is random, and that all quakes start the same. => Earthquake prediction is impossible.
40 And some parting thoughts from a USGS interview with Charles Richter Q: What are your thoughts on the possibility of predicting earthquakes in the next two decades. A: None. Q: How can the study of United States earthquakes be improved? A: By continuous earnest efforts to find out what is going on, without running after prediction. Q: If the building you are in now started to shake and you knew an earthquake was occurring, what would you do? A: I would walk - not run - to the nearest seismograph.
41 Acknowledgements Alan Felzer, my dad, for years of high quality editing, commentary, and incredible support. Mike and Emmet Harrington, husband & 12 mo. old son, for giving me the gift of time to write this talk! Laura Felzer and Andee Harrington, mom and mom-in-law extraordinaire, for taking care of everyone!
42 The difference in time between consecutive earthquakes does not vary with magnitude range California earthquakes,
43 Even if magnitude is completely random a sudden increase in the seismicity rate would increase the odds of a large earthquake. But we rarely see larger rate increases than during aftershock sequences => aftershock forecasts are the best predicting we can do!
44 Important! -- if the magnitude of each earthquake is truly random there is no way to predict earthquake magnitude unless there is some signal right before/after the earthquake starts. But it has never been demonstrated that earthquakes of different magnitudes show any differentiation until the small one stops and large one keeps going.
Clinical Trials: What You Need to Know Clinical trials are studies in which people volunteer to test new drugs or devices. Doctors use clinical trials to learn whether a new treatment works and is safe
Climate Surveys: Useful Tools to Help Colleges and Universities in Their Efforts to Reduce and Prevent Sexual Assault Why are we releasing information about climate surveys? Sexual assault is a significant
Problem 3 If A is divided by B the result is 2/3. If B is divided by C the result is 4/7. What is the result if A is divided by C? Suggested Questions to ask students about Problem 3 The key to this question
How to Study Mathematics Written by Paul Dawkins Before I get into the tips for how to study math let me first say that everyone studies differently and there is no one right way to study for a math class.
Is Connectivity A Human Right? For almost ten years, Facebook has been on a mission to make the world more open and connected. For us, that means the entire world not just the richest, most developed countries.
Critical analysis Be more critical! More analysis needed! That s what my tutors say about my essays. I m not really sure what they mean. I thought I had written a really good assignment this time. I did
Do-it-Yourself Recovery of Unpaid Wages How to Represent Yourself Before the California Labor Commissioner A Publication of: The Legal Aid Society-Employment Law Center 600 Harrison Street, Suite 120 San
MOST DIFFICULT INTERVIEW QUESTIONS When preparing for an interview, go through the following list of questions and write down appropriate responses. Don t memorize your answers; review them for content
THE CARNEGIE FOUNDATION FOR THE ADVANCEMENT OF TEACHING Problem Solution Exploration Papers WHAT COMMUNITY COLLEGE DEVELOPMENTAL MATHEMATICS STUDENTS UNDERSTAND ABOUT MATHEMATICS James W. Stigler, Karen
A Self-Directed Guide to Designing Courses for Significant Learning L. Dee Fink, PhD Director, Instructional Development Program University of Oklahoma Author of: Creating Significant Learning Experiences:
USE-CASE 2.0 The Guide to Succeeding with Use Cases Ivar Jacobson Ian Spence Kurt Bittner December 2011 USE-CASE 2.0 The Definitive Guide About this Guide 3 How to read this Guide 3 What is Use-Case 2.0?
The Impact of Maine s One-to-One Laptop Program on Middle School Teachers and Students Phase One Summary Evidence Research Report #1 Report prepared by David L. Silvernail Director Dawn M. M. Lane Research
Students Perceptions of Their Community College Experiences Students Perceptions of Their Community College Experiences by Thad Nodine, Laura Jaeger, Andrea Venezia, and Kathy Reeves Bracco, with research
Who Sends the Most Text Messages? Written by: Anna Bargagliotti and Jeanie Gibson (for Project-SET) Loyola Marymount University and Hutchison School firstname.lastname@example.org, email@example.com, www.project-set.com
Equal Access Participatory Monitoring and Evaluation toolkit The Most Significant Change technique A manual for M&E staff and others at Equal Access Developed by June Lennie February 2011 Contents Acknowledgements
The Statistician (1999) 48, Part 1, pp. 1±40 Some statistical heresies J. K. Lindsey Limburgs Universitair Centrum, Diepenbeek, Belgium [Read before The Royal Statistical Society on Wednesday, July 15th,
CENTERS FOR MEDICARE & MEDICAID SERVICES 2015 Choosing a Medigap Policy: A Guide to Health Insurance for People with Medicare This official government guide has important information about: Medicare Supplement
Chapter 4: MEASURING PROGRESS An indicator is something that helps you understand where you are, which way you are going, and how far you are from where you want to be. An effective indicator alerts you
Chapter 11 Markov Chains 11.1 Introduction Most of our study of probability has dealt with independent trials processes. These processes are the basis of classical probability theory and much of statistics.
Sawtooth Software RESEARCH PAPER SERIES How Many Questions Should You Ask in Choice-Based Conjoint Studies? Richard M. Johnson and Bryan K. Orme, Sawtooth Software, Inc. 1996 Copyright 1996-2002, Sawtooth
Are You Financially Fit? A workbook to help you become and stay ly fit for life. Courtesy: Andy Newson / FreeDigitalPhotos.net Florida State University Learning Systems Institute What is Financial Fitness?
THE RECENT SLOWDOWN IN THE COST OF DELAYING ACTION TO STEM CLIMATE CHANGE The Council of Economic Advisers July 2014 Executive Summary The signs of climate change are all around us. The average temperature
Perfect For RTI Getting the Most out of STAR Math Using data to inform instruction and intervention The Accelerated products design, STAR Math, STAR Reading, STAR Early Literacy, Accelerated Math, Accelerated
karl e. case Wellesley College robert j. shiller Yale University anne k. thompson McGraw-Hill Construction What Have They Been Thinking? Homebuyer Behavior in Hot and Cold Markets ABSTRACT Questionnaire
How Do You Know It? How Can You Show It? Penny Reed Wisconsin Assistive Technology Initiative Gayl Bowser Oregon Technology Access Program Jane Korsten Responsive Centers for Psychology and Learning Wisconsin
HOW TO GET THE MOST (FOR YOURSELF) FROM TRAINING WHY TRAINING? WHY NOW? If training is not important to a particular job today, that job probably will not be important much longer, either. What makes training
A Student Handbook for Choosing and Using Assistive Technology by Gayl Bowser OREGON TECHNOLOGY ACCESS PROGRAM OTAP Penny Reed WISCONSIN ASSISTIVE TECHNOLOGY INITIATIVE WATI Hey! Can I Try That? (Bowser