How good can hazard maps be & how good do they need to be Seth Stein, Earth & Planetary Sciences, Northwestern University Jerome Stein, Applied Mathematics,

Slides:



Advertisements
Similar presentations
TOPIC 3: HOW WELL CAN WE PREDICT EARTHQUAKE HAZARDS? Predictions are important for hazard mitigation policy How much should we believe them?
Advertisements

USING EARTHQUAKE SCIENCE TO PREDICT EARTHQUAKE HAZARDS AND REDUCE EARTHQUAKE RISKS: USING WHAT WE KNOW AND RECOGNIZING WHAT WE DON’T Seth Stein Department.
Playing against nature: formulating cost- effective natural hazard policy given uncertainty Tohoku, Japan 3/2011 New Orleans 8/2005 Seth Stein, Earth &
Triggering of New Madrid Seismicity by Late Pleistocene Erosion Eric Calais & Andy Freed Purdue University Roy Van Arsdale, University of Memphis Seth.
SEISMIC HAZARD Presentation is based on: Allen, R., Earthquake hazard mitigation: New direction and opportunities, in "Treatise on Geophysics”, Bilham,
What are Earthquakes? The shaking or trembling caused by the sudden release of energy Usually associated with faulting or breaking of rocks Continuing.
WHAT COULD BE THE NEXT EARTHQUAKE DISASTER FOR JAPAN  A difficult question, but ---  It is the one that was being asked long before the March 11, 2011.
LESSONS LEARNED FROM PAST NOTABLE DISASTERS JAPAN PART 1A: EARTHQUAKES Walter Hays, Global Alliance for Disaster Reduction, Vienna, Virginia, USA.
Since New Madrid's not moving... A complex system view of midcontinental seismicity and hazards Seth Stein Northwestern Eric Calais Purdue Qingsong Li.
NEW MADRID: A dying fault? GPS seismology geology Heat flow Recent data, taken together, suggest that the New Madrid seismic zone may be shutting down.
A New Approach To Paleoseismic Event Correlation Glenn Biasi and Ray Weldon University of Nevada Reno Acknowledgments: Tom Fumal, Kate Scharer, SCEC and.
Chapter 12 - Forecasting Forecasting is important in the business decision-making process in which a current choice or decision has future implications:
8/23/2011 Washington Post Mineral, VA, earthquake illustrates seismicity of a passive-aggressive margin Seth Stein 1, Frank Pazzaglia 2, Emily Wolin 1,
NEW MADRID: A dying fault? GPS seismology geology Heat flow Recent data, taken together, suggest that the New Madrid seismic zone may be shutting down.
2011 Tōhoku Earthquake and Tsunami. MODIS satellite image on 26 FEB, before the tsunami. Scale bar is 10 km.
8: EARTHQUAKE SOURCE PARAMETERS
You must unlearn what you have learned. Alan Kafka 8. Faults switching on and off.
CONTRASTING SEISMIC RATES BETWEEN THE NEW MADRID AND WABASH VALLEY SEISMIC ZONES: STRESS TRANSFER OR AFTERSHOCKS? Miguel Merino, Seth Stein & Emile Okal.
EARTHQUAKE RECURRENCE Crucial for hazards, earthquake physics & tectonics (seismic versus aseismic deformation) Recordings of the east-west component of.
03/09/2007 Earthquake of the Week
Migrating earthquakes and faults switching on and off: a new view of intracontinental earthquakes Seth Stein Northwestern University Mian Liu University.
Time-dependent seismic hazard maps for the New Madrid seismic zone and Charleston, South Carolina areas James Hebden Seth Stein Department of Earth and.
TOPIC 2: How does the challenge of predicting hazards differ between earthquakes - at plate boundaries -In plate boundary zones -within plates?
5: EARTHQUAKES WAVEFORM MODELING S&W SOMETIMES FIRST MOTIONS DON’T CONSTRAIN FOCAL MECHANISM Especially likely when - Few nearby stations, as.
April Exploring a collaboration between Caltech and the Singaporean government Forecasting giant earthquakes of the Sumatran subduction zone and.
Seth Stein, Earth & Planetary Sciences, Northwestern University
2007 NSTA: St. Louis, Missouri Earthquake Prediction and Forecasting: A Case Study of the San Andreas and New Madrid Faults Sponsored by: IRIS (Incorporated.
China Earthquakes Tangshan, Overview Earthquake in China Tectonic overview - Causes of Earthquakes –Damage –Recovery.
Earthquakes (Chapter 8)
Ch 8.1 Numerical Methods: The Euler or Tangent Line Method
Bad assumptions or bad luck: Tohoku’s embarrassing lessons for earthquake hazard mapping What’s going wrong and what to do? Tohoku, Japan March 11, 2011.
Some Background Assumptions Markowitz Portfolio Theory
What are Earthquakes? The shaking or trembling caused by the sudden release of energy Usually associated with faulting or breaking of rocks Continuing.
Lecture 12 Statistical Inference (Estimation) Point and Interval estimation By Aziza Munir.
The global seismic energy to moment ratio: a tool for basic research and real-time identification of “Tsunami Earthquakes” Jaime Andres Convers Dr. Andrew.
The Hunting of the SNARF Giovanni F. Sella Seth Stein Northwestern University Timothy H. Dixon University of Miami "What's the good of Mercator's North.
12: Choosing Mitigation Policies "It is our choices that show what we truly are, far more than our abilities.” J. K. Rowling, Harry Potter and The Chamber.
Earthquakes (Chapter 13). Lecture Outline What is an earthquake? Seismic waves Epicenter location Earthquake magnitude Tectonic setting Hazards.
Searching for Long Duration Aftershocks in Continental Interiors Miguel Merino, Seth Stein Northwestern University.
Chapter 4 Earthquakes Map is from the United States Geological Survey and shows earthquake hazard for the fifty United States.
2. Shallow versus deep uncertainties Why many predictions / forecasts fail.
Research opportunities using IRIS and other seismic data resources John Taber, Incorporated Research Institutions for Seismology Michael Wysession, Washington.
In the past ~15 years we’ve learned a lot and have new questions: Paleoseismology shows that continental intraplate seismicity often migrates, is episodic,
Are we successfully addressing the PSHA debate? Seth Stein Earth & Planetary Sciences, Northwestern University.
Earthquake forecasting using earthquake catalogs.
Earthquake hazard isn’t a physical thing we measure. It's something mapmakers define and then use computer programs to predict. To decide how much to believe.
SEISMIC HAZARD. Seismic risk versus seismic hazard Seismic Hazard is the probability of occurrence of a specified level of ground shaking in a specified.
Karen Felzer & Emily Brodsky Testing Stress Shadows.
Bad assumptions or bad luck: Why natural hazard maps (forecasts, warnings, etc…) often fail and what to do about it Seth Stein, Northwestern University.
Yan Y. Kagan Dept. Earth and Space Sciences, UCLA, Los Angeles, CA , Evaluation.
9. As hazardous as California? USGS/FEMA: Buildings should be built to same standards How can we evaluate this argument? Frankel et al., 1996.
What to do given that earthquake hazard maps often fail
Department of Mathematics and Geosciences 1 Department of Earth & Planetary Sciences and Institute for Policy Research, Northwestern University, Evanston,
A GPS-based view of New Madrid earthquake hazard Seth Stein, Northwestern University Uncertainties permit wide range (3X) of hazard models, some higher.
Migrating earthquakes and faults switching on and off: a new view of intracontinental earthquakes Seth Stein Northwestern University Mian Liu University.
Lessons from Tohoku: why earthquake hazard maps often fail and what to do about it Tohoku, Japan March 11, 2011 M 9.1 NY Times CNN Seth Stein, Northwestern.
9. As hazardous as California? USGS/FEMA: Buildings should be built to same standards How can we evaluate this argument? Frankel et al., 1996.
You must unlearn what you have learned. Alan Kafka 8. Faults switching on and off.
1 Department of Earth & Planetary Sciences and Institute for Policy Research, Northwestern University, Evanston, Illinois, USA 2 Department of Statistics.
Understanding Earth Sixth Edition Chapter 13: EARTHQUAKES © 2011 by W. H. Freeman and Company Grotzinger Jordan.
Antonella Peresan F. Vaccari, A. Magrin, G.F. Panza, S. Cozzini B.K. Rastogi, I. Parvez Antonella Peresan F. Vaccari, A. Magrin, G.F. Panza, S. Cozzini.
Seth Stein Northwestern University Timothy Dixon University of Miami
Metrics, Bayes, and BOGSAT: Recognizing and Assessing Uncertainties in Earthquake Hazard Maps Seth Stein 1, Edward M. Brooks 1, Bruce D. Spencer 2 1 Department.
Seismic Hazard Analysis for Guam & the Northern Mariana Islands Chuck Mueller U.S. Geological Survey Golden, Colorado, USA.
Question of the Day What is a natural disaster?
What are Earthquakes? The shaking or trembling caused by the sudden release of energy Usually associated with faulting or breaking of rocks Continuing.
BREVIA Time-Variable Deformation in the New Madrid Seismic Zone Eric Calais 1 and Seth Stein 2 velocities relative to the rigid interior ofNorth Amer-
Earth & Planetary Sciences, Northwestern University
Why aren't earthquake hazard maps better. Seth Stein1, M
Understanding Earth Chapter 13: EARTHQUAKES Grotzinger • Jordan
Presentation transcript:

How good can hazard maps be & how good do they need to be Seth Stein, Earth & Planetary Sciences, Northwestern University Jerome Stein, Applied Mathematics, Brown University

How good can hazard maps be & how good do they need to be Seth Stein, Earth & Planetary Sciences, Northwestern University Jerome Stein, Applied Mathematics, Brown University What specifically are hazard maps designed to accomplish? How can we measure how well they do it? What is the best they can realistically do?

Insights from weather forecasting Different users want different things from forecasts Thus it’s important to have agreed criteria for "good" and "bad" forecasts of various types. "it is difficult to establish well-defined goals for any project designed to enhance forecasting performance without an unambiguous definition of what constitutes a good forecast." (Murphy, 1993)

Insights from weather forecasting There are fundamental limits to how accurate weather forecasts can be. If weather weren’t chaotic, every year storms would form at the same time & follow the same tracks (Lorenz, 1995) Ebeling and Stein, 2011

Insights from weather forecasting Communicating uncertainty is crucial and valuable Weather and climate forecasts should seek to to “routinely provide the nation with comprehensive, skillful, reliable, and useful information about the uncertainty of weather, water, and climate forecasts… Forecast uncertainty information can improve decisions and outcomes in various socioeconomic areas.” (Hirschberg et al., 2011)

Tohoku earthquake broke many segments 2011 Tohoku Earthquake 450 km long fault, M 9.1 (Aftershock map from USGS) J. Mori Expected Earthquake Sources 50 to 150 km segments M7.5 to 8.2 (Headquarters for Earthquake Research Promotion)

Tsunami runup approximately twice fault slip (Plafker, Okal & Synolakis 2004) M9 generates much larger tsunami Planning assumed maximum magnitude 8 Seawalls 5-10 m high CNN NYTStein & Okal, 2011

Assumed lack of M9s in record seemed consistent with model that M9s only occur where lithosphere younger than 80 Myr subducts faster than 50 mm/yr (Ruff and Kanamori, 1980) Disproved by Sumatra 2004 M9.3 and dataset reanalysis (Stein & Okal, 2007) Short record at most SZs didn’t include larger multisegment ruptures Stein & Okal, 2011

2008 Wenchuan earthquake (Mw 7.9) was not expected: map showed low hazard based on lack of recent earthquakes Didn’t use GPS data showing 1-2 mm/yr (~Wasatch) Earthquakes prior to the 2008 Wenchuan event Aftershocks of the Wenchuan event delineating the rupture zone Stein et al., 2012

Similar problems occur worldwide The earth often surprises us Do these reflect systemic problems with hazard mapping or simply low probability events (someone wins the lottery)?

Off Sanriku-oki North ~M8 0.2 to 10% Off Sanriku-oki Central ~M to 90% Off Fukushima ~M7.4 7% Off Ibaraki ~M6.7 – M7.2 90% Off Miyagi ~M7.5 > 90% Not a PSHA vs DSHA issue - issue is estimating poorly known model parameters Any map assuming no M9 would have done no better

Hazard maps are hard to get right: successfully predicting future shaking depends on accuracy of four assumptions over years Where will large earthquakes occur? When will they occur? How large will they be? How strong will their shaking be? Uncertainty & possible map failure result because these are often hard to assess, especially in plate interiors & other slowly deforming zones

GSHAP 1999 NUVEL-1 Argus, Gordon, DeMets & Stein, 1989 Swafford & Stein, 2007 Slow plate boundary Africa-Eurasia convergence rate varies smoothly (5 mm/yr)

Swafford & Stein, 2007 GSHAP 1999 NUVEL-1 Argus, Gordon, DeMets & Stein, 1989 M 6.4 M 6.3 Slow plate boundary Africa-Eurasia convergence rate varies smoothly (5 mm/yr)

Long record needed to see real hazard Swafford & Stein, M M 7.2

“Our glacial loading model suggests that earthquakes may occur anywhere along the rifted margin which has been glaciated.” Stein et al., Concentrated hazard bull's-eyes at historic earthquake sites 2005 Diffuse hazard along margin GSC Map depends greatly on assumptions & thus has large uncertainty

Plate Boundary Earthquakes Major fault loaded rapidly at constant rate Earthquakes spatially focused & temporally quasi-periodic Past is fair predictor Intraplate Earthquakes Tectonic loading collectively accommodated by a complex system of interacting faults Loading rate on a given fault is slow & may not be constant Earthquakes can cluster on a fault for a while then shift Past can be poor predictor Plate A Plate B Earthquakes at different time Stein, Liu & Wang 2009

during the period prior to the period instrumental events Earthquakes in North China Large events often pop up where there was little seismicity! Ordos Plateau Shanxi Graben Bohai Bay Beijing 1303 Hongtong M 8.0 Liu, Stein & Wang 2011 Weihi rift

during the period prior to the period instrumental events Earthquakes in North China Large events often pop up where there was little seismicity! Ordos Plateau Shanxi Graben Bohai Bay Beijing 1556 Huaxian M 8.3 Weihi rift Liu, Stein & Wang 2011

during the period prior to the period instrumental events Earthquakes in North China Large events often pop up where there was little seismicity! Ordos Plateau Shanxi Graben Bohai Bay Beijing 1668 Tancheng M 8.5 Weihi rift Liu, Stein & Wang 2011

during the period prior to the period instrumental events Earthquakes in North China Large events often pop up where there was little seismicity! Ordos Plateau Shanxi Graben Bohai Bay Beijing 1679 Sanhe M 8.0 Weihi rift Liu, Stein & Wang 2011

during the period prior to the period instrumental events Earthquakes in North China Large events often pop up where there was little seismicity! Ordos Plateau Shanxi Graben Bohai Bay Beijing 1966 Xingtai M Tangshan M Haicheng M 7.3 Weihi rift Liu, Stein & Wang 2011

No large (M>7) events ruptured the same fault segment twice in past 2000 years In past 200 years, quakes migrated from Shanxi Graben to N. China Plain Historical Instrumental Shanxi Graben Weihi rift

Aftershocks of 1976 Tangshan earthquake continue today Liu & Wang, 2012

Rate-state friction (Dieterich, 1994) predicts aftershock duration  1/loading rate Plate boundary faults quickly reloaded by steady plate motion after large earthquake Faults in continents reloaded much more slowly, so aftershocks continue much longer Stein & Liu, 2009 Often see long aftershock sequences in slowly deforming continental interiors Stein & Liu 2009

Tuttle (2009) Within plates or other low-deformation zones, present seismicity may be better indicator of locations of past events & future small events than future large ones Obermeier, (1998) Wabash: M~7 6 Kybp

Newman et al., % 275% Hazard maps involve assumptions about - Mmax of largest future events -Ground motion model -Timing of future earthquakes (time- independent or time- dependent) Since all have large uncertainties, wide range of plausible hazard models

154% %106 Hazard maps involve assumptions about - Mmax of largest future events -Ground motion model -Timing of future earthquakes (time- independent or time- dependent) Since all have large uncertainties, wide range of plausible hazard models

Stein et al, 2012 Stein et al., 2012 Uncertainty typically factor of 3-4 Often can’t be reduced much due to earthquake variability Hazard is essentially unknowable within broad range One can chose a particular value depending on preconception, but the uncertainty remains and only time will tell how good the choice was

Seismological assessment of hazard maps Various metrics could be used, e.g. compare maximum observed shaking in subregion i, x i to predicted maximum shaking p i Compute Hazard Map Error HME(p,x) =  i (x i - p i ) 2 /N and compare to error of reference map produced using a null hypothesis HME(r,x) =  i (x i - r i ) 2 /N using the skill score SS(p,r,x) = 1 - HME(p,x)/HME(r,x) Positive score if map does better than null

Some testing challenges 1)Short time record: can be worked around by aggregating regions. 2) Subjective nature of hazard mapping, resulting from need to chose faults, maximum magnitude, recurrence model, and ground motion model. This precludes the traditional method of developing a model from the first part of a time series and testing how well it does in the later part. That works if the model is "automatically" generated by some rules (e.g. least squares, etc). In the earthquake case, this can't be done easily because we know what happens in the later part of the series.

3) New maps made after a large earthquake that earlier maps missed are problem for counting statistics. Frankel et al, 2010 Before 2010 Haiti M7After 2010 Haiti M7 4X

4) Overparameterized model (overfit data): Given a trend with scatter, fitting a higher order polynomial can give a better fit to the past data but a worse fit to future data Analogously, a seismic hazard map fit to details of past earthquakes could be a worse predictor of future ones than a smoothed map How much detail is useful? Linear fit Quadratic fit

Consider map as means, not end Assess map’s success in terms of contribution to mitigation Even uncertain or poor maps may do some good Societal assessment of hazard maps

Goal: Assess the hazard in order to develop cost-effective mitigation policy Mitigation diverts resources from present needs - education, health, environment, economic growth, etc - in hope of reducing possible future losses Mitigation policies typically made - without considering uncertainties in hazard estimate (map) - without analysis of costs & benefits Challenge: how much is enough?

Societally optimal level of mitigation minimizes total cost = sum of mitigation cost + expected loss Expected loss = ∑ (loss in i th expected event x assumed probability of that event) Compared to optimum Less mitigation decreases construction costs but increases expected loss and thus total cost More mitigation gives less expected loss but higher total cost Stein & Stein, 2012 For earthquake, mitigation level is construction code Loss depends on earthquake & mitigation level Optimum

Consider marginal costs C’(n) & benefits Q’(n) (derivatives) Optimum is where marginal curves are equal, n* More mitigation costs more But reduces loss Stein & Stein, 2012 Benefit (loss reduction ) cost Less mitigation costs less But increases loss

Loss estimate scenarios based on hazard model Estimate loss as function of magnitude, ground shaking model, recurrence rate, and mitigation level This case Current mitigation fatalities ~ $100B damage Examine range of parameters & use to find optimum / php

Present Value of Future Losses Average loss over T years is L T Interest rate i PVFL = L T  t 1/(1+i) t = L T D T D T = 1/(1+i) + 1/(1+i) /(1+i) T = ((1+i) T -1 ) / (i(1+i) T ) ≈ 1/i for T large For interest rate i=0.05, D T = 15.4 for 30 years, and 19.8 for 100 years. For long enough times, the limit as T becomes infinite is D T = 1 / I, so if i = 0.05, D = 20. This is essentially the same as the value for 100 years.

Even without uncertainty, mitigation rarely will be optimal for societal reasons,but can still do some good Net benefit when mitigation lowers total cost below that of no mitigation Net loss when mitigation raises total cost above that of no mitigation

Within range, inaccurate hazard maps produce nonoptimal mitigation, raising cost, but still do some good (net benefit)

A hazard map that greatly overestimates the hazard can do net harm Resulting overmitigation cost can cause net loss

Epistemic uncertainty gives range of hazard estimates Select mitigation level between optima

Including risk aversion & aleatory uncertainty Consider marginal costs C’(n) & benefits Q’(n) (derivatives) Aleatory uncertainty in hazard model & mitigation efficiency causes uncertainty in expected loss. We are risk averse, so add risk term R(n) proportional to uncertainty in loss, yielding higher mitigation level n** cost Benefit (loss reduction ) Stein & Stein, 2012

Summary Limitations in our knowledge about earthquakes, notably space-time variability, limit how accurately hazard maps can be made Although uncertain maps likely produce nonoptimal mitigation, they still do some good if they’re not too bad Testing maps & quantifying uncertainties will help some Need to recognize & accept uncertainties

Haiti 2001 hazard map M7 earthquake shaking much greater than predicted for next 500 years

Epistemic uncertainty gives range of hazard estimates Select mitigation level between optima