history and future of operational earthquake forecasting at the usgs andrew j. michael

History and Future ofOperational Earthquake Forecasting

at the USGS

Andrew J. Michael

What Is Operational Earthquake Forecasting?a.k.a. OEF

Currently OEF = Earthquake Clusteringa.k.a. Foreshocks and Aftershocks

Earthquakes trigger other earthquakes.

When you have one earthquake,the probabilities for other earthquakes go up.

Semantics

When earthquake happens without obvious ancestors, it is a mainshock.

If smaller earthquakes follow, they are aftershocks.

If an aftershock occurs that is bigger than the mainshock,then we rename all previous events foreshocks,and the biggest aftershock is now the mainshock.

If smaller earthquakes follow, they are aftershocks.

…… The Physics Fine Print: all earthquakes are the same, we can’t tell foreshocks, mainshocks and aftershocks apart.

Standard Aftershock WarningIssued after M>=5 Earthquakes in California

PROBABILITY REPORT Published on May 24, 2009 @ 16:58:34 GMT

Version 3: This report supersedes any earlier probability reports about this event.

MAINSHOCK Magnitude : 4.6 Mw (A light quake) Time : 23 May 2009 03:58:32 PM, PDT : 23 May 2009 22:58:32 GMT Coordinates : 36 deg. 25.28 min. N, 117 deg. 46.24 min. W : 36.4213 N, 117.7707 W Depth : 0.0 miles ( 0.1 km) Quality : Fair Location : 7 mi. ( 11 km) SE from Keeler, CA : 20 mi. ( 32 km) SE from Lone Pine, CA Event ID : NC 66099

STRONG AFTERSHOCKS (Magnitude 5 and larger) -

At this time (18 hours after the mainshock) the probability of a strong and possibly damaging aftershock IN THE NEXT 7 DAYS is less than 10 PERCENT

EARTHQUAKES LARGER THAN THE MAINSHOCK -

Most likely, the recent mainshock will be the largest in the sequence. However, there is a small chance (APPROXIMATELY 5 TO 10 PERCENT) of an earthquake equal to or larger than this mainshock in the next 7 days.

WEAK AFTERSHOCKS (Magnitude 3 to 5) -

In addition, up to approximately 5 SMALL AFTERSHOCKS are expected in the same 7-DAY PERIOD and may be felt locally.

This probability report is based on the statistics of aftershocks typical for California. This is not an exact prediction, but only a rough guide to expected aftershock activity. This probability report may be revised as more information becomes available.


Background Information About Aftershocks

Like most earthquakes, the recent earthquake is expected to be followed by numerous aftershocks. Aftershocks are additional earthquakes that occur after the mainshock and in the same geographic area. Usually, aftershocks are smaller than the mainshock, but occasionally an aftershock may be strong enough to be felt widely throughout the area and may cause additional damage, particularly to structures already weakened in the mainshock. As a rule of thumb, aftershocks of magnitude 5 and larger are considered potentially damaging.

Aftershocks are most common immediately after the mainshock; their average number per day decreases rapidly as time passes. Aftershocks are most likely to be felt in the first few days after the mainshock, but may be felt weeks, months, or even years afterwards. In general, the larger the mainshock, the longer its aftershocks will be felt.

Aftershocks tend to occur near the mainshock, but the exact geographic pattern of the aftershocks varies from earthquake to earthquake and is not predictable. The larger the mainshock, the larger the area of aftershocks. While there is no "hard" cutoff distance beyond which an earthquake is totally incapable of triggering an aftershock, the vast majority of aftershocks are located close to the mainshock. As a rule of thumb, a magnitude 6 mainshock may have aftershocks up to 10 to 20 miles away, while a magnitude 7 mainshock may have aftershocks as far as 30 to 50 miles away.


STRONG AFTERSHOCKS (Magnitude 5 and larger) -At this time (18 hours after the mainshock) the probability of a strong and possibly damaging aftershock IN THE NEXT 7 DAYS is less than 10 PERCENT

EARTHQUAKES LARGER THAN THE MAINSHOCK -Most likely, the recent mainshock will be the largest in the sequence. However, there is a small chance (APPROXIMATELY 5 TO 10 PERCENT) of an earthquake equal to or larger than this mainshock in the next 7 days.

WEAK AFTERSHOCKS (Magnitude 3 to 5) -In addition, up to approximately 5 SMALL AFTERSHOCKS are expected in the same 7-DAY PERIOD and may be felt locally.

California Advisories

Mineral, VA

STEP

Gerstenberger et al.Nature, 2005USGS OFR 2004

Reasenberg and Jones, Science, 1989

Probability of earthquakesduring an aftershock sequenceas a function of time andmagnitude.

Initial estimates are based onparameters for a “generic”California earthquake sequence.

Sequence specific parametersare used once they can bedetermined.

Spatial Clustering

Aftershocks occur in the general region of the mainshock.

Up to about 1 – 2 fault lengths away.

In the case of a large aftershock at the edge of the zone, the zone will get larger.

The people we are warning, felt the mainshock or should be aware of it.

How Many Aftershocks?

Bigger Mainshocks Produce More Aftershocks

Magnitude of Mainshock

Expected number of aftershocksMagnitude 3 or larger

In the first week

5 6.7

6 67

7 670

8 6700

The Sizes of Aftershocks(including bigger ones)

Minimum Magnitude

Observed Number Expected Number

3 310 340

4 42 34

5 6 3.4

6 0 0.34

7 0 0.034

Most Aftershocks Are Smaller

1994 Northridge Earthquake Magnitude 6.7Number of Aftershocks within 30 km of Mainshock

In the first week

The Times of Aftershocks

When Observed Number Expected Number

First Day (½ – 1½) 420 733

Tenth Day 44 70

100th Day 4 7.1

1000th Day 0 0.71

10000th Day ? until 2021 0.071

The Rate of Aftershocks Decays at About 1/time

1994 Northridge Earthquake Magnitude 6.7Number of Aftershocks in 1 day

Magnitude 2 or larger, within 30 km of Mainshock

Forecast After a Magnitude 7

MinimumMagnitude

Expected Number Probability of 1 or more events

3 670 >99%

4 67 >99%

5 6.7 99%

6 0.67 50%

7 0.067 6.5%

8 0.0067 0.66%

Magnitude 7 MainshockForecast for the First Week


MinimumMagnitude


3 6.7 99%

4 0.67 49%

5 0.067 6.5%

6 0.0067 0.67%

7 0.00067 0.067%

8 0.000067 0.0067%

Magnitude 5 MainshockForecast for the First Week Wide Range of

Probabilities


MinimumMagnitude


3 0.31 27%

4 0.031 3.1%

5 0.0031 0.31%

6 0.00031 0.031%

7 0.000031 0.0031%

8 0.0000031 0.00031%

Magnitude 5 MainshockForecast for the Fifth Week

The Long Lives of Aftershocks

Aftershocks of the Magnitude 8¼ Nobi Earthquake of 1891

Forecasts With Respect toUrban Planning for Recovery

Stage

1. Emergency Response (search and rescue, fire fighting, shelters, damage assessment)

2. Restoration(restore utilities, debris removal, temporary repairs)

3. Reconstruction(structures replaced to pre-disaster levels)

4. Betterment(major projects improve community to a new standard)

5. Long-Term(life with a new normal)

Forecast for a Magnitude 7 Mainshock


Stage Dominant Time Period(Kobe, Northridge, Christchurch)


0 to 14 days


14 days to 1 year


1 to 3 years


3 to 10 years


10 to 50 years



Stage Dominant Time Period(Kobe, Northridge, Christchurch)

Probability of an Aftershock with Magnitude 6 or larger


0 to 14 days 61%


14 days to 1 year 33%


1 to 3 years 11%


3 to 10 years 11%


10 to 50 years 13%


Uncertainties

Variability from sequence to sequence by at least a factor of 10. Our forecasts will adapt to each sequence as we collect data.

Agnew and Jones, JGR, 1991:

“But it ought to be possible to do better:

Should we say the same thingafter every event?

the probability of a very large earthquake should be higher if the candidate foreshock were to occur near a fault capable of producing that mainshock than if it were located in an area where we believe such a mainshock to be unlikely.

Moreover, the chance of a candidate earthquake actually being a foreshock should be higher if the rate of background (nonforeshock) activity were low.”

Foreshock Modelto a

Specific Mainshock

M4.8 Event At Bombay Beach On March 24, 2009Could It Be A Foreshock To A Larger Earthquake In The Next 3 Days?

Mainshock:SAF, Coachella Seg.UCERF2:Length = 69 kmM 75-yr Prob. = 5%3-day Prob.= 0.009%



Reasenberg &Jones, 1989:Probabilityof M4.8 beingfollowed byan M≥7 eventPF = 0.05%



R&J 1991:PF= 0.05%

Agnew andJones, 1991:PF = 4%


Past Efforts - Parkfield

Past Efforts - Parkfield

Alert versusStatus


MinimumMagnitude

Probability of 1 or more events

Probability of 1 or more events if

M≥7 occur 100xmore frequently

3 >99% >99%

4 49% 52%

5 6.5% 12%

6 0.67% 7%

7 0.067% 6.4%

Magnitude 5 MainshockForecast for the First Week Integrated Aftershock

and Foreshock Forecast

Uncertainties

The probabilities for the largest events could be underestimated by a factor of 10 to 100.

Future USGS Plans

Expand the standard aftershock warnings nationally and internationally (for internal government use) with updated messaging. This will start as a Reasenberg and Jones calculation.

Develop the UCERF3 clustering model and associated products. This will combine Reasenberg and Jones type calculations with Agnew and Jones type calculations.

1-year hazard assessments for induced seismicity.

Test models and products, develop new methods.

Are these statements worth saying?

• They are true, everyone knows it, and they will ask.

• New Zealand and Italy experiences

• When probabilities are high enough then public warnings have been issued in conjunction with the State of California.

• "During the August 8, 1989 advisory, all our departments ran drills to prepare for an imminent earthquake. This made a tremendous difference in the city's response when the earthquake struck," said Henry Renteria, Emergency Services Manager, Oakland, California

•"The probability of aftershocks given by the U.S. Geological Survey was one of the factors we used in deciding how many firefighters to keep on duty after the 1989 San Francisco Bay area earthquake," said Greg Abell, Battalion Chief in the San Francisco Fire Department.

•The important thing is what you want to do with them.

history and future of operational earthquake forecasting at the usgs andrew j. michael

Documents

aftershocks earthquakes

aftershocks of magnitude

mainshock magnitude

small aftershocks

numerous aftershocks

recent mainshock

strong aftershocks magnitude

weak aftershocks magnitude