Earthquake prediction. Southwest BC: Our local dangerous fault zones

Transcription

1 Earthquake prediction Prediction: specifies that an earthquake of a specific magnitude will occur in a defined region during a particular period: There will be a M 7 or larger earthquake in southern California in March of 25 Forecast: provides a probability of the above, usually over to year scales 2% chance of a M 6.5 to M 7 earthquake on the San Andreas Fault in the next thirty years 62% chance of a M 6.5 or greater earthquake in the San Francisco Bay Area in the next thirty years Southwest BC: Our local dangerous fault zones Queen Charlotte Fault M 8 Shallow crustal faults M < 8 Cascadia Subduction Zone Fault M 9.? Faults in the subducting slab M < 7?

2 Use earthquake statistics and assumption about maximum earthquake size if details of faults are unknown Mazzotti et al., 2 Forecasting when a large earthquake is likely to happen, if you know something about the fault and its past large earthquakes! earthquakes do not happen at regular intervals (even at Parkfield CA, famous for regular SAF earthquakes, of M 6, the spacing is not actually regular)! but earthquakes are not perfectly random, either 24

3 Three types of forecasts (for when you know something about the fault and its past large earthquakes) Cascadia Subduction Zone Fault M 9.? Queen Charlotte Fault M 8 Conventional Forecast Say all we know is an average recurrence interval. We have no clue how regular or random the large quakes are. In this case, the estimated probability is constant. Example: a fault has earthquakes on average every 5 years. That means that each year, there is a in 5 chance of an earthquake. In 3 years, there is a 3 in 5, or 2% chance.

4 Conventional Forecast Queen Charlotte Fault: average recurrence interval for M 8 earthquakes is about every years, as little as 5, as much as 2. Thus, conventional forecast is about a % chance per year, or a % chance per decade. Elastic stress on the fault increases gradually, so the chance of a big earthquake grows with. In this case the probability in 9 was 3% in the next 3 years but the probability in 2 was 48% in the next 3 years We have renewal forecasts for the Casadia Subduction Zone fault. We need to know: Renewal forecast mean recurrence interval standard deviation if the recurrence interval since the last large earthquake

5 Renewal Forecast in the SF Bay Area Probability of a M 6.7 or larger earthquake in the San Francisco Bay Area between 23 and 232 Probability of a large quake in a region includes probabilities on all of the local faults This is one place where all of the major faults are mapped and their recurrence intervals and most recent events are known. Renewal forecast with earthquake interaction Effects of stress changes caused by nearby earthquakes may cause probabilities of another shock to rise or fall temporarily. These stress changes can be from permanent deformation of the ground or from passing seismic waves. In this case the probability increased from 48% in 3 years to 65% in the next 3 years due to stresses from a large quake nearby. This is the most sophisticated type of forecasting and it is not very common.

6 Renewal forecast requires Who tells us this? mean recurrence interval standard deviation of recurrence interval since the last earthquake record of earthquakes vs. Geologists. Trenching across active faults, dating offset former ground surfaces and getting a history of large earthquakes

7 Signs of prehistoric earthquakes along the SAF: many trenches AD 2-4 AD 6-8 AD Cascadia Subduction Zone Fault drowned ancient stumps from tress killed by sudden subsidence in a Cascadia earthquake M9+ earthquakes every (2 to 3) years wood fragments sandwiched between sand layers where land surface dropped suddenly and killed a mature forest photos by Steve Carlson plus dates of catastrophic undersea debris flows (submarine landslide deposits)... (Trenching is not possible)

8 From Saanich Inlet drill cores (Stevens et al., 2): 8 DFDs (3, 4 435, , , , 33, , , , 694 8, , , , , , , , !yrs ago from A.D.!2 datum) were correlated among two or more cores during this period, suggesting an average return period of strong shaking from earthquakes of about 22!yr. Nine of the DFDs overlap with the age ranges for great plate-boundary earthquakes that have been determined by other paleoseismic studies: coastal subsidence and offshore turbidity deposits. The remaining nine events give an average return period of about 47!yr for strong shaking from local earthquakes. ( DFD = debris flow deposit - EHH) Make a bar chart of the recurrence intervals - they sort of fit a Gaussian (normal) distribution Probability density function φ = σ (x µ) 2 2π e 2σ 2 x in EQ above is small σ σ = µ large σ σ = standard deviation of the recurrence interval µ = mean recurrence interval σ 5 +σ (for the red line) years (for the red line) = mean = in above EQ µ µ µ this axis is ( - )/ There were typos in the version I showed in class. April fool? no - I am sorry.

9 Integrate this Gaussian function to get the probability of an earthquake over a interval (area under the curves on the previous slide) Cumulative Probability : from to small σ σ = µ large σ Φ = µ ( + erf(x 2 σ 2 )) x in EQ above is σ = standard deviation of the recurrence interval µ = mean recurrence interval σ 5 +σ (for the red line) years (for the red line) = mean = in above EQ µ µ µ this axis is ( - )/ There were typos in the version I showed in class. April fool? no - I am sorry. Cumulative Probability Probability density function (based on recurrence s) t t t small σ t2 is just after the last big quake. t2 t t2 t t2 large σ

10 Cumulative Probability Probability density function (based on recurrence s) t t t2.5.5 t2 t t2 t t2 Cumulative Probability Probability density function (based on recurrence s) now now Predicting the probability for a interval in the future. T % chance during T T 2% chance during T

11 Onur and Seeman, 24 Cumulative Probability Probability density function (based on recurrence s) If we know the quake has not happened yet... and we are at no earthquake occurred now now T =3yr.5.5.4/.9 = 45% chance during next 3 years no earthquake occurred then only use this part of the PDF.2/.7 = 3% chance during next 3 years T =3yr

12 3th World Conference on Earthquake Engineering Vancouver, B.C., Canada August -6, 24 Paper No. 65 PROBABILITIES OF SIGNIFICANT EARTHQUAKE SHAKING IN COMMUNITIES ACROSS BRITISH COLUMBIA: IMPLICATIONS FOR EMERGENCY MANAGEMENT Tuna ONUR, and Mark R. SEEMANN Earthquake forecasting summary We can usually forecast where damaging earthquakes will be (seismic gaps on known faults) but we are still often surprised (e.g., blind faults, intraplate faults) We can usually forecast their effects (e.g., strength and duration of shaking, tsunami genesis) We cannot predict the timing of earthquakes very well (though probabilities over long periods can somes be estimated)