1. SIO 226: Introduction to Marine Geophysics Seismicity LeRoy Dorman Scripps Institution of Oceanography March, 2006

2. Plate Geography

3. Questions about earthquakes ● The first question usually asked about an earthquake is “where was it?” - the usual answer is that it is on a plate boundary. ● The second question is usually “How big was it?”, and this leads us into the subject of earthquake magnitudes.

4. Earthquake Magnitude ● Among the first questions anyone asks about an earthquake is “How big was it?” ● There are four size estimates which we encounter today. ● The first in use in California was the Richter magnitude, based on the Wood-Anderson seismometer, denoted M L. ● The second was the surface wave magnitude, denoted M S. ● The third was the body-wave magnitude, denoted m b. ● The fourth is moment magnitude, denoted M W. This is the most useful for tectonic purposes.

5. Basis for Richter scale ● Upperpanel show measurement of amplitude. ● Lower panel shows establishment of distance correction, the basis for which is that the same magnitude should be given by the calculation regardless of the distance from the seismograph to the epicenter. ● Similar work was done by Wadati in Japan.

6. Richter (local) Magnitude ● Measured on Wood- Anderson torsion seismograph ● Peak amplitude -from surface waves- is measured ● Calibrated for California, but used elsewhere as well, especially for smaller quakes. ● Ar right is a nomogram for graphical calculation of the formula on the previous slide.

7. Surface wave magnitude ● The surface wave magnitude is made on the surface wave train, typically at a period around 20 seconds, and is most useful for distant earthquakes ● Calculated using Amplitude A and period T from

8. Body Wave Magnitude ● Calculated from ● Where, as before,A is amplitude, T is period, Δ is distance and h is depth. ● Q is taken from a contour plot determined empirically for different arrivals and components.

9. The Q function is determined empirically. ● These are determined for each arrival (P, S, PP) separately, allowing use of several measurements at the same seismic station. ● from Wysession, 1996.

10. Moment and Moment Magnitude ● Moment is defined as ● Here A is area of the fault slip zone and μ h is the shear modulus at the source and D is the displacement across the fault during the earthquake. ● Moment is tectonically useful since this displacement is that which should be consistent with motion across the plate boundary on which the quake occurs. ● The moment is related to moment magnitude by ● Where M 0 is in units of dyne-cm. ● The constants in this equation and the previous definitions of magnitude are chosen so that magnitudes measured in several ways will be approximately equal.

11. Source representations ● Any seismic source which is small in size can represented by a set of couples-pairs of force vectors. ● This representation is called the moment tensor, and can be used to describe quake and explosive sources. ● Each of these will generate a different seismogram.

12. Moment Tensor ● Earlier we noted that fault slip could be represented as slip across three orthogonal faults, with two components of slip on each fault. ● This is important since the synthetic seismogram observed at some station can be represented as a sum of Green's functions or fundamental seismograms. ● Thus what a station observes is the sum of the seismograms from all the moment components (multiplied by the moment for that component) ● These fault plane solutions are distributed by email within a few hours of major earthquakes, using data from a small number of stations.

13. ● This process only works when the seismic model for the region is good enough to make very accurate synthetic seismograms. It is usable for M W > 4.0.

14. Harvard CMT solutions, M W >5.3 Here is the solution for the recent event. January 26, 2006, PACIFIC-ANTARCTIC RIDGE, MW=5.6 Natasha Maternovskaya CENTROID, MOMENT TENSOR SOLUTION HARVARD EVENT-FILE NAME C012606B DATA USED: GSN L.P. BODY WAVES: 61S,101C, T= 40 SURFACE WAVES: 72S,141C, T= 50 CENTROID LOCATION: ORIGIN TIME 10:43:30.6 0.2 LAT 65.23S 0.01;LON 179.79W 0.03 DEP 17.4 1.0;HALF-DURATION 1.5 MOMENT TENSOR; SCALE 10**24 D-CM MRR=-0.52 0.05; MTT= 2.83 0.06 MPP=-2.32 0.05; MRT=-1.23 0.14 MRP= 0.56 0.12; MTP=-0.62 0.04 PRINCIPAL AXES: 1.(T) VAL= 3.34;PLG=19;AZM=188 2.(N) -0.85; 67; 329 3.(P) -2.49; 14; 93 BEST DOUBLE COUPLE:M0=2.9*10**24 NP1:STRIKE=230;DIP=67;SLIP= 176 NP2:STRIKE=321;DIP=87;SLIP= 23e t ########### ################### --##################### ------################----- ---------##########---------- -----------#####--------------- -------------#----------------- -------------##------------------ -----------######------------ - ---------#########----------- P - -------#############--------- - -----###############----------- ----#################---------- --####################------- -#####################----- ######## ##########-- ###### T ########## ## ######

15. Why so many magnitude scales? ● The problem with the conventional magnitude scales is that they saturate. ● The moment-based scale does not. ● The figure at right shows spectra for idealized earthquakes of various sizes. ● Geller, BSSA, 1976.

16. Shortcomings of magnitude scales ● Magnitude, being a single scalar, cannot characterize all properties of an earthquake. ● The two signals at right are from two earthquakes in the same source region, plotted at the same scale. ● The duration of shaking of the larger event is much longer, and the peak shaking is not in the time interval used to determine m b. ● Figure from C. Ammon, PSU.

17. Fault slip areas ● Much of the complexity of of the seismic signal from great earthquakes is that the rupture lasts a long time. ● For a slip velocity of 3km/s, the time to traverse the 150-km scale bar at the bottom of the figure is 50 seconds.

18. Cumulative moment is dominated by great earthquakes (figure from CMT site)

19. Earthquake statistics ● Are embodied in the Frequency- magnitude relation, due to Gutenberg and Richter. ● Their result is ● Here N is the number of earthquakes with magnitude greater than M occurring in a given time. A depends on the activity and length of time period and b is generally about 1. ● Plot at right is from H. R. DeShon, S. Y. Schwartz, S. Bilek, L. M. Dorman, V. Gonzalez,. M. Protti, E. Flüh, T. Dixon, JGR, 2003, for Costa Rica's subduction zone. ● Complete for 1.9-4.

20. Earthquake seismic tomography ● Consider that we have a set o travel time observations along the paths shown at right. ● The travel time for each path will be ● Where d j is the distance traveled in box j and Δ u j is the perturbation of slowness for that box. ● For a set of travel times, we can write a system of equations ● d = Gm for data vector d and model m. ● If the size of d is greater than the size of m, we can usually solve for m and adjust the model to fit the data more closely.

21. Tomographic solution for velocity ● This is the P-velocity structure beneath the Lau backarc basin, with the plate at right. ● Boxes are OBSs, red dots are land stations. ● Not shown are 188+ earthquakes. ● From Zhao, Xu, Wiens, Dorman, Hildebrand and Webb, Science 1997

22. Attenuation across Lau Basin ●

23. Single and double seismic zones ● Some Benioff zones exhibit a double seismic zone, as shown at right in a paper of Peacock, 2001. The upper zone represents the slip zone between the downgoing and overriding plates. ● The cause of the lower zone is not yet clear, but may be caused by a phase transformation, as in the next slide.

24. Antigorite dehydration ● The arrows in the lower box show the P-T path followed by the edge of the subducting plate.