Seismicity Rates from GPS and other Deformation Rate Estimates in Washington State Roy Hyndman, Lucinda Leonard & Stephane Mazzotti Geological Survey of Canada, Pacific Geoscience Centre, Sidney, B.C.; SEOS, University of Victoria; Univ. Montpellier, France Forearc crustal earthquakes Back Olympic Penin. Puget Sound arc x x x x x xx x x x x x xx Subduction thrust earthquakes x x Benioff-Wadati x slab earthquakes PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Catalogue Seismicity Rates Large event rates based on smaller event rates 10 Are there large events without Puget-Georgia Basin corresponding smaller magnitude 1 ) 1 - seismicity? r M 6 y ( ~50 yrs i.e., "characteristic . q e 0.1 earthquakes"? r f M 7 e v i ~400 yrs Cascadia t a l megathrust u 0.01 m & central Van. Is. m u Mx=7.5 two M7 with little C other seismicity 0.001 7.7 b~0.77 7.3 0.0001 3 4 5 6 7 8 R.D. Hyndman Magnitude PGC Pacific Geoscience Centre Geological Survey of Canada Long-term Crustal Seismicity Rate in Pacific Northwest --Can we estimate the earthquake rate from that required to accommodate deformation rate assuming all seismic? (GPS, paleoseismic fault displacement rates, tectonic models, etc.) (1) Puget Sound, (2) Olympic Penin., (3) E. Washington --Problem of short duration of catalogue seismicity; possibility of large infrequent earthquakes where few smaller ones, i.e., Cascadia megathrust, M7 central Van. Is. events, etc. "characteristic earthquakes"? --also issue of large crustal events in Puget Sound after megathrust as predicted by deformation data PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Seismicity rate from deformation rate (Kostrov, 1974; Anderson, 1979; Hyndman & Weichert, 1983; Hyndman et al., 2003; Mazzotti et al., 2011) Convergence Rate: s' = C Mo' / (2 µ A' ) where: A' = WL is the total cross-sectional area (i.e., need seismic thickness) Mo' = rate of seismic moment release per unit time in area, from recurrence Mo from magnitude vs moment relation µ = shear modulus C = faulting orientation constant, ~1.0 for 45o faulting From Hyndman et al. (2003). Better to use strain rates rather than uniaxial shortening but less intuitive, e.g., detailed discussion by Mazzotti et al. (2011) PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Washington Seismicity Areas Studied (mainly from Mazzotti et al., 2011; study with BC Hydro for mainly BC) Olympic Puget North-Eastern Peninsula Sound Washington PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Crustal Seismicity Study Areas 50N Comparison of long-term seismicity rate from catalogue seismicity and from deformation (GPS etc) Vancouver 49 1. Puget Sound 1 7 0 High seismicity, high 0 Back deformation rate arc 48 Olympic Seattle 2. Olympic Peninsula Peninsula Low seismicity, high Puget M6.8 deformation rate Sound 47 3. Eastern Washington Low seismicity, moderately low deformation rate 46 127W 126 125 124 -123 -122 -121 PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Puget Sound & Olympic Penin. N-S Deformation 130 126 122W 52 NORTH AMERICA N PLATE British 1 Oregon forearc Columbia V a 1 n block c 1 shortening . Is Vancouver zone is rotating . northward ? 1 A c Victor ia 48 c r compressing Puget e Seds. }? t e d Seattle Sound against S ? e 1 d ? i Vancouver Island m ? Washi ngton e n 1 1 t s backstop JUAN DE FUCA Por tland PLATE s (Wells &Simpson, McCafrey, 1 e o and others) ? n a 1 c l R o 44 V o ta 1 e b ti l n d oc g ? Oregon k a c 1 s Weak Olympic accr. a 1 C sediments focus ? forearc deformation ? 1 California in Puget area T S ra ie ns rr la 40 a ti b N n lo ev g 0 200 km ck ad a PGC R.D. Hyndman Pacific Geoscience Centre Modified from Wells and Simpson (2000) Geological Survey of Canada 10 ) 1 Oregon Block - N Measured GPS S Puget Sound N-S r 8 y Puget-Georgia m Basin 6 Deformation Rate m ( Vancouver Is. h t r 4 from GPS o n V 2 Stable continent reference Seattle 0 Coast mountain stations 3.0 -+ 0.7 mm/yr -2 Current rate relative to 50 48 46 44 Vancouver Island Latitude ( oN) 2 (which is moving north Gt. earthquake interseismic due to great earthquake 0 cycle elastic deformation) 10 ) 1 - N Residual GPS S r 8 y Long-term through great m 6 With respect m ( to DRAO earthquake cycle h t i.e., relative to stable r 4 o n continent V 2 S N 0 4.8 -+ 0.6 mm/yr -2 50 48 46 44 PGC (Hyndman et al., 2003) Updated GPS (Mazzotti et al.) 3.2 mm/yr N-S Shortening Rates for Puget Sound Area 1. GPS shortening rate (Sherrod, Mazzotti, Haugerud, 2008; Hyndman et al., 2003): Current: 3.2 ± 0.8 mm/yr: Long-term: 4.8 ± 0.6 mm/yr 2. GPS rotation models for Oregon Block (e.g., McCafrey et al., 2007): Puget Sound northerly motion: 4.4 ± 0.3 mm/yr. 3. Paleomagnetic models (e.g., Wells et al., 1998; Wells and Simpson, 2001): Northerly motion of southern Puget Sound: ~6 mm/yr 4. Active fault estimates: total north–south rate: 3.6 mm/yr (Sherrod, Mazzotti, & Haugerud, 2008; Wells and Simpson, 2001); Seattle fault: 0.7–1.1 mm/yr, Johnson et al., 1999) What seismicity rate is required to accommodate this deformation rate (assuming all seismic)? PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada GPS Data and Strain Map of Pacific Northwest (Mazzotti et al., 2011) Strain rate Estimate smoothed NE Washington strain field Puget (not just shortening) Estimate seismicity rate required to accommodate strain GPS data (rel. N. Am.) PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Seismicity Rate from GPS GPS based strain rate (Mazzotti et al., 2011): Catalogue Puget-Sound agrees well with catalogue GPS based Extra seismicity could be E. Washington spread over all magnitudes & Olympic (recurrence in figures), or accommodated by Penin. Catalogue a number of ~M7 much greater "characteristic" earthquakes than catalogue Mazzotti et al., 2011 (x10) PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Moment Rates from Catalogue Seismicity compared to from GPS Strain (Mazzotti et al., 2011) (1017 Nm yr-1) GPS Seismicity Ratio Puget Sound 2.8 2.2 ~ 1 Olympic Mtns. 4.1 0.23 18 E. Washington 1.9 0.20 10 PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Olympic Peninsula and Central/Eastern Washington Seismicity Rates from GPS Data 1. Catalogue seismicity for Puget Sound agrees well with estimate from deformation 2. Large N-S shortening rate for Olympic Peninsula (greater than for Puget Sound) but very little seismicity. Is deformation mainly aseismic in this accretionary sedimentary prism (as in such prisms elsewhere)? 3. Seismicity for Eastern Washington required to accommodate GPS strain is much greater than that in the catalogue. This is a common occurrence for low deformation rate areas. Explanation unclear; aseismic? But, more frequent large events than indicated by the catalogue cannot be excluded. ... low weighting? PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Good agreement of deformation estimates and catalogue seismicity: 1. Puget Sound N-S shortening 2. San Andreas fault system 3. Queen Charlotte fault zone 4. Denali fault zone, Alaska 5. Ofshore transform faults of J. de F. ridge 6. Charlevoix (eastern Canada seismic area) Deformation estimates much greater: Olympic Mtns, E. Washington, most of B.C. interior, other slow deformation examples; possible large "characteristic" events not shown by catalogue Large recorded earthquakes with few smaller events: Cascadia, Central Vancouver Island... PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Implications for PSHA: ratios of ground shaking derived from GPS vs seismic data (Mazzotti et al. (2011) Puget Sound: ~1:1 good agreement Olympic Peninsula and E. Washington: > 10 x seismicity; 2-3 times greater PGA from GPS- derived model compared to catalogue seismicity PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Conclusions: (1) Can long-term seismicity rates be estimated quantitatively from deformation data? (GPS, fault slip rates, tectonic models, etc.) Yes, but... probably only for areas with rapid deformation and well-organized "mature" faults or deformation zones (Hypothesis?) ----------- (2) Deformation confirms seismicity rates from catalogue for Puget Sound --GPS deformation larger than seismic for Olympic Mtns. and E. Washington, so could have large "characteristic" events, but give low weighting? PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada PGC R.D. Hyndman Pacific Geoscience Centre Geological Survey of Canada Very few catalogue earthquakes in Olympic Peninsula area Puget Olympic Sound Peninsula PGC R.D. Hyndman from McCrory et al. Pacific Geoscience Centre Geological Survey of Canada Puget Sound Seismicity Rates from Deformation Data 1. Seismicity rate to accommodate deformation is very close to rate from catalogue seismicity. 2. Vancouver Island receding to north due to locked megathrust elastic deformation. Will be recovered to south in megathrust. i.e., ~up to seven M > 7 events after megathrust possible. Crustal events occurring some "short" time after megathrust could be more important hazard than the megathrust itself, -but not really apparent in paleoseismic data? Sherrod et al. conclude about half of Holocene shortening occurred over short time period (~1100 yrs ago), including Seattle fault earthquake M≥7 (1020-1050 cal yr BP, Atwater 1999).