10:45 AM - 11:15 AM
[S08-2-02] Application of the paleoseismic record of great Cascadia earthquakes for use in the 2015 and 2020 National Building Code of Canada seismic hazard maps
invited
The 2005 National Building Code of Canada (NBCC2005) used a deterministic model to derive the shaking from great Cascadia subduction interface earthquakes. The resultant values were used for design only if they exceeded the hazard values from other sources. NBCC2015 introduced a probabilistic treatment of Cascadia's Juan de Fuca segment. The rate of great earthquakes used the complete-rupture chronology established from turbidite analysis by Goldfinger et al. (2012 pre-print). Partial, southern ruptures were neglected as being too distant from Canada. The analysis used 18 turbidites of a 10,097-year record, giving a mean interval of 532 years and a standard deviation of 234 years. The rupture intervals (time since the last earthquake) were turned into magnitudes using fixed rupture length and width together with choices for slip rate and crustal rigidity to give central, upper and lower recurrence rates. Magnitudes were in the range 8.55-9.28. Three choices for the down-dip extent of the rupture were used, to capture the uncertainty in how close the energy release would be to sites on land. The model was combined with other crustal and inslab sources to give probabilistic seismic hazard estimates. The analysis neglected issues of partial rupturing and event clustering. Shaking from the offshore Cascadia subduction fault dominates 2015 design values in Victoria and Vancouver for periods > 1 s. For the NBCC2020 maps an updated turbidite history adds 4 events off Washington that decreases the mean interval to 434 years and reduces the standard deviation to 168 years, making the events appear more regular. The change in rate increases the onshore seismic hazard by about 9%.