[S07-3] Hazard and risk assessment II

[S07-3-01] Reconciliation of Canada's 5th Generation Seismic Hazard Model results with those from the OpenQuake-engine

^○John Adams¹, Trevor Allen^{2, 3}, Stephen Halchuk¹ (1.Geological Survey of Canada, Ottawa, Canada, 2.Geological Survey of Canada, Sidney, Canada, 3.Geoscience Australia, Canberra, Australia)

The Geological Survey of Canada (GSC) has contributed earthquake hazard information for the National Building Code of Canada (NBCC) since the 1953 edition. The 2015 national hazard model update included many important advances on its predecessors, including: probabilistic treatment of the Cascadia subduction zone; reconfigured seismic source zones and special consideration of rare large eastern earthquakes; explicit definition of crustal fault sources in the Yukon Territory and offshore western margin faults (north of Cascadia) based on GPS observations and paleoseismic slip rates; catalogue magnitudes expressed consistently in terms of moment magnitude for improved magnitude-frequency statistics; and the use of a suite of representative backbone ground-motion models. The GSC is now working towards the 2020 building code model. A major element will be changing from 1980s-era FRISK88 software to the OpenQuake-engine for hazard computation. To ensure comparability we have performed comprehensive tests using our 2015 model implemented in OQ. Without adjustments for how some elements of the model are treated by the two codes, the engines give results within 2-3% for much of Canada, but up to 15% different in some places. When adjustments are made for magnitude-recurrence distribution, integration slice size, cutoff distance, and floating-rupture distribution along fault sources, the differences reduce to <2-3% everywhere.