2018-04-02_REVISION - M1977342 - Laserfiche WebLink

Home Browse Search

SECT NFIVE Seismic Nwid Results the Abrahamson et al. (2014)model gives the highest hazard, while Boore et al. (2014)gives the lowest(Figure 29). At a return period of 10,000 years, there is a factor of 2 difference in PGA between the models predicting the highest and lowest hazard. By comparison,the difference in the 5 h and 95th percentile PGA at 10,000 years is a more than a factor of 4 (Figure 11). At 0.2 sec SA,there is less difference between the four models (Figure 30).At 1.0 sec SA,Abrahamson et al. (2014) model gives the highest hazard, while Campbell and Bozorgnia (2014) gives the lowest(Figure 31). On Figure 32, mean UHS for a range of return periods from 500 to 30,000 years is shown. The UHS are also provided in Table 6. Fi ure 33 shows the range in epistemic uncertainty in the 10,000-year return period UHS. The 5t and 95th percentile UHS differ by a factor of 4.0 to 5.0 for periods less than 0.75 sec. This factor increases to between 5.0 and 6.8 for periods between 0.75 sec and 5 sec. At very long periods, the factor drops to about 3.6. These ranges represent the uncertainty modeled in the PSHA based on our current knowledge of all inputs. Additional information on seismic sources or future refinement of ground motion models would likely decrease the range of uncertainty, but the mean may shift. The 9501 percentile 10,000-year return period UHS is similar to the mean 30,000-year return period UHS, which is also shown on Figure 33. 5.2 COMPARISON WITH 2O13 RESULTS The current study includes an updated characterization of the WFMF and Frontal fault, (Section 4.1.1), an update of the background seismicity (Section 4.1.2), revised site conditions (Section 4.2), and an update of the ground motion prediction models (Section 4.3). Figure 34 shows the mean PGA hazard curves for the sources contributing the most to the hazard along with the total mean hazard curves compared to the 2013 results (Olig et al., 2013). There is an increase in PGA hazard from the WFMF and Frontal fault along with an increase in the total mean PGA hazard. As shown on Figure 35, the 1.0 sec SA hazard is increased slightly for the WFMF and Frontal fault,but the total mean 1.0 sec SA hazard is similar to the 2013 results. Figures 36 and 37 illustrate the differences in the 2017 and 2013 results due to site conditions, ground motion model, and seismic source model. The increase in Vs30 (from 940 m/sec to 1,500 m/sec) results in a decrease in the PGA hazard as shown by the comparison of the solid green and solid black curves on Figure 36, which result from using the 2017 seismic source model and NGA-West2 ground motion models,but the two different site conditions. Changes to the seismic source model result in an increase in the PGA hazard as shown by comparison of the dashed black and dashed red curves on Figure 36. The update of the ground motion models from NGA-Westl to NGA-West2 has very little impact on the PGA hazard for the site as shown by the dashed red and solid green curves on Figure 36. A similar trend can be seen for the 1.0 sec SA hazard results (Figure 37), although a slightly larger decrease in hazard due to the change in site conditions causes the final total 1.0 sec SA hazard to be very similar for return periods greater than about 2,000 years. Table 7 compares ground motions computed from this study with the 2013, 2011, and 1995 results. Figure 38 graphically illustrates the PGA and 1.0 sec SA values from these studies. As discussed above, there is an increase in mean ground motion from 2013 to 2017 studies due to the changes in the seismic source model. There is also a slight increase in the range of the 5th to 95d'fractiles. 21