Chiou and Youngs PEER-NGA Empirical Ground Motion Model for ...

Chiou and Youngs PEER-NGA Empirical Ground Motion Model for ... Chiou and Youngs PEER-NGA Empirical Ground Motion Model for ...

peer.berkeley.edu
from peer.berkeley.edu More from this publisher
08.04.2013 Views

shown on each figure. The depths for southern California sites in the PEER-NGA data base, as well as the modeling performed by Day et al. (2006), are based on the SCEC Version 2 3- D velocity model (Magistrale et al., 2000). The SCES web site contains an updated Version 4 velocity model for southern California. The major change between Version 2 and Version 4 is in the velocity gradient at shallow depths, with the depths to VS of 1.0 km/s in the Version 4 model being systematically 60 percent of the depths in the Version 2 model. The depths to VS of 2.5 km/s is similar in Version 2 and Version 4 of the SCEC models. At this time, we are not prepared to propose a model for the effect of sediment depth because of the uncertainty in the Z1.0 depths for the southern California sites. We think that Z1.0 is a more practical parameter to use than Z2.5 because it is more likely to be available to a user based on site data. The residuals do indicate that large sediment depths lead to larger ground motions. The residuals also indicate that very thin sediment depths lead to smaller long period motions (Figures 37c and 37d). resid 2 1 0 -1 -2 SCEC Version 2 0 200 400 600 800 1000 1200 1400 Z1 (m) C&Y2006 Page 53 resid 2 1 0 -1 -2 SCEC Version 4 0 200 400 600 800 1000 1200 1400 Figure 37a: Intra-event residuals versus Z1.0 for spectral period of 0.01 seconds (pga). Plot on left shows sites with velocities based on SCEC Version 2 3-D velocity model and plot on right shows sites with velocities based on SCEC Version 4 3-D velocity model. resid 2 1 0 -1 -2 SCEC Version 2 0 200 400 600 800 1000 1200 1400 Z1 (m) resid 2 1 0 -1 -2 Z1 (m) SCEC Version 4 0 200 400 600 800 1000 1200 1400 Figure 37b: Intra-event residuals versus Z1.0 for spectral period of 0.2 seconds. Plot on left shows sites with velocities based on SCEC Version 2 3-D velocity model and plot on right shows sites with velocities based on SCEC Version 4 3-D velocity model. Z1 (m)

esid 2 1 0 -1 -2 SCEC Version 2 0 200 400 600 800 1000 1200 1400 Z1 (m) C&Y2006 Page 54 resid 2 1 0 -1 -2 SCEC Version 4 0 200 400 600 800 1000 1200 1400 Figure 37c: Intra-event residuals versus Z1.0 for spectral period of 1.0 seconds. Plot on left shows sites with velocities based on SCEC Version 2 3-D velocity model and plot on right shows sites with velocities based on SCEC Version 4 3-D velocity model. resid 2 1 0 -1 -2 SCEC Version 2 0 200 400 600 800 1000 1200 1400 Z1 (m) resid 2 1 0 -1 -2 Z1 (m) SCEC Version 4 0 200 400 600 800 1000 1200 1400 Figure 37d: Intra-event residuals versus Z1.0 for spectral period of 3.0 seconds. Plot on left shows sites with velocities based on SCEC Version 2 3-D velocity model and plot on right shows sites with velocities based on SCEC Version 4 3-D velocity model. Comparisons with Data for Individual Earthquakes: Appendix E contains plots showing the model fit to the data for individual earthquakes. The predicted ground motions in these plots include the inter-event random effect. Nonlinear Soil Model: Figures 37a through 37b compare the site amplifications versus VS30 and ground motion amplitude predicted by the nonlinear soil model developed as part of this ground motion model with the site amplifications computed by Silva (2004) using equivalent linear site response analyses and by Choi and Stewart (2003) using empirical ground motion data. The soil model developed in this study compares well with the site response results computed by Silva (2004) for spectral periods of 0.01, 0.2, and 1.0 seconds and shows greater amplification at a spectral period of 3.0 seconds. The model compares well to Choi and Stewart’s results at spectral periods of 0.01 and 1.0 seconds, is more nonlinear at a spectral period of 0.2 seconds, and shows greater amplification a period of 3.0 seconds. Z1 (m)

esid<br />

2<br />

1<br />

0<br />

-1<br />

-2<br />

SCEC Version 2<br />

0 200 400 600 800 1000 1200 1400<br />

Z1 (m)<br />

C&Y2006 Page 54<br />

resid<br />

2<br />

1<br />

0<br />

-1<br />

-2<br />

SCEC Version 4<br />

0 200 400 600 800 1000 1200 1400<br />

Figure 37c: Intra-event residuals versus Z1.0 <strong>for</strong> spectral period of 1.0 seconds. Plot on left shows<br />

sites with velocities based on SCEC Version 2 3-D velocity model <strong>and</strong> plot on right shows sites with<br />

velocities based on SCEC Version 4 3-D velocity model.<br />

resid<br />

2<br />

1<br />

0<br />

-1<br />

-2<br />

SCEC Version 2<br />

0 200 400 600 800 1000 1200 1400<br />

Z1 (m)<br />

resid<br />

2<br />

1<br />

0<br />

-1<br />

-2<br />

Z1 (m)<br />

SCEC Version 4<br />

0 200 400 600 800 1000 1200 1400<br />

Figure 37d: Intra-event residuals versus Z1.0 <strong>for</strong> spectral period of 3.0 seconds. Plot on left shows<br />

sites with velocities based on SCEC Version 2 3-D velocity model <strong>and</strong> plot on right shows sites with<br />

velocities based on SCEC Version 4 3-D velocity model.<br />

Comparisons with Data <strong>for</strong> Individual Earthquakes: Appendix E contains plots showing<br />

the model fit to the data <strong>for</strong> individual earthquakes. The predicted ground motions in these<br />

plots include the inter-event r<strong>and</strong>om effect.<br />

Nonlinear Soil <strong>Model</strong>: Figures 37a through 37b compare the site amplifications versus VS30<br />

<strong>and</strong> ground motion amplitude predicted by the nonlinear soil model developed as part of this<br />

ground motion model with the site amplifications computed by Silva (2004) using equivalent<br />

linear site response analyses <strong>and</strong> by Choi <strong>and</strong> Stewart (2003) using empirical ground motion<br />

data. The soil model developed in this study compares well with the site response results<br />

computed by Silva (2004) <strong>for</strong> spectral periods of 0.01, 0.2, <strong>and</strong> 1.0 seconds <strong>and</strong> shows<br />

greater amplification at a spectral period of 3.0 seconds. The model compares well to Choi<br />

<strong>and</strong> Stewart’s results at spectral periods of 0.01 <strong>and</strong> 1.0 seconds, is more nonlinear at a<br />

spectral period of 0.2 seconds, <strong>and</strong> shows greater amplification a period of 3.0 seconds.<br />

Z1 (m)

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!