Project Abstract
|
We have simulated broadband (BB) ground motion synthetics for the M6.7 Northridge, CA, earthquake using the hybrid method by Mai et al. (2010). The bias of spectral accelerations between data and synthetics at 28 near-field stations is less than 0.5. We then generate map-based goodness-of-fit (GOF, Olsen and Mayhew, 2010) values for Northridge inelastic/elastic (IE) ratios for the bandwidths 0.2-0.5 s, 0.75-1.5 s, and 2-5s. The GOF scale ranges from 0 (poor) to 100 (best). The short-period IE ratios are poorly modeled, generally with GOF values < 40-50. The IE GOF values increase somewhat in the intermediate range, with most near-field station values in 45-65 range. Finally, the long-period IE ratios are generally in the range 55-75. When the IE ratios are plotted as a function of period and the ratio R (elastic displacement to yield displacement) the data and synthetics are very close for periods between 0.4 and 10 s, but start to diverge at shorter periods. Usually, but not for all stations, the synthetic IE ratios underpredict data for the intermediate to shorter periods. This matches the results from Olsen and Mayhew (2010) for the smaller M5.4 Chino Hills earthquake, and suggests the need for refining the BB method by Mai et al (2010) when used to compute IE ratios at shorter and intermediate periods. Future work should include additional scenarios, and compute the IE ratios using additional BB methods, such as the Graves and Pitarka (2010) and Liu et al. (2006) methods. |