Scaling relations for strong ground motion prediction models with M2 terms

1996 ◽  
Vol 86 (2) ◽  
pp. 329-336
Author(s):  
Yoshimitsu Fukushima

Abstract In most previous studies on the attenuation relation of strong ground motion, the increase in amplitude with earthquake magnitude was expressed by a linear relation. However, if the ω−2 source model is assumed, the corner period of the spectrum varies with the magnitude, and the scaling law of the source spectrum amplitude becomes a complicated function of magnitude. The scaling relation of the spectrum amplitude with MW can be simply approximated as a quadratic function, and the coefficient of the MW2 term should be negative. On the contrary, positive coefficients of the ML2, mbLg2, and MJ2 (MJ: magnitude of Japan Meteorological Agency) terms have been derived by regression analyses of strong ground motions. Based on the ω−2 model, semi-empirical relations between Mo and the magnitudes were derived. Further, the seismological scaling law of the source spectrum obtained from the relation between Mo and the magnitudes was approximated as a quadratic function of the magnitude, and positive coefficients of the squared magnitude terms were obtained. These results stem from the definitions of magnitude derived from amplitudes of seismograms around a specific period of 1.0 sec for ML and mbLg and of 5.0 sec for MJ.

2013 ◽  
Vol 438-439 ◽  
pp. 1474-1480
Author(s):  
Ju Fang Zhong ◽  
Long Wei Zhang ◽  
Jun Wei Liang

The key to near-field strong ground motion simulation based on stochastic finite fault method is to determine the spectrum of ground motion. We present an improved source spectrum model for simulation near-field strong ground motion acceleration time history. We combine Masudas source spectrum model with scaling factor Hij to keep radiation energy conservation and reflect the energy decrease with frequency at low to mid frequencies. We calculate the Fourier amplitude spectrum Fa, accelerate response spectrum Sa, velocity response spectrum Sv and displacement response spectrum Sd of simulation time histories. By comparative analysis of the laws of spectrum values (Fa, Sa, Sv, Sd) with the variation of frequency or period, we discusses the effects of sub-fault dividing scheme, the method of determining scale factor and source spectrum model on spectrum values (Fa, Sa, Sv, Sd). The results show that sub-fault dividing scheme has slightly effect on the model presented in this paper, and the model enable to reflect the sink laws of source spectrum value in mid-to-low frequencies well. We demonstrate that the improved model is superior to other commonly used models.


Science ◽  
2014 ◽  
Vol 343 (6169) ◽  
pp. 399-403 ◽  
Author(s):  
M. A. Denolle ◽  
E. M. Dunham ◽  
G. A. Prieto ◽  
G. C. Beroza

Sign in / Sign up

Export Citation Format

Share Document