Statistical Characteristics of Velocity Pulses for Near-fault Strong Motions Based on Wavelet Approach

SUN Yiou; DING Haiping

doi:10.13512/j.hndz.2014.03.005

Abstract：In this paper, 180 records are selected based on wavelet approach. These records show three common features:their rupture distances are less than 20km, their peak ground velocity (PGV) are greater than 10cm/sec and they all show characteristics of velocity pulses for near-fault strong motions. All Records are divided into groups according to different site conditions (rock vs. soil), different fault types (strike-slip fault vs. non-strike-slip fault) and different PGV threshold levels. Regression coefficients of several period-magnitude relationships in different situations are then given. With statistics in hand, several conclusions can be made: ① pulse periods in strike-slip faults are greater than those in non-strike-slip faults, but things changed with moment magnitude greater than 7, which is believed to have relationship with insufficient severe earthquake records. With limited severe earthquake records, the result may be affected. ② In both strike-slip faultsˊ and non-strike-slip faultsˊ cases, pulse periods in rock sites are greater than those in soil sites with moment magnitude less than 6.5 or so, and pulse periods in rock sites will be greater than those in soil sites as moment magnitude increases, but the boundary line on which pulse periods change differs according to different fault types. ③ The period-magnitude relationships derived from different PGV threshold levels varies a lot. For non-strike-slip faults, both soil sites and rock sites show the same property, that is, when magnitude is relatively small, the higher the PGV threshold level is, the smaller the pulse periods will be, however, pulse periods increase with the moment magnitude, when the moment magnitude reaches a certain degree, pulse periods from lower PGV threshold levels will be greater than those from higher ones. However, for non-strike-slip faults, in soil sitesˊ cases, it is just the other way around, while in rock sitesˊ cases, such trend is not obvious, which may be caused by limited data volume.