Seismological Understanding of Accelerogram Amplitude Scaling for Engineers with Implications to Seismic Risk Analysis

Abstract

Due to the paucity of strong recorded accelerograms, earthquake engineering analysis relies on accelerogram amplitude scaling for structural damage/collapse assessment and target spectrum matching. This paper investigates seismological characteristics of scaled accelerograms so as to inform future ground motion selection and seismic risk assessment methods. If a recorded accelerogram is scaled linearly by multiplying it with a positive factor, it is shown using the Representation theorem and the accelerogram Fourier spectrum that moment magnitude scales logarithmically and static stress drop scales linearly. Other seismic parameters such as the Joyner-Boore distance metric and the effective rupture area are invariant to scaling an accelerogram. This proposed interpretation of scaling is validated in the time as well as the frequency domains using a hybrid method for ground motion simulation. Finally, a discussion is made over the seismological correctness of accelerogram scaling. It is suggested that a suite of scaled accelerograms can be considered as being seismologically correct if this suite's magnitude given rupture area and stress drop distributions are similar to empirical observations.