Spectral decay parameter (?) analysis for a single fault source and an automatic selection procedure for spectral flattening frequency (fx)

Abstract

Lately the spectral decay parameter, kappa, which essentially represents the decay of the high frequency ground motion amplitudes, is highly utilized and also debated in the scientific community due to the uncertainties and the inconsistencies of different evaluation processes. To reduce source-based uncertainties, a dataset with a good spread of magnitude variety and near as well as far site recordings is needed. Furthermore, the selection of the frequency where the exponential decay of the acceleration spectrum becomes flat, f(x), should be performed based on a consistent procedure to minimize the contamination introduced by the noise spectrum. In this study, to reduce the source effect on kappa evaluations, only the events generated from Sivrice-Puturge segment of East Anatolian Fault are considered. This also provided a regional Q correction that is not influenced by different source generated events during the estimation of the zero epicentral spectral decay parameter, kappa(0) for stations under consideration. An automated selection procedure for f(x) is developed which yielded a good fit to the dataset. The stations were also examined for the peak frequencies of horizontal-to-vertical spectral ratios (HVSR) whether these peak frequencies intercepted with the frequency band considered in the kappa evaluations. The HVSR calculations additionally provided information on the soil properties at the station locations. The findings are considered to be useful for understanding the effect of kappa versus epicentral distance (R-epi) fit on the estimation of station-based kappa(0) values. Furthermore, the proposed procedure for obtaining spectral flattening frequency (f(x)), may improve compatibility among kappa evaluations, and reduce selected frequency band related uncertainties.