Author/Authors :
Atefi, Sanam Department of Geophysics - Science and Research Branch - Islamic Azad University, Tehran , Heidari, Reza Department of Geophysics - Science and Research Branch - Islamic Azad University, Tehran , Mirzaei, Noorbakhsh Institute of Geophysics - University of Tehran , Siahkoohi, Hamid Reza Institute of Geophysics - University of Tehran
Abstract :
This study introduce a new frequency parameter called sfcwt, which can be used to estimate earthquake magnitude on the basis
of the first few seconds of P-waves, using the waveforms of earthquakes occurring in Japan. This new parameter is introduced
using continuous wavelet transform as a tool for extracting the frequency contents carried by the first few seconds of P-wave.
The empirical relationship between the logarithm of sfcwt within the initial 4 s of a waveformand magnitude was obtained. To
evaluate the precision of sfcwt, we also calculated parameters smax
p and sc. The average absolute values of observed and
estimated magnitude differences (jMest Mobsj) were 0.43, 0.49, and 0.66 units of magnitude, as determined using smax
p , sc,
and sfcwt, respectively. For earthquakes with magnitudes greater than 6, these values were 0.34, 0.56, and 0.44 units of
magnitude, as derived using smax
p , sc, and sfcwt, respectively. The sfcwt parameter exhibited more precision in determining the
magnitude of moderate- and small-scale earthquakes than did the sc-based approach. For a general range of magnitudes,
however, the smax
p -based method showed more acceptable precision than did the other two parameters.