Synthesis Method of Sound Which Has Same Duration Time as a Seismogram

Abstract

Listening to a sound representing ground motion is valuable to understand an earthquake intuitively compared with only seeing a waveform. In this paper, we propose a new method to make a sound of earthquake, in which several rules different from existing methods are used. In our new method, a sound of earthquake is generated by modifying the frequency information of generating function based on the theory of symmetric Fourier analysis. Applying the method to the JMA Kobe seismogram in 1995 South Hyogo Prefecture Earthquake and predicted waveform at Sannomaru, Nagoya, in future Tokai and Tonankai earthquake, three features of the sound generated by this method are confirmed: the sound has same duration time as the seismogram, high sound is heard when the seismogram has high frequency, and loud sound is heard when the seismogram has large amplitude.