Development of a Spectrogram Analysis Tool for Seismic Waveform Data and its Application to MeSO-net for Noise Survey

We developed a spectrogram analysis tool working on Matlab to check on a seismic waveform data easily. This tool can display waveforms, power spectrum densities (PSD), and PSD spectrograms of ground motion recorded in WIN format. It can be used to evaluate a temporal variation of ground motion, and compare waveforms and spectrums for data from different channels. In this paper, we outline this tool and show its application to MeSO-net (Metropolitan Seismic Observation network) data recorded at Yayoi station (YYIM). From the analysis of whole day waveform, it is found that the noise level from AM to PM is more than ten times larger than that in the midnight because of a higher human activity, and the noise caused by train is recorded from 5 AM to 0:30 AM, which are trains’ operation hours. We then compare ground noise levels between YYIM, YYI and ASO. YYIM station is located in Tokyo University and its sensor is installed about 20m beneath the ground. The sensor at YYI is installed at the surface just beside the YYIM borehole. ASO is a permanent seismic station operated by Tokyo University, and one of the quietest sites in Kanto plain. From the analysis using this tool, YYIM (20m deep) is 10 to 20 dB quiet than YYI0 in the range more than 5 Hz, and ASO is more than 20 dB quiet than YYIM in the range lower than 20 Hz.＜論説