Noise Attenuation in Shallow Holes (III)

Improvement of Signal to Noise Ratio

Abstract

The present observations and a nalyses form the third and final experiment for the study on “Noise Attenuation in Shallow Holes ”(T. HIRONO, S. SUYEHIRO, M. FURUTA and K. SATO,1968 and 1969), the first and second experiments having been made at the Meteorological Research Institute, Tokyo and the Onahama Weather Station on the Pacific coast of Northeastern Japan, respectively.
The final test site was at the Niigata Meteorological Observatory on the Japan Sea coast of Northeastern Japan, where the geological founda tion is very weak and the microseisms are very large in winter. The final depth of the bore-hole was 200 m, and observations were made at depths of 26,45,105,146 and 200 m. The instrumentation and method of analyses were basically the same as in our previous experiments. The results are summarized in the following:
1) The observatory premises are in a recently reclaimed ground. The high frequency portion of both background noise and earthquake motions are already much absorbed by the soft foundati on. The shortlived man-made noise, which is generated locally and mainly of high frequencies, is also attenuated very rapidly by depth for the same reason.
2) The background noise in Niigata predominates at a frequency band of 4 to 2.5 cps at the surface. The amplitude of this freq uency band, however, becomes very small in the ground by a rapid a ttenuation (see Figs.2,3 and 4).
3) In our previous tests, almost no attenuation was observed. even at a depth of more than 100 m in a frequency band of microseisms. In the present experiment, however, the amplitudes of a frequency band of 0.4∼0.25 cps is attenuated down to about half at a depth of 200 m. This is probably because of the surface layer being very soft ground, which was recently reclaimed (see Fig.3)
4). Amplitudes of earthquakes become smaller with increasing depth, and this effect is larger at higher frequencies as recognized in our previous experiments (see Fig.5).
5) The improvement of signal to noise ratio, which depends on the depth attenuation of both background noise and earthquake motion, is greater with increasing depth. At a depth of 200 m, the theoretical lower limit of earthquake detectability is expectantly lowered by 0.5 in M. If it is taken into account that the seismogram interpretation is made much easier because of the elimination of confusing background noise, the lower limit is effectively lowered by 1 in M as shown by a routine test (see Fig.7).
6) Due to large microseisms in winter, it is very difficult to identify earthquakes in seismograms. To eliminate those disturbing microseisms, a low-cut (F_L=1 cps) was connected. Clear seismograms were obtained, which are free both from artificial background noise and natural microseisms (see Fig.8)1.
7) By the three experiments since 1967, the bore-hole seismograph has been proved effective in improving the signal to noise ratio for local earthquakes and already, employed in the JMA routine operation. It has also been shown that this noise attenuation becomes much greater when the, depth reaches a certain hard layer. To study the local geology near the surface is therefore important to bury at an optimum depth.