Microseisms associated with winds blowing across a lake, appear as ground noise in the range 1 - 2.5 c/s. These transient "Lake Microseisms" obscure the true geothermal ground noise.

Geothermal ground noise surveys can be adversely influenced by transient seismic phenomenon. Microselsms occurring from lakes produce major transients that may be difficult to identify without sufficient record. This paper is based upon observations made upon the shores of Lake Taupo, North Wand, New Zealand in 1966. The same seismoniter tape recorder unit sited at kinloch, about 3.2 Kilometres from the lake shore, and later at Acacia Bay. about 800 metres from the lake shore (Figure M. local microseisms were recorded at both sites. An analysis of these events and a comparison with meteorological data from the Wairakei Meteorological Observatory show that these microseisms were closely related to the magnitude of the wind.

Records taken during the period 25 February to 19 March 1966 at the Kinloch and Acacia Bay sites were studied, and chart records of the overall amplitude in the frequency range 0.1 cps to 15 cps were plotted and frequency analysed. The frequency spectrum analysis samples (Figure 2) of which many were made, all showed a very much higher amplitude than the geothermal ground noise. These varied in length from 1 hour to 18 hours, which was the longest wind storm recorded during this study, and coincided with the meteorological data of the wind storms.
The frequency spectrum and the average amplitude of the local microseisms were not greatly influenced by the direction of the wind, although microseisms with low frequencies were always accompanied by winds from the South. Frequency spectrum, average amplitude. and duration of the microseisms showed a relationship with the wind speed. For example, seismic disturbances of short duration with frequencies of around 2.0 c/s (Hz) were in most cases accompanied by winds, with speeds less than 5 knots. Microseismic events which lasted longer than three hours with frequencies between 1.0 and 2.5 cps were in each case accompanied by winds with speeds greater than 10 knots. At the onset of the microseism higher frequencies are dominant; lower dominant frequencies occur later. The time lag between the beginning of the microseismic disturbance and the occurrence of the lowest frequencies in the frequency spectrum (around 1.2 cps) is about 3 to 6 hours. A peculiarity in the spectra is that signals with a frequency of about 1. 5 cps are low in energy.

Published in "The Geothermal World Directory" 1973. Page 56.
A relationship between average amplitude of the microseism and average wind speed is indicated. (Figure 3) especially during the period 8 to 18 March, 1966. but peaks of high winds do not always correspond with peaks in the average high microseism ground noise. Volcanic, earthquake. and man made noise can be recognized by vertical noise columns in (Figure 2a, b, and c). The persistent horizontal less dominant lines represent geothermal ground noise. (Figure 3) compares the recorded noise levels in the frequency range 0.1 to 15 cps and meteorological data. This noise includes. besides local events, microseisms generated by the ocean (frequencies below 1 cps).


The comparlson between stations suggests, that more than wave action on the shore is responsible for the lake microseisms. Winds from the South produce lower frequencies possibly because the Wind has to cross the lake with consequent longer wave actions, whereas winds from other directions start closer to where the seismic stations are located. North of the lake. The onset action on the water has to be considered. It appears to produce the observed higher frequencies for all but the South winds. especially during the earlier part of the wind storms. To accurately identify this phenomenon and to insure a correct identification of true geothermal ground noise; it would appear essential to record by the direction of the wind, although microseisms before and after, as well as during, the wind Storms.


The conclusions drawn have been corroborated by proprietary surveys in the vicinity of Lago de Chopala Mexico, when the author was with the University of Mexico. Lago, de Chopala has the same surface area as Lake Taupo in New Zealand. There have also been recorded wind microseisms several days long generated by the Great Salt Lake in Utah, and also in the Salton Sea California, as well as other lakes in the U.S.A.

NOTE Using the listening technique., the wind microseisms were easy to identity. Then they could be excluded from the seismic tremor. Ocean microseisms am much lower frequency and out of the seismic tremor range. It is worth noting that the record when played back sounded like the doppler effect. Another good reason for using the listening technique.

FIGURE 3 Relative amplitude of microseisms at Kinloch. Wind speed in knots at Wairakei, from the 25th of February to 18th of March 1966

FIGURE 1. Map showing location of Lake Taupo, North Island New Zealand, and location of Seismic Station and Meteorological station.
FIGURE 2 Frequency analysis of three microseisms recorded at Kinloch between 26th February and 19th March, 1966.

White represents the seismic signal. The white below the 1c/s level is very low frequency . Using the listening technique, these phenomena sound like the doppler effect, which identifies these events easily.

Go to Download page for seismic audio files and discussion by G.R.T. Clacy

Return to main page