Tack hammer source test

Tests of a small hammer source on a fractured outcrop of the Palisades sill
(See field notes for detailed descriptions of source-receiver geometry and acquisition notes.)

Linear array record section

This is an unfiltered record section from shots 91__. See NOTES for geometry. Amplitudes are TRACE equalized.

The trigger times are suspect on these very short time scales. A rough velocity however was obtained from the slope of the first arrivals. Using the arrival time at distances of 0.1 and 2.2 meters, a velocity of 1750 m/s was obtained. It is still unclear what type of wave this is however.

Spectral characteristics

These are overlayed spectrums from the unfiltered record sections shown above. The amplitude scale is linear. Notice the highly dominant frequency peak between 1900-2000 Hz. This is likely to be a resonant frequency of the geophone. This ringing is not present before the impulsive signal but continues to be strong after the signal decays away.

Also note that this frequency does not seem to shift with the source distance. The frequencies between 3000-4000 Hz do seem to decay with distance (Pink shades represent close sources, dark shades are further away)

Depth of source comparison
   CROSS SECTION

[phone]       [1]
*****************[2]
*****************[3]
*******ROCK******[4]
*****************[..]
This is are unfiltered records from shots 92__. These tests display the effect of having the impulsive source on the surface and at shallow depths on the side of the rock outcrop. See NOTES for geometry. Amplitudes are WINDOW equalized.
Source repeatability


These records are from files 94__. They demonstrate the repeatability of this source. They are single shots at a distance of 2.3 m. Seismograms have NOT been trace equalized. Though there may be a static time offset due to the trigger mechanism, these plots show that the waveform is is quite consistant from shot to shot. Stacking would further improve this.

Influence of cracks

An interesting observation was made on two similar records. Both records at right are from the same geophone with a small hammer source at a distance of 40 cm. Both are stacks of five shots. They have not been trace equalized and no filters have been applied. The only known difference is that the bottom source-receiver path (blue spectrum) crossed a sizeable crack. On the surface this crack was 1 cm across. Its geometry is unknown below the top cm or two.

Something quite different is happening to the early part of each signal. The signal crossing the crack is severely attenuated during its early arrival but during later portions of the signal. Not surprisingly, it is this first part of the signal which contains the highest frequencies. Spectrums of the full waveform of each signal (window equalized, logarithmic scales) were made. The blue spectrum corresponds to the lower seismogram - the one which crossed the crack.

4.7-6.2 ms 6.2-7.8 ms 7.8-9.4 ms

The waveform was analyzed into three blocks of similar size corresponding to the beginnin, middle and end of the wave. The spectrums of the middle and end are very similar. However, during the first part of the signal, the path crossing the crack showed a marked attenuation of higher frequencies. One explanation is simply that the crack (or the soft material filling it) acts like a low pass filter. This is intuitive but we do not know how deep the crack penetrates. A second explanation could work if the crack is shallow. If this wave is a surface wave than a shallow crack would most affect higher frequencies. Lower frequencies decay more slowly with depth and could pass under a shallow crack.
Conclusions