Because reflection surveying uses arrays of geophones to cancel surface waves and other horizontally propagating energy, existing pre-recorded reflection records may well yield poor results from these analysis techniques. New data may have to be taken. If only geophone strings are available, they can be set in a clump or pot, at one effective surface location. Alternatively the strings might be stretched out perpendicularly to the trend of the refraction line, thus mitigating energy propagating across the line without canceling the record of waves travelling along the line.

Another important component of this experimental setup was the use of a relatively long (15.24 m) interval between each geophone. The so-called ``takeout interval'' of the cables we used is on the long side of those already in the market, but not unusual. The use of hundreds of meters of multichannel seismic recording cable does make deployment of this technique difficult in congested urban areas, where the cable would have to be protected at every vehicle crossing. An array of independent, stand-alone recorders will be easier to deploy over a grid of streets.

Our experiments used a Bison Galileo-21 48-channel recorder. This device employs an instantaneous floating-point digitizer to 21-bit floating-point samples. Before digitization, the analog signals from the geophones pass through pre-amplifiers with configurable gain and filter settings. During tests of the system's response to low frequencies, using clustered geophones (not shown), we found that setting the higher ranges of pre-amplifier gain yielded the most coherent recording of noise and microtremor frequencies below 4 Hz. Cross-correlation tests gave correlation coefficients of 97.8±2.0% among geophones in the cluster, at 0-5 Hz when we used a 60 dB gain. Background noise did not saturate digitally at such a gain, although hammer-blow recordings would. I expect that recordings made with seismographs yielding 24-bit integer samples would also need high pre-amplifier gain to record low-frequency microtremor accurately. Seismographs producing 12 and even 16-bit integer data probably cannot record low-frequency microtremor without an analog high-cut filter.

For the velocity-spectral analysis described here I selected two records; on the basis of their geometry but not quality. Both have the 2 kg blast near the east end of the line of geophones. I derived the velocity-spectral ratio plot in the upper part of figure 5 from a record in the middle of the dry playa lake bed at the center of southern Dixie Valley. As might be expected for a reflection survey, most of the energy represented in this plot appears as body waves and air waves above 12 Hz. These waves truncate sharply in the record against the edge of the geophone array by the source, so the p-tau transform has smeared the truncations out over large ranges of velocity, with the frequency-shifting artifact. But the most significant parts of the dispersion trends are outside the aliased areas.

The reflection and gravity surveys show that playa and alluvial sediments at this location are more than 1 km deep, and floored by volcanic deposits filling an additional 1 km or more of basin depth. Figure 1, upper, shows a surface-wave dispersion trend sloping downward and to the right from 3 to 12 Hz, easily distinguished from the p-tau truncation artifacts. The phase velocities in this trend do not exceed 0.27 km/s.

Figure 1: Velocity spectral power-ratio plots of two records from the March 1998 Dixie Valley fault reflection survey. Both records had a 2 kg explosion source just off the east end of a 48 channel, 716 m linear array of the 8 Hz single refraction geophones. The ``playa'' record generating the upper plot was taken in the middle of the Dixie Valley Basin, where Quaternary alluvial and lacustrine sediments are >1 km thick. The lower ``piedmont'' record was taken with the west end of the array against the basin-bounding Dixie Valley fault, with granitic basement no more than 350 m below the east end of the array. Aliasing in the spatial frequency domain would occur below the dotted curve in each plot.
(Click here for a larger image; on the image for plot in Adobe Acrobat PDF format; or here for a monochrome PDF plot.)

The lower plot of figure 1 analyzes a record taken with the west end of the geophone array against the fault, and thus the basin boundary against granitic rock with a 4 km/s P-wave velocity. At this piedmont locality, 4 km west of the playa record, Abbott et al. (1998) showed that the granitic basement is not deeper than 350 m. In addition, no playa sediment underlies this record, only sand and gravel. The plot shows a similar surface-wave dispersion trend sloping downward and to the right from 5 to 20 Hz. At this location, however, the velocities in this trend are never less than 0.28 km/s.

Thus the velocity spectral analysis has very clearly distinguished between two sites in Dixie Valley that ought to have very different shallow S-wave velocity structure. In addition, this analysis was performed on an existing data set that met certain parameters, and did not require any additional data collection. Most existing seismic reflection surveys would not meet these requirements, having used source and/or receiver arrays to mitigate surface waves. But most refraction survey data sets would be suitable, having used single or clumped geophones and single shots.