Kidd Creek Multi-Level Crossline VSP Experiment

Acquisition of multi-level crossline VSP data for BH-5171 was completed between June xx-yy. The data were acquired by firing dynamite charges (500 g) into a total of 12 receiver levels in the borehole, between depths of .. to ++ 
(Fig. ). The receiver levels were chosen so that data from the previous crossline experiment could be merged with it, to provide the uppermost trace in the first 70 source gathers. Recording conditions were favourable, and the equipment (a 12 channel Bison recording system operated by the UofA) performed
very well. Consistent source signatures were achieved for each shot position,
verifying the quality of the data and simplifying subsequent processing and
interpretation steps.
For the methods and equipment used in the survey, these results indicate that
summer acquisition is preferable to winter acquisition in this environment.



Preliminary processing of the data has been undertaken, and the
results so far look very encouraging. Processing steps implemented to date have included:
1) data reformat, 2) merge new data into a single file (145 MByte), 3) add trace header
information, 4) sort into common source gathers, 5) pick first arrivals on
surface traces. Intermediate processing steps will
essentially follow the procedures used to process the VSP data from phase II
of the IPP program, and will include: 1) correction for shot delays, 2) rotation of horizontal components, 3) removal of the downgoing wavefield, 4) semblance
and frequency filters. The goal of the intermediate processing steps will be
to produce clean images of the upgoing (scattered) wavefield for each
receiver level. Final processing steps will be designed to image the position
of scattering bodies in 3-D. One possible approach to this, making use of the
directional information from particle-motion analysis, is outlined below.

Fig. .. shows a plot of the first three common source gathers from the new
data. There is a 20 ms delay for the second receiver level (which is also
visible in the surface traces) that will be accounted for in the first step
of the intermediate processing sequence. Otherwise, first breaks are exceptionally consistent, and the signal-to-noise level appears to be very
high. There is no visible evidence for 60 Hz contamination in these records.

The average amplitude spectrum for the traces displayed in Fig. xx is shown
in Fig. yy. For the most part, the amplitude spectrum lacks the sharp 
spikes at 60 Hz and harmonics, which plagued the previous survey. Useable energy
continues to beyond 500 Hz, implying a theoretical resolution of 3 m or better.
However, in practice, it may not be possible to achieve this level of resolution.

Estimation of Absolute Particle-Motion Direction from crossline VSP data

Where three-component seismic measurements are available, particle-motion studies can be used to infer
the propagation direction for recorded body-wave arrivals. This information may be particularly useful in
a complex structural setting, where it may not be obvious (on the basis of traveltime information alone)
from which direction a scattered signal is incident. Here, particle motion directions for the side-scanning
experiment are estimated for the direct-P arrival, the direct-S arrival, and a reflected event recorded
between 0.1784 s and 0.2516 s from trace 1 to trace 70.

In order to estimate true particle-motion directions it is necessary to establish an absolute frame of
reference for the three-component receiver at a depth of 450 m. A deterministic estimate for geophone
orientation was obtained by assuming that the first motion for the direct P arrival should be directed away
from the source. For trace 25, the source was as assumed to be 300 m due north of the receiver. Inspection
of the first particle motion for the direct P arrival in the vicinity of trace 25 showed that the H1 channel
was oriented approximately NNE (first motion negative), the H2 channel was oriented WNW (first motion
weakly positive) and the V channel was oriented downwards (first motion positive). The processing steps
applied to the data to obtain absolute particle-motion measurements were as follows:

1. Bandpass filter (40-50-140-150 Hz).
2. Determine rotation angle to minimize the energy on the H2 channel in the first-break window (±5 ms)
for trace 25. This was estimated to be counterclockwise 20o.
3. Determine scaling factor for both horizontal channels so that the mean particle motion in the first-break
window (for trace 25) points toward the source. This was found to be 1.16.
4. Apply transformation determined in steps (2) and (3).
5. Calculate average particle motion direction in a 10 ms window for all three arrivals.

The results from step (5) are plotted in a stereographic projection in the accompanying figures. For the
direct P wave, the particle motion vectors (squares) intersect the upper half plane. A systematic shift is
seen from NW-directed motion for traces 1-10 to slightly east of north for traces 51-60. Polarization
estimates for traces 61-70 show considerable scatter. Similar results are seen for the direct-S wave, but
with predominantly horizontal particle motion.

Three possible scenarios are considered for the reflected event, as constrained by the particle motion
directions. First, we note that the particle motion is almost completely horizontal. If the reflection is a P
wave, the direction of propagation must therefore be horizontal. Similarly, if it is an SH wave, the
propagation direction must be horizontal; if it is an SV wave, the propagation direction is vertical. We also
note that the orientation swings from N-S for traces 1-10, to NW-SE for traces 31-40. Past trace 40, there
is much more scatter in the data. The known near-vertical dip for structures in this area suggest that the
reflection is unlikely to be an SV wave, since this would require a nearly horizontal reflector below the
receiver. If the reflection is interpreted as a P wave, the particle motion directions can be thought of as
approximate dip vectors; if it is interpreted as an SH wave, they indicate the approximate strike direction
of the reflector.