Results of seismic reflection lines at California Wash Fault and the Astor Pass Tufa Tower. May 6, 2011 GPH492

Transcription

1 CA Wash Results Results of seismic reflection lines at California Wash Fault and the Astor Pass Tufa Tower May 6, 2011 GPH492 A bandpass filter of Hz was used to process the data. This filter did not provide good enough clarity and very low amplitudes at zero offset, where we would like to see reflections. Additionally, in the vicinity of the air- wave/ Rayleigh wave data was highly aliased resulting in a checkerboard pattern. This area of high amplitudes was also characterized by a steep dip. To help remove the steeply dipping features and boost the amplitude of flatter features a dip filter was used. This helped remove the aliased checkerboard pattern, and high amplitude steeply dipping features. The result was a more cohesive product, which was more easily used to pick stacking velocities. Plane 22. After bpfilter. On left no dip filter, lots of aliasing and high amplitude steep dips. On right, subdued steep dips, reflections identifiable at zero offset, and less aliasing. Initial analysis of the data consisted of picking single reflections and calculating NMO- velocity, frequency, wavelength, and simple depth based on time 0 at zero offset. Picking was not consisitent, and no single plane contained easily recognized reflection traces. The results of the simple analysis were inconclusive. These results were from multiple planes/shot- points and therefore the Dix interval velocities are not representative. The cvstack process was completed using stacking velocities of 500 m/s to 3,000 m/s. Picks were made at velocities generating consistent reflections. Velocites were

2 generally 1,100 to 1,600 m/s. Based on these picks a cmpstack was generated and is below. Velocity models from cvtstack picks. CA Wash cmpstack (W- E cross- section, time scale is s) Two first order observations are of note. 1) shallow reflections (about 0.05 s two- way travel time) are truncated at the east end of the line, and to the east low amplitudes are non- descript. 2) Two deeper reflections (at about 0.1 s two- way travel time) appear coherent at the east and west end of the lines. These features are generally consistent with the location of the high velocity (>3,500 m/s) regions in the SeisOpt at 2D refraction velocity model. a. Based on NMO velocities the depth to the deep reflector is 80 m. At the east end of the line these observations are generally consistent with a) the truncations of shallow beds at the location of the fault scarp and b) the deeper reflection is not apparent immediately west of the main scarp. Combined this line suggest that the deeper reflection is absent from the west side of the scarp because it has been down- dropped. The presence of the deep reflector at the west end of the line also suggests a graben structure, however additional work is necessary with the refraction velocity model to confirm that the high velocity region is real, and not an edge effect. In addition, a depth of 80 m is 2-3 times deeper than the high velocities depicted in the refraction analysis (20-30 m). This is not consistent and I doubt the accuracy of the reflection.

3 These results are not concrete by any means, and a more intensive initial analysis is necessary to discern more reflections to pick during cvstacking. The Dix velocity model shows a very generalized velocity model that lacks picks throughout the entire stack. Additionally, the use of the dipfilter has the potential to generate false data in the final result. It is necessary to complete and compare cvstacks and cmp stacks with, without, and with variations of the dipfilter to confirm that aberrations are not being introduced into the data. Astor Pass Results A bandpass filter of Hz was used to process the data. This filter did not provide good enough clarity and very low amplitudes at zero offset, where we would like to see reflections. Additionally, in the vicinity of the air- wave/ Rayleigh wave data was highly aliased resulting in a checkerboard pattern. This area of high amplitudes was also characterized by a steep dip. To help remove the steeply dipping features and boost the amplitude of flatter features a dip filter was used. This helped remove the aliased checkerboard pattern, and high amplitude steeply dipping features. The result was a more cohesive product, which was more easily used for identifying reflections. These images are similar in nature to the above examples from CA Wash. Initial analysis of the data consisted of picking single reflections and calculating NMO- velocity, frequency, wavelength, and simple depth based on time 0 at zero offset. Picking was not consisitent, and no single plane contained easily recognized reflection traces. In addition to difficulties picking reflections in on plane, reflectors were not easily traced from plane to plane. This implies that reflections have a variety of dips at small scales, shallow velocity gradients are high, or a combination of both. The results of the simple analysis were inconclusive. These results were from multiple planes/shot- points and therefore the Dix interval velocities are not representative. Cvstacking generated very few if any continuous reflections. On the first order this is a result of the wavy, inconsistent presence of reflections on each plane. So it is not surprising that cvstacking did not allow us make pick velocities. Based on a fairly common and expected velocity of 1,200 m/s one reflector was identified. Based of that velocity and its two- way travel time its depth was 120 m. This is generally consistent with the drilled depth to the Tertiary volcanic section m deep. The cvstack is below.

4 West- East cross- section at 1,200 m/s stacking velocity. Arrow pointing to reflection at depth comparable to sediment depth. We have included a cmpstack based loosely on the refraction results. This does not work well, especially since NMO- velocities are needed to stack and refraction results are loosely analogous to Dix interval velocities.

5 West- East cross- sections: top cmpstack loosely based on the velocity model (bottom). Time scale is s (top). The reflection results at Astor Pass, with the processing completed here, do not add any insights into the presence of a tufa tower fault. Generally speaking, the processing tools, time and knowledge base utilized were not equipped to extract precision, uniform data analysis from the survey. The major point of contention was two- fold. The amount of wind during the survey necessitated turning up gain on the recorder to capture data at long offsets, thus introducing and capturing excess noise. Second, the undulatory nature of reflections in the data do not lend themselves to be nicely stacked at one velocity. In otherwords, the reflections were dipping wilding (or similar) and the plain concept of trying to straighten them is both impractical and precisely obscures their true nature.