Rocky Mountain (53rd) and South-Central (35th) Sections, GSA, Joint Annual Meeting (April 29–May 2, 2001)

Paper No. 0
Presentation Time: 1:00 PM-5:00 PM

DATA IMPROVEMENT BY SUBTRACTION OF HIGH-AMPLITUDE HARMONICS FROM SEISMIC DATA: IMPLICATIONS FOR INTERPRETATION OF CD-ROM REFLECTION DATA


SHOSHITAISHVILI, Elena1, JOHNSON, Roy A.1 and MIKHAILOV, Oleg V.2, (1)Department of Geosciences, Univ of Arizona, Tucson, AZ 85721, (2)Chevron Petroleum Technology Co, 6001 Bollinger Canyon Rd, Building D, San Ramon, CA 94583, elenas@geo.arizona.edu

Where seismic data is recorded in populated areas, it is common to have a high-amplitude 60 Hz noise present in the data. This noise is produced by power transmission lines and other equipment and often has amplitude orders of magnitude higher than the amplitudes of desirable signals in the seismic data. Any single geophone records a sum of the high-amplitude harmonics coming from all sources in the surrounding area. Since the frequency of the noise produced by different equipment is not exactly the same, the "60 Hz" noise is a combination of frequencies, and contaminates the data together with its sub-harmonics. Notch filtering, which works well for eliminating a single problem frequency, is not effective on data that contain multiple problem frequencies.

We used a time-domain-based method of automatic estimation of noise frequencies and their amplitudes, followed by subtraction of these estimated anomalous harmonics from the data. Since the technique estimates the best fit of noise for the entire trace, subtraction of the noise preserves the actual signal. Filtering reflection data from the northern CD-ROM transect across Archean and Proterozoic terranes in southern Wyoming and northern Colorado in this way significantly improved data quality even in areas where first breaks were completely obscured by noise. A new analysis of the improved first breaks from CD-ROM multicomponent reflection data provides a more accurate near-surface velocity model and better estimates of rock properties based on P- and S-wave arrivals. This filtering also improves our ability to correlate the seismic data with geological exposures and provides more robust three-dimensional information about deeper events.