Surface Wave Dispersion Analysis Using Time-Frequency Filtering and an Interactive Normal Move Out (NMO) Tool with Uncorrelated Garner Valley Vibroseis Data

Wednesday, 17 December 2014
Neal Edward Lord1, Herbert F Wang1, Dante Fratta1, Chelsea Lancelle1, Jonathan Alexander Baldwin1, Robert L Nigbor2, Athena Chalari3 and Ethan Castongia4, (1)University of Wisconsin Madison, Madison, WI, United States, (2)University of California Los Angeles, Los Angeles, CA, United States, (3)Silixa Ltd., Hertfordshire, United Kingdom, (4)ConocoPhillips Company Anchorage - COP, Anchorage, AK, United States
Frequency swept seismic sources (vibroseis) recorded by seismic arrays can be used to directly measure the surface wave apparent velocity as a function of frequency. Time-frequency filtering (TFF) passes only a narrow frequency band centered on the time varying frequency of the seismic source. This greatly improves the quality of the uncorrelated seismic data by removing noise outside of the narrow band filter. Sources of noise include traffic, harmonics generated by the seismic source and the propagating seismic waves. TFF also allows the separate analysis of the source fundamental frequency and each harmonic. The filtered data is used with an interactive normal move out (NMO) tool to adjust the time and apparent velocity parameters to extract a surface wave dispersion curve.

This technique was used on the vibroseis data collected in the September 2013 seismic experiment conducted at the NEES@UCSB Garner Valley field site. Three vibroseis sources were used: a 45 KN shear shaker, a 450 N portable mass shaker, and a 26 KN vibroseis truck. They were recorded by two lines of 1 and 3 component accelerometers and geophones, and a Distributed Acoustic Sensor (Silixa-iDAS™) system connected to 762 m of trenched fiber optical cable in a larger rectangular area.