Recent Experience Using Active Love Wave Techniques to Characterize Seismographic Station Sites

Abstract:

Active-source Love waves recorded by the multi-channel analysis of surface wave (MAS_LW) technique were recently analyzed in two site characterization projects. Between 2010 and 2011, the 2009 American Recovery and Reinvestment Act (ARRA) funded GEOVision to conduct geophysical investigations at 189 seismographic stations—185 in California and 4 in the Central Eastern U.S. (CEUS). The original project plan was to utilize active and passive Rayleigh wave-based techniques to obtain shear-wave velocity (V_S) profiles to a minimum depth of 30 m and the time-averaged V_S of the upper 30 meters (V_S30). Early in the investigation it became evident that Rayleigh wave techniques, such as multi-channel analysis of surface waves (MAS_RW), were not effective at characterizing all sites. Shear-wave seismic refraction and MAS_LW techniques were therefore applied. The MAS_LW technique was deployed at a total of 38 sites, in addition to other methods, and used as the primary technique to characterize 22 sites, 5 of which were also characterized using Rayleigh wave techniques. In 2012, the Electric Power Research Institute funded characterization of 33 CEUS station sites. Based on experience from the ARRA investigation, both MAS_RW and MAS_LW data were acquired by GEOVision at 24 CEUS sites—the remaining 9 sites and 2 overlapping sites were characterized by University of Texas, Austin. Of the 24 sites characterized by GEOVision, 16 were characterized using MAS_LW data, 4 using both MAS_LW and MAS_RW data and 4 using MAS_RW data. Love wave techniques were often found to perform better, or at least yield phase velocity data that could be more readily modeled using the fundamental mode assumption, at shallow rock sites, sites with steep velocity gradients, and, sites with a thin, low velocity, surficial soil layer overlying stiffer sediments. These types of velocity structure often excite dominant higher modes in Rayleigh wave data, but not in Love wave data. At such sites, it may be possible to model Rayleigh wave data using multi- or effective-mode techniques; however, in many cases extraction of adequate Rayleigh wave dispersion data for modeling was difficult. These results imply that field procedures should include careful scrutiny of Rayleigh wave-based dispersion data in order to collect Love wave data when warranted.