Study of the horizontal-to-vertical spectral ratio (HVSR) method for characterization of deep soils in the Mississippi embayment.

摘要

Soil deposits can significantly influence the amplitude and frequency content of surface ground motions during earthquakes. Estimating the fundamental frequency (f0) of a site is often needed for improved planning and design for future earthquakes. A cost-effective method of obtaining an estimate of f0 is the Horizontal-to-Vertical Spectral Ratio (HVSR) method (also termed Nakamura's method), which utilizes ambient energy recorded in the horizontal and vertical directions from a single, three-component sensor. In addition to estimating the fundamental frequency, average shear wave velocity (VS,AVG) values have been estimated using the HVSR method and a simple approximate relationship relating VS,AVG to f0 and the depth to bedrock. This procedure was performed by Bodin et al. (2001) to develop a relationship between VS,AVG and soil depths in the Mississippi embayment. However, this relationship predicts average velocity values that are about 25% higher than values predicted by another relationship developed using a different method (Chen et al. 1996). Although this inconsistency is known, the relationship of Bodin et al. (2001) is often cited as the best information on the velocity structure in the embayment. In addition, Bodin et al. (2001) identified a second frequency peak of unknown origin in their HVSR plots. The objectives of this study are to answer these unresolved questions, specifically: (1) why the shear wave velocity relationships developed for the Mississippi embayment using the HVSR method are inconsistent with other findings, and (2) the origin of the second frequency peak observed in HVSR plots from the Mississippi embayment. To meet these objectives, the following three studies were performed: (1) a parametric study of site factors influencing the fundamental frequency and average velocity estimates from the HVSR method, (2) an investigation into the influence of the HVSR processing parameters using experimental data collected at eleven deep soil sites in the Mississippi embayment, and (3) comparison of experimental and simulated HVSR results for Mississippi embayment sites. With regard to the first objective, it was found that the HVSR method yielded reliable values of the fundamental site frequency for conditions of high velocity contrast between soil and rock and saturated soil conditions (conditions that are met in the Mississippi embayment). Also, it was shown that varying the HVSR processing parameters had a negligible impact on the HVSR frequency estimates. However, it was demonstrated that use of the approximate method to estimate VS,AVG systematically over-predicted the true VS,AVG values. With regard to the second objective, it was shown that the second frequency peak observed in the experimental HVSR plots can be attributed to either a higher-mode resonance of horizontally-polarized shear waves reflecting from the soil/bedrock boundary or local site resonance due to a shallow contrast in VS within the soil deposit. Based on the findings of this study it is recommended that the VS relationship for the sediments of the Mississippi embayment developed by Chen et al. (1996) should be preferred to the Bodin et al. (2001) relationship.