Effect of data uncertainty and inversion non-uniqueness of surface wave tests on VS,30 estimation

The effect of data measurement uncertainty and inversion uncertainty on VS,30 estimation is studied in this article. An attempt has also been made to separate out the inversion and data measurement uncertainties, and their effect on VS,30 is quantified separately. A relaxation to the misfit value is introduced to study the consequence of inversion uncertainty, whereas data variation up to one standard deviation is considered to study the effect of data uncertainty. All possible profiles are generated within this limiting misfit value (after relaxation) in case of inversion uncertainty and within measured data uncertainty bound in case of data uncertainty. Finally, few profiles are selected from entire range of generated profiles for VS,30 estimation. For the profiles falling short of 30 m depth, shear wave velocities are extrapolated using a simplified method proposed by Boore (2004) [4]. The VS,30 values show limited variation in terms of CoV (< 1%) as a results of inversion uncertainty. Although higher variation is observed in CoV (1.7 - 3.4%) values in the case of data uncertainty, site classification as per NEHRP seems to be unaffected.