When it comes to seismic design of structures shear wave velocity V_s belongs to the basic characteristics of the soil column at any building site. Shear wave velocity can be used for analysis in various ways. Either we have an average value for the upper layers of the crust or we deal with a shear wave velocity profile – a fitting velocity for each actual layer. In the first case the so-called V_s,30 – the average shear wave velocity of the upper 30 meters is a generally accepted description, used in building codes worldwide (e.g. EN, ASCE, etc.). The latter is used when conducting more advanced analyses, generally applying a discretization method by which correct wave velocities can be set for each soil layer. Velocity profiles also allow for ground motion amplification and liquefaction predictions.

The determination of the shear wave velocities is a major task in a microzonation study. There exist various methods to determine the shear wave profile at a location. Downhole methods, which belong to the intrusive group, are widely applied. Downhole shear wave examination is a variation of the seismic refraction method employing a downhole tri-axial geophone and a surface energy source or a downhole source. The energy propagates from the source to the geophone and can be recorded as compression and/or shear waves. The data can be used to evaluate the compression wave and shear wave velocities of the subsurface soils versus depth. The disadvantage of the method is the requirement of a borehole, which may limit its application especially in urban areas. Therefore various surface methods have been developed, based on the idea that the dispersive properties of surface waves can be used to infer near-surface elastic properties. The common feature of surface methods is that they use an array of sensors to pick up the vibrations of the site. These tests are commonly referred to as array measurements. From the data the f-k-spectrum (frequency-wave number spectrum) also called dispersion curve is extracted. The methods differ especially in the numerical methodology of dispersion curve extraction from tests. Additionally, the test setup might be altered to suit the requirements of the analysis technique.

One of these surface methods is the spatial autocorrelation (SPAC) method, which likewise hypothesizes that ambient noise is stationary both in time and space, and it is composed of surface dispersive waves. Spatial autocorrelation is the correlation of a variable with itself through space. To support this, a relatively large number of sensors is required to measure simultaneously.
A widespread alternative is the spectral-analysis-of-surface-waves (SASW) method. The profile of the material tested is a stack of horizontal layers. All field testing is performed at small-strain levels. The SASW uses only two sensors, however, requires multiple spacings to obtain the necessary range of wavelengths. Afterwards the phase velocities are extracted from the frequency spectra.
The harmonic-wavelet-analysis-of-waves (HWAW) is another method to receive dispersive characteristics of a soil site. This method relies on a test with two sensors and an impact. The extraction is based on the harmonic wavelet transform (HWT), which is a time-frequency analysis. It basically expands the time signal into wavelets at a range of frequencies. The phase velocities at various frequencies are then computed from the time-frequency information.

Our aim is to adopt and implement the HWAW method as the basic tool to determine dispersive characteristics of a soil site. Afterwards shear wave velocity profile is computed by an inversion procedure which requires mathematical modelling of the stack of soil layers and an optimization algorithm. The complete set of software tools needed for shear wave velocity identification is under development.