Simulated Annealing Algorithm for Subsurface Shear Wave Velocity Investigation Using Ground Vibration Data

Subsurface shear wave velocity plays an important role in designing earthquake-resistant structures. Average shear wave velocity up to 30 m depth, known as V_s30, is used as a common design parameter. Ambient noise data that is generated by ground vibrations due to the passing of vehicles or other passive sources carry an important information about the subsurface shear wave velocity structure of the region. The seismic instrument records these disturbances along with three different directions i.e. two horizontal and one vertical. The ratio of the spectral amplitude of the horizontal to the vertical component is an important factor that is dependent on subsurface velocity structure of the investigating area. Ground vibration data from ambient noise recording system has been recorded at Bayasi site in the Garhwal Himalaya. A computer code DISHV in FORTRAN has been developed in this work for obtaining HV curve having a limited data from a large sampled HV curve having similar spectral property. The obtained output from DISH has been effectively used as an input to the HV-Inv software developed by Garcia et al. (Comput Geosci 97:67–78, 2016). The HV curve obtained from recorded data is modelled with a simulated annealing algorithm. The obtained velocity model from Simulated Annealing algorithm is further refined using forward modelling approach to match obtained and theoretical HV curve.