INVESTIGATION OF SUBSURFACE VELOCITY STRUCTURE USING THE GENETIC ALGORITHM

Abstract

The Kumaon I-limalaya is of utmost seisimctectonic importance due to the very high seismic activity and elaborate thrust system in the region. During the last 100 years, this region has suffered from 14 Earthquakes of magnitude 6.0 or more. This dissertation deals with the study of subsurface velocity structure of Bageshwar district in the Kumaon Himalayas. Traditionally, the seismic risk study focuses on the seismic hazard estimation. But, over the years another factor of site amplification has been found to be as important as the seismic risk estimation. The site amplification effects in the area predominantly depend on the soil compaction. The soil compaction study requires the knowledge of the subsurface shear wave velocity structure. This velocity structure can then be used to classify the soil in the region in number of classes, depending upon the compaction levels. A large number of passive and active source methods are available to measure the Vs profile. But most of these methods require special instruments and are expensive to carry out. This dissertation makes use of a relatively fast and inexpensive technique to estimate the shear velocity profile based on the 1-lorizontal to Vertical spectral ratio (H/V) of ambient noises recorded at the different stations. In this technique, the observed l-lVSR are fitted with the synthetic HVSR generated using the trial and error forward modeling technique. This dissertation makes use of the GA-Haskell method to achieve its purpose. The GA-Haskell method uses SR waves transfer function to model I-I/V spectral curves, which are then constrained by the Genetic Algorithm so to get the best fit. For this dissertation, we also undertook the fieldwork in the month of October so as to collect the data necessary for the completion of this dissertation work. llere, we have presented the velocity structures obtained from the Bageshwar station ambient noise data and the one collected along the profile line. This dissertation also talks about the challenges we face with the active source data along the profile and why the results obtained for such data are not that precise. Finally, we end this dissertation, explaining how we can collect the data along the profile so to able to infer the lateral shear wave velocity structure changes along the profile.