Estimation of lithologic parameters from seismic data using genetic algorithm

The present study proposes an alternate method based on genetic algorithm (GA) to estimate the subsurface lithologic parameters such as P-wave velocity, the S-wave velocity and the density for subsurface earth layers occurring at a particular location. These are useful parameters to discriminate lithology and help in detecting hydrocarbons from seismic data. However, estimation of the lithologic parameters is an inverse problem which is highly nonlinear and non-unique and therefore requires use of some sort of global optimization method. The present study focus on GA based optimization method which offers significant advantages over other existing nonlinear optimization algorithms. The effectiveness of GA in estimating lithologic parameters have been illustrated considering synthetic seismic data which is generated using one-dimensional ray-tracing method. Results of simulation at different levels of noise demonstrate that the proposed GA based method can successfully estimate the seismic parameters under significant levels of noise and adequate convergence can be achieved.