| Abstract Submission No. | ABS-2.4-0200 |
| Title of Abstract | Basis Pursuit aided seismic sparse-layer reflectivity inversion: A case study from KG Basin, India |
| Authors | Suman Konar*, Maheswar Ojha, Kalachand Sain |
| Organisation | National Centre for Polar and Ocean Research, Vasco da Gama, Goa |
| Address | National Centre for Polar and Ocean Research Vasco Da Gama, Goa, India Pincode: 403804 E-mail: konarsuman2@gmail.com |
| Country | India |
| Presentation | Poster |
| Abstract | A conventional sparse-spike inversion assumes that seismic trace is represented by a convolution of a sparse series of spikes - representing subsurface reflectivities and a wavelet. In contrast, Basis Pursuit Inversion (BPI) is an approach which assumes that subsurface reflectivities can be decomposed into a series of odd and even pairs of spikes with varying thicknesses. A kernel dictionary matrix is required for the purpose of decomposing the reflectivities. This dictionary matrix consists a set of odd and even pairs of spikes with thicknesses may vary from a minimum of one sample rate to a maximum of a tuning thickness. BPI will then find a minimum combination of odd and even pairs of spikes from the dictionary matrix which can be used to represent the seismic trace. We have investigated the performance of the technique on both synthetic wedge model and real seismic data from Krishna-Godavari (KG) basin. Results show that BPI is superior in resolving thin bed problems than sparse-spike technique. This can be observed from the closed similarity of the obtained reflectivities from the well log data. |
| Keywords | Sparse-Spike inversion, BPI, dictionary matrix, tuning thickness, wedge model |