Constrained Pre-Stack LSRTM: A Better Conditioned Imaging Technique in Presence of Erroneous Migration Velocity Model
B. Tavakoli F., R. Baina and V. Duprat
Event name: 81st EAGE Conference and Exhibition 2019
Session: Least-Squares RTM
Publication date: 03 June 2019
Info: Extended abstract, PDF ( 7.27Mb )
Price: € 20
We propose a new parametrization for pre-stack Least Square Reverse Time Migration (LSRTM) based on amplitude variation along redundancy axis. Contrary to conventional methods which in each shot is inverted independently from others, our formulation constrains the inverted shot-images such that amplitude variation of each imaging point along redundancy axis is parameterized based on orthogonal bases. This results in a robust subsurface imaging algorithm which is insensitive to erroneous migration velocity model and returns shot-images free from migration artifacts (such as migration smiles and artifacts from strong velocity contrasts). An application on Sigsbee synthetic model demonstrates the potential of proposed approach.