Acoustic full‐waveform inversion using least‐squares method with active constraint balancing technique

Abstract

We have developed a full-waveform inversion algorithm using spatially variant damping factor and normalized wavefield for surface seismic data. The Active Constraint Balancing (ACB) technique which is used to obtain spatially variant Lagrangian multipliers satisfies both high resolution and stable inversion. We also used the normalized wavefield to overcome the lack of precise knowledge of the source. To verify our inversion algorithm, we tested proposed algorithm to the synthetic anticline model and SEG/EAGE overthrust model. Through comparison of two inversion results with and without ACB, we could present that the inversion results using ACB method converge faster and more stable under the circumstance of insufficient data than those by the GaussNewton method. In addition, our inversion algorithm could works successfully without the source information, because the the source spectrum is eliminated during the normalization procedure.