Reverse time migration (RTM) is capable of imaging the underground structures even in complex geological conditions. However, RTM does not guarantee a migration result with accurate amplitude and high resolution. Therefore, Least-Squares RTM methods are developed to reduce migration artifacts, balance the amplitudes, and improve the resolution. For computational considerations, implementing LSRTM in the image-domain is preferred, which comes down to deblurring the RTM result against the point spread functions (PSFs). Traditionally, the PSFs are acquired by demigration-migration on the impulse-scatter model, which alone takes twice the computation of RTM. And to avoid interference, the impulse scatters cannot be sampled densely enough. In this paper, we deduced an analytic expression of Hessian/PSF under the WKJB-approximation. And with it we are able to explicitly calculate the PSFs at any sampling distance, without interference concerns. Based on those PSFs, the result from a common RTM can be deblurred to approach the accuracy of LSRTM. In our numerical tests, this method exhibits flexibility and efficiency.
Presentation Date: Wednesday, October 14, 2020
Session Start Time: 1:50 PM
Presentation Time: 3:30 PM
Location: Poster Station 8
Presentation Type: Poster