![Sample our Earth Sciences journals, sign in here to start your access, latest two full volumes FREE to you for 14 days]({"type":"image" , "src" : "/sda/5930/TOC-Subject-Sample-2021-Earth-Sciences.jpg"})
Abstract
Advances in computational capabilities as well as ongoing improvements in storage strategies have made reverse time migration (RTM) a feasible method for capturing images of complex structures. However, large storage requirements still restrict RTM applications, especially in anisotropic media. Utilising a first-order quasi-P-wave equation in vertically transversely isotropic (VTI) media, we investigate anisotropy and deduce an RTM equation for a staggered-grid high-order finite difference (FD) scheme incorporating a perfectly matched layer (PML) boundary in this study. We also develop an improved source wavefield storage strategy via a PML boundary method for VTI medium RTM using graphic processing unit (GPU) accelerated computation. Our proposed method significantly reduces the total volume of data storage required for conventional RTM while increasing calculation time by just a small amount. Checkpoints can be set based on GPU memory size, leading to the generation of high precision and high efficiency subsurface images. We carried out a series of numerical tests on simple anisotropic media and complex Hess 2D VTI models to verify the effectiveness of our proposed method.
The reverse time migration (RTM) equations for staggered-grid high order finite difference scheme incorporating a perfectly matched layer boundary for vertically transversely isotropic (VTI) media are proposed, and checkpoints and GPU accelerated techniques are utilised for data storage and computation efficiency. Hess 2D model tests demonstrate the effectiveness of the proposed algorithm.
Acknowledgements
This research was supported by a Project of the National Natural Science Foundation of China (grant nos 41574117, 41474118), the Excellent Youth Foundation of Heilongjiang Province, China (grant no. JC2016006) and The Northeast Petroleum University Excellent Scientific Talent Fund (grant no. GLJHB201601).