scholarly journals Acoustic finite-difference modeling beyond conventional Courant-Friedrichs-Lewy stability limit: Approach based on variable-length temporal and spatial operators

2021 ◽  
Vol 34 (0) ◽  
pp. 1-14
Author(s):  
Hongyu Zhou ◽  
◽  
Yang Liu ◽  
Jing Wang ◽  
Geophysics ◽  
2011 ◽  
Vol 76 (4) ◽  
pp. T79-T89 ◽  
Author(s):  
Yang Liu ◽  
Mrinal K. Sen

Most finite-difference simulation algorithms use fixed-length spatial operators to compute spatial derivatives. The choice of length is dictated by computing cost, stability, and dispersion criteria that are satisfied globally. We propose finite-difference schemes with adaptive variable-length spatial operators to decrease computing costs significantly without reducing accuracy. These schemes adopt long operators in regions of low velocity and short operators in regions of high velocity. Two methods automatically determine variable operator lengths. Dispersion analysis, along with 1D and 2D modeling, demonstrates the validity and efficiency of our schemes. In addition, a hybrid absorbing boundary condition helps reduce unwanted reflections from model boundaries. Our scheme is more efficient than those based on variable-grid methods for modeling, migration, and inversion of models with complex velocity structures because the latter require local grid refinement, which usually increases memory requirements and computing costs.


Geophysics ◽  
2018 ◽  
Vol 83 (3) ◽  
pp. T139-T157 ◽  
Author(s):  
Shigang Xu ◽  
Yang Liu

Current temporal high-order finite-difference (FD) stencils are mainly designed for isotropic wave equations, which cannot be directly extended to pseudoacoustic wave equations (PWEs) in tilted transversely isotropic (TTI) media. Moreover, it is difficult to obtain the time-space domain FD coefficients for anisotropic PWEs based on nonlinear dispersion relations in which anisotropy parameters are coupled with FD coefficients. Therefore, a second-order FD for temporal derivatives and a high-order FD for spatial derivatives are commonly used to discretize PWEs in TTI media. To improve the temporal and spatial modeling accuracy further, we have developed several effective FD schemes for modeling PWEs in TTI media. Through combining the [Formula: see text] (wavenumber)-space operators with the conventional implicit FD stencils (i.e., regular-grid [RG], staggered-grid [SG], and rotated SG [RSG]), we establish novel dispersion relations and determine FD coefficients using least-squares (LS). Based on [Formula: see text]-space operator compensation, we adopt the modified LS-based implicit RG-FD, implicit SG-FD, and implicit RSG-FD methods to respectively solve the second- and first-order PWEs in TTI media. Dispersion analyses indicate that the modified implicit FD schemes based on [Formula: see text]-space operator compensation can greatly increase the numerical accuracy at large wavenumbers. Modeling examples in TTI media demonstrate that the proposed FD schemes can adopt a short FD operator to simultaneously achieve high temporal and spatial modeling accuracy, thus significantly improve the computational efficiency compared with the conventional methods.


2016 ◽  
Vol 26 (3/4) ◽  
pp. 698-721 ◽  
Author(s):  
J I Ramos

Purpose – The purpose of this paper is to both determine the effects of the nonlinearity on the wave dynamics and assess the temporal and spatial accuracy of five finite difference methods for the solution of the inviscid generalized regularized long-wave (GRLW) equation subject to initial Gaussian conditions. Design/methodology/approach – Two implicit second- and fourth-order accurate finite difference methods and three Runge-Kutta procedures are introduced. The methods employ a new dependent variable which contains the wave amplitude and its second-order spatial derivative. Numerical experiments are reported for several temporal and spatial step sizes in order to assess their accuracy and the preservation of the first two invariants of the inviscid GRLW equation as functions of the spatial and temporal orders of accuracy, and thus determine the conditions under which grid-independent results are obtained. Findings – It has been found that the steepening of the wave increase as the nonlinearity exponent is increased and that the accuracy of the fourth-order Runge-Kutta method is comparable to that of a second-order implicit procedure for time steps smaller than 100th, and that only the fourth-order compact method is almost grid-independent if the time step is on the order of 1,000th and more than 5,000 grid points are used, because of the initial steepening of the initial profile, wave breakup and solitary wave propagation. Originality/value – This is the first study where an accuracy assessment of wave breakup of the inviscid GRLW equation subject to initial Gaussian conditions is reported.


Geophysics ◽  
2020 ◽  
Vol 85 (2) ◽  
pp. T57-T70 ◽  
Author(s):  
Yang Liu

Time-space domain finite-difference modeling has always had the problem of spatial and temporal dispersion. High-order finite-difference methods are commonly used to suppress spatial dispersion. Recently developed time-dispersion transforms can effectively eliminate temporal dispersion from seismograms produced by the conventional modeling of high-order spatial and second-order temporal finite differences. To improve the efficiency of the conventional modeling, I have developed optimal variable-length spatial finite differences to efficiently compute spatial derivatives involved in acoustic and elastic wave equations. First, considering that temporal dispersion can be removed, I prove that minimizing the relative error of the phase velocity can be approximately implemented by minimizing that of the spatial dispersion. Considering that the latter minimization depends on the wavelength that is dependent on the velocity, in this sense, this minimization is indirectly related to the velocity, and thus leads to variation of the spatial finite-difference operator with velocity for a heterogeneous model. Second, I use the Remez exchange algorithm to obtain finite-difference coefficients with the lowest spatial dispersion error over the largest possible wavenumber range. Then, dispersion analysis indicates the validity of the approximation and the algorithm. Finally, I use modeling examples to determine that the optimal variable-length spatial finite difference can greatly increase the modeling efficiency, compared to the conventional fixed-length one. Stability analysis and modeling experiments also indicate that the variable-length finite difference can adopt a larger time step to perform stable modeling than the fixed-length one for inhomogeneous models.


Sign in / Sign up

Export Citation Format

Share Document