Wave Equation Forward In FWI

With the surface conditions and subsurface structures of exploration area increasinglycomplex, how to improve the accuracy of seismic inversion, and how to obtain more accurateseismic reservoir parameters has become a serious problem. Therefore, full waveforminversion method emerges. It rebuilds structure of the formation by using both kinematics anddynamics information of pre-stack seismic data. Compared with the traditional method ofseismic inversion, the former is more accurately reflect the geological structure and reservoirinformation of complicated geological conditions.Wave field forward modeling is the basis of FWI, and it plays a vital role in the FWI.Wave equation forward numerical simulation is based on the wave theory, and can reservekinematics and dynamics characteristics of shock. It is an important means to understand theseismic wave propagation and can provide a theoretical basis for the identification of seismicwave field and also is the basis of pre-stack or post-stack inverse time depth migrationimaging. Numerical simulation technology almost has been widely used in seismic dataacquisition, processing, interpretation and development of oil and gas reservoir. With theexploration tasks more and more complex and exploration accuracy increasingly high, weneed study this further to help us have more awareness of propagation of seismic waves in thecomplex exploration targets, and solve the thorny problems which we are facing in modernoil and gas exploration and development work.This paper studies the two-dimensional acoustic, elastic wave high-order staggered gridfinite difference numerical modeling technique to solve accuracy, stability, absorbingboundary conditions problem of wave equation forward numerical simulation. Based onprevious research, we discuss the precision error sources, the mechanism of instability, andboundary reflection mechanism in detail. And we combined improved PML absorbingboundary conditions with FCT flux correction, calculate wave field information. This methodimproves the precision of numerical simulation, and weakens the numerical simulationdispersion problems. Numerical simulation results show that the algorithm has high accuracy,waveform characteristics is easy to identify, measure process is stable, and no boundaryreflections and so on.