【摘要】 为了解决在长时间和复杂结构的地震波场数值模拟情况时的数值频散问题,现在辛算法的基础上,主要结合通量校正传输(FCT)、褶积微分算子、完全匹配层(PML)等数值模拟技术,寻找一种更为优秀的地震波场数值模拟方法.地震波场的数值模拟结果表明,辛算法不仅具有保持体系结构的特性,并且具有长时间跟踪能力,具有很强的数值模拟稳定性;FCT方法基于通量守恒原理,压制网格频散效果明显;褶积微分算子突出了空间微分的局部属性.因此,通过合理应用各种技术,可以对地震波场特征进行更精确的数值模拟.更多还原

【Abstract】 To find out a solution of numerical dispersion in long-time tracing and complicated structure modeling,based upon the symplectical algorithm,this paper introduced a 2D seismic wavefield modeling technique which combins with the Flux-corrected Transport method(FCT) and convolutional differentiator,and uses the Perfectly Matched Layer(PML) as a absorbing boundaries.The numerical modeling shows that the Symplectical algorithm not only has a structure-preserving feature but also has a long-time tracing ability,it has a good stability during numerical modeling.Due to adopting convolutional differentiator and FCT techniques,the spatial feature is enhanced,and the numerical dispersion caused by the gridding is suppressed to produce a more accurate seismic wavefield.So we can get a better result with combins these technique reasonable.更多还原