变网格间距速度层析成像方法

Variable grid spacing velocity tomography

变网格间距速度层析成像,是指反演网格模型中采用不同的网格间距反演地下地层的速度分布,它可以在观测数据较少的情况下,提高浅部地层速度反演的分辨率和深层速度反演的精度。变网格间距与规则网格间距速度层析成像方法类似,但有两个最主要的不同之处:射线正演网格单元的射线段被划分到相应的较大反演网格单元中,平滑约束方程在变网格间距界面的上一平面和下一平面的网格处需要做出适当的修改。文中首先采用变网格方法对二维和三维检测板模型进行恢复测试,反演结果对大、小异常体的刻画效果均比规则网格方法好;然后,采用变网格方法对实际资料进行走时反演,反演结果对两个低速异常区的刻画比规则网格方法的效果好。因此,变网格方法反演的速度剖面分辨率较高,尤其对浅部地层的速度成像效果具有较大的改善作用;另外,在观测数据较少的情况下,通过调整不同网格的数量比例,变网格反演方法还可以克服层析成像过程中解的欠定问题。

In variable grid spacing tomography the underground velocity distribution is parameterized with model cells of different sizes.This method can simultaneously take into account the spatially varying resolution inherent in most datasets.Due to experimental design or logistic constraints,the shallow and deep subsurface velocity distribution may be very differently determined by the available data.The variable grid spacing tomography and regular grid spacing tomography are similar in most other aspects.There are two main differences between the variable grid spacing method and the regular grid spacing method.First,ray segments calculated in the forward model cells should be divided into the larger cells of the inversion model.Second,the smoothness constraint equations needed to inhibit wild velocity variations need to be modified where the cells change in size.In a synthetic checkerboard reconstruction test with differently sized checkers,the variable grid spacing method recovers the small and large checkers better than the regular grid spacing method in both two and three-dimensional test cases.For a real dataset,the variable grid spacing method distinguishes two low velocity zones better than the regular grid spacing method.Finally,it can be concluded that when the effort is spent to identify the regions in a model with best data coverage.The variable grid spacing method can produce velocity images with higher fidelity than when a uniform cell size is used.Especially,in many cases this method can enhance the fidelity of the shallow subsurface velocity distribution.In addition,variable grid spacing tomography can reduce the underdetermined regions in an inverse problem when the data coverage is irregular by adjusting quantitative proportion of cells with different sizes.