论文部分内容阅读
在试图解决复杂地质构造问题时,使用 CDP 方法的缺陷就变得显而易见了。此外,将不同偏移距的地震道进行叠加,会扭曲用于地层分析所必需的振幅信息。为了保护数据中真实的振幅和构造信息,必须进行叠前处理。给定共偏移距数据和反射层之上的速度值,叠前声波克希霍夫反演能恢复反射界面的位置,当数据中的振幅信息被保护下来时,该方法还附带地计算出各界面点的反射系数。对于有限带宽地震数据,反演算子生成类似正弦信号段组成的反射界面图形,有限带宽奇异函数的振幅峰值等于与角度有关的反射系数。这种反演的推导是基于被插入到常规3-D 反演算子中的高频克希霍夫效据。用四维平稳相位法渐近地计算四个结果积分的方法,要选择反演振幅函数,以便使反演算子生成反映反射界面的位置、并由反射系数加权的奇异函数。把三维反演算子作一些特殊的处理,使其成为2.5-D 的算子,便可用于单测线共偏移距数据的处理。合成数据的例子说明了这种方法对常速克希霍夫数据的精确性,同时也指出了将常速数据用于多速模型时所存在的问题。
The pitfalls of using the CDP approach become apparent when trying to solve complex geological problems. In addition, superimposing seismic traces of different offsets distorts the amplitude information necessary for stratigraphic analysis. In order to protect the true amplitude and structure information in the data, pre-stack processing is necessary. Given the co-offset data and the velocity values above the reflector, prestack sonic Kirchhoff inversion can restore the position of the reflector interface and is calculated incidentally when the amplitude information in the data is preserved The reflection coefficient of each interface point. For the finite bandwidth seismic data, the inversion operator generates a reflection interface graph similar to the sinusoidal signal segment. The peak amplitude of the finite-bandwidth singular function is equal to the angle-dependent reflection coefficient. The derivation of this inversion is based on the high-frequency Kirchhoff’s result that is inserted into the conventional 3-D inversion operator. The method of asymptoticly calculating the four integrals of results using the four-dimensional stationary phase method is to choose the inverse amplitude function so that the inversion operator generates a singular function that reflects the position of the reflection interface and is weighted by the reflection coefficient. The three-dimensional inversion operator for some special treatment, making it a 2.5-D operator, can be used for single measurement line offset data processing. The example of synthetic data shows the accuracy of this method for constant-speed Kirchhoff data, and also points out the problems in using constant-speed data for multi-speed models.