基于振幅谱互相关的薄层厚度计算

Thin bed thickness calculation from spectral correlation

反射波振幅和峰值频率两种属性当前被广泛应用于预测薄层砂体厚度。然而,实际地震资料的振幅通常不保幅会降低振幅计算薄层厚度的可靠性;峰值频率法计算薄层厚度具有泰勒展开近似前提,会降低其计算薄层厚度的精度。为了提高薄层厚度估计的精度,提出了一种利用振幅谱互相关函数计算薄层厚度的新方法。首先,构建目的层的振幅谱与子波振幅谱和反射系数褶积后振幅谱之间的相关系数函数,其最大值所对应的地层厚度为目的层砂体的厚度;然后,利用高斯迭代方法求取相关系数函数最大值,从而计算出薄层厚度。通过楔状体模型测试证实新方法是可行、有效的。相较于峰值频率方法,新方法能够更精确地计算薄层厚度。实际资料的应用进一步证实新方法适用于薄层厚度计算。

The reflection amplitude and peak frequency,two basic attributes,are widely used in predicting the thickness of thin bed.The real seismic data without amplitude-preserved sometimes and the peak frequency obtained by some approximation methods will reduce the accuracy for thin bed thickness calculation using the reflection amplitude or the peak frequency.In order to improve the accuracy of thin bed thickness calculation,we present a new method for calculating thickness from spectral correlation.For a thin bed,the correlation coefficient is examined between the amplitude spectrum of the wavelet multiplied by a reflection response filter and that of a target thin bed.The thin bed thickness can be determined when the correlation coefficient reaches the maximum which can be searched by the Newton iteration scheme.A wedge model test indicates that the proposed method is valid for calculating thin bed thickness.By contrast to the peak frequency method,the new proposed method is accurate for thin bed thickness calculation.Moreover,application in a field data further shows that the proposed new method is effective and practicable for the thin bed thickness estimation.