【摘要】 品质因子Q值是地震勘探的重要物理参数,可用于提高地震资料处理质量和油气藏描述精度。本文提出了一种新的基于整形正则化反演和S变换的Q值估计方法。首先通过S变换得到不同旅行时刻的振幅谱,然后引入整形正则化算子稳定求解不同时刻的频谱比值,最后通过线性拟合估计出地震品质因子。该方法通过正则化技术消除了噪声、反射相干以及变换窗口等引起的算法不稳定问题,得到光滑稳定的频谱比值,有利于最终的线性回归求取Q值。与传统的频谱比值法相比,基于整形正则化技术的方法估计的Q值更为准确。模拟算例和实际地震资料处理结果均表明,该方法可以有效稳定地估计地震资料品质因子。更多还原

【Abstract】 The seismic quality factor Q is an important physical parameter in seismic exploration.The important information for hydrocarbon exploration can be provided by the factor Q.In this paper,we present a novel method for Q estimate using S-transform with regularized inversion.We first use S-transform to obtain the amplitude spectrum at different given travel time.When using the spectral ratio method,we do the division with shaping regularization,which constrains the ratio to have a desired behavior,such as smoothness.It can avoids the instability of division,which might be caused by noise,interfering reflectors,windowing distorts,etc.We use synthetic examples and a field data example to verify our method.These examples show that this method can estimate Q in stable way and can obtain more precise Q value compared with conventional method.更多还原