论文部分内容阅读
根据炮点和检波器空间展布规律的差异,大多数地震数据处理过程可分为2-D、2.5-D、弯线或3-D。本文介绍的层析法能克服这些观测系统上的困难,并通过拾取的初至波时间求取稳定的静校正解。拾取的初至波中包含了短波长和长波长的地表静校量。本文通过使用加权最小二乘法和共轭梯度法解广义的地表一致性时间延迟方程来计算这些静校量。迭代中,要对每个拾取的初至进行评估以保证最小二乘解的一致性。基于一致性,我们对拾取的旅行时进行加权,并将它们用于下次迭代中。对用户来说,这些权值也可作为异常拾取值的一种指示。长波长解在最小二乘解中留下了较大的剩余误差。利用菲涅尔带的预期长度区分长波长和短波长静校正解。消除了长波长静校量的影响后,我们用短波长静校正法去进一步减少剩余误差。将此方法应用到一实际资料上,结果证明了该方法的有效性。消除了长波长和短波长静校量,改善了数据,使它产生了更加相符的速度,并且仅留下了短波长的剩余反射静校量,它们一般小于1/4的子波周期延时。这消除了大部分可能造成相位跳跃的原因,消除了从剩余静校计算中获得较差的速度估计,尤其是对于3-D数据。
According to the difference between the distribution of shots and the detector space, most seismic data processing can be divided into 2-D, 2.5-D, curve or 3-D. The chromatographic method described in this paper overcomes the difficulties of these observing systems and obtains a stable static correction solution from the first arrival time of the pick-up. The first arrivals picked up contain surface short-wavelength and long-wavelength surface statics. In this paper, we calculate these statics by using the weighted least square method and the conjugate gradient method to solve the generalized surface-consistent time delay equations. In iterations, the first arrival of each pick is evaluated to ensure the consistency of the least squares solution. Based on the consistency, we weight the pick-up trips and use them for the next iteration. For users, these weights can also be used as an indication of abnormal pick values. The long wavelength solution leaves a large residual error in the least squares solution. Use the expected length of the Fresnel zone to distinguish between long and short wavelength statics. After eliminating the effects of long-wavelength statics, we use short-wavelength statics to further reduce the residual error. Applying this method to an actual data, the result proves the validity of this method. Eliminating the long wavelength and short wavelength statics, improving the data to produce a more consistent velocity, and leaving only the short wavelength residual reflection statics, which are generally less than 1/4 of the wavelet period delay . This eliminates most of the possible causes for phase jumps and eliminates the need for poor speed estimates from residual statics calculations, especially for 3-D data.