论文部分内容阅读
明确表达了用局部相关法在数据域层析成像的不适合标准。与基于差分的反演相比,基于相关的反演对局部极小值敏感度较低。可是相关值却常常在同相轴之间受到串音的影响,另外,由于是沿整个时间轴求和,所以全局相关值仅仅给出一个模型运动误差的一般概念。作为一种选择,Gaussian时窗局部相关法是能够提取模拟数据和观测数据之间不适合中的局部运动误差值的。与全局相关法相比,局部相关法对地震同相轴的串音敏感度也较低,原因是作为一个时间函数只需进行局部求和。因为低相关串音造成清洁伴随震源,所以产生清洁梯度值。采用与地震资料带宽一致的补偿函数改善梯度值,与用于消除无限带宽资料的线性补偿函数相比,该补偿函数更具有现实意义。
Explicitly expressed by the local correlation method in the data tomography imaging unsuitable standards. Compared with the differential-based inversion, the correlation-based inversion is less sensitive to the local minima. However, the correlation values are often affected by crosstalk between the in-phase axes. In addition, because they are summed along the entire time axis, the global correlation values give only a general concept of model motion error. As an alternative, the Gaussian time-window local correlation method is able to extract the value of the local motion error that is not suitable between the simulated data and the observed data. Compared with the global correlation method, the local correlation method is also less sensitive to the crosstalk of seismic events because the local summation is only required as a function of time. A clean gradient value is generated because of the clean associated with hypocenters due to low correlation crosstalk. Using a compensation function consistent with the bandwidth of the seismic data to improve the gradient value, the compensation function is more practical than the linear compensation function used to eliminate the infinite bandwidth data.