基于汉宁窗函数滤波时间域高斯束成像方法

Time domain Gaussian beam migration method based Hanning window filtering

实际地震数据中通常存在背景噪声及人为干扰信号,传统射线类方法对低信噪比数据成像适应性较差,基于此本文发展了一种适应于低信噪比数据的时间域高斯束成像方法.相比于频率域类束方法,新方法的主要优点包括:(1)频率域高斯束偏移方法需计算每个成像点处各个频率的格林函数,基于时间域高斯束偏移方法无需上述过程,大大提高了计算效率;(2)时间域高斯束公式推导中,直接对地震记录进行局部倾斜叠加并加入汉宁窗滤波处理,对低信噪比数据具有更好的适应性.在实现成像方法的基础上,本文通过对不同信噪比Marmousi炮记录进行模型试算及含噪声实际资料成像试处理表明:时间域高斯束偏移方法能够较为明显地提高低信噪比数据的偏移效果,提高成像方法对实际数据的适应性.

Actual seismic data usually contains background noise and artificial coherence signal. The traditional ray methods have bad imaging effect with low Signal to Noise Ratio（SNR） data.Accordingly,we developed a time domain Gaussian beam migration method which is adapted to low SNR data. Compared with the frequency domain beam migration,the advantages of our method are as follows： firstly,the frequency domain migration methods have to calculate the Green function of every frequency at each imaging point,however the time domain Gaussian beam migration method does not need the process. This improves efficiency greatly;secondly,we directly accomplish the work of partial local slantstack,meanwhile Hanning window filtering process is adopted in the time domain formula derivation,which has better adaption for low SNR data. Based on the implementation of migration,we tried to compute several different SNRs common-shot records of Marmousi model and image real noisy seismic data. The results verify that the time domain Gaussian beam migration method obviously improve the imaging efficiency of low SNR data and enhance the compatibility of imaging method for actual data.