低信噪比地震资料初至自动拾取技术

为提高低信噪比地震资料初至自动拾取的精度,提出了一种联合应用曲波变换与希尔伯特变换进行初至拾取的方法。首先采用基于Wrapping的快速离散曲波变换算法对检波点静校正和线性动校正后的炮域数据进行曲波变换,然后根据随机噪声和有效初至信号的曲波系数在不同尺度不同方向上的分布差异,设置合适的阈值对曲波系数进行"去噪"处理,最后将处理后的曲波系数进行反变换,获得压制随机噪声后的地震记录。利用希尔伯特变换计算各地震道的瞬时振幅,然后利用改进的瞬时强度比公式逐道计算给定的时窗内各采样点的瞬时强度比,最后根据瞬时强度比极大值确定单炮的初至时间。理论数据和实际资料处理结果表明,对于低信噪比地震资料,利用曲波变换法进行去噪后,数据的信噪比得到了提高。结合具有一定抗噪能力的改进型瞬时强度比初至拾取方法,可以有效地提高自动拾取初至的精度,减少人工修改错误初至的工作量,该方法具有一定的实用价值。

Automatic first-arrival picking from seismic data with low signal-to-noise ratio

Automatic first-arrival picking from seismic data with low signal-to-noise ratio(SNR)is a challenging task because of the random noise generated by activities in the environment.To address this issue,a method combining a curved wave transform and the Hilbert transform was proposed.First,a fast discrete curvelet transform(FDCT)algorithm based on wrapping was employed for the curvelet transform on the shot gathers,after receiver static correction and LMO.Next,according to the distribution differences of the random noise and the effective first arrival in different scales and directions,an appropriate threshold value was set to‘denoise’the curve coefficient.Then,multiple thresholds,depending on the scale and noise variance,were applied to suppress the curvelet transform coefficients,and the inverse curvelet transform was applied to reconstruct the uncorrupted data.Finally,the Hilbert transform was applied,so as to calculate the instantaneous strength of each trace.The instantaneous intensity ratio of each sample within a given time window was calculated using the modified formula with respect to instantaneous intensity ratio,and the first arrival was identified as the time when the instantaneous intensity ratio reached the maximum value.Tests on both synthetic and real data showed that the proposed method can effectively remove the random noise while preserve the first arrival.In combination with the Hilbert transform,the first arrival can be accurately picked from seismic data with low SNR.