动态地震子波估计

目前常用的地震子波估计方法有确定性子波估计和统计性子波估计,各有其优缺点.2类子波估计方法都是基于传统的褶积理论模型,即子波在传播过程中保持波形不变,没有考虑子波在传播过程中的动态性.实际上由于地层对子波具有吸收作用,子波在传播过程中逐步衰减,子波传播是一个动态衰减过程.改进静态褶积模型提出动态褶积模型,从而估计动态子波.假设地层反射系数为白噪序列,震源子波在地层中传播形成衰减地震道,对此地震道作小波变换时频谱分析,采用箱状平滑对衰减地震道频谱作平滑,去除反射系数的影响,从而获得子波振幅谱,再假设子波为最小相位,通过希尔伯特变换求得子波相位,从而实现动态子波估计.理论模型分析与实例计算验证了方法的可行性.

Dynamic seismic wavelet estimation

determinant wavelet estimation and statistical wavelet estimation are two commonly-used methods at present, both of which have excellence and disadvantages. Conventional convolution model believes that wavelet pattern remain the same during propagating process thus ignores the dynamic feature of wavelet, the two method introduced here are base on the conventional convolution model, but practically proved that wavelet would be absorbed by stratum during propagating process, and the wavelet pattern was fading, which means that the propagation process is actually a dynamically fading process. Stationary convolution model is improved and dynamic convolution model is proposed to estimate the dynamic wavelet. Supposing the reflectance is white noise, the wavelet traveling in the stratum will form an attenuating seismic channel, analyzing the seismic trace in time-frequency, and smooth it with boxcar smoother to obtain wavelet spectrum. Wavelet phase could be obtained through Hilbert transform of wavelet spectrum, and then dynamic wavelet estimation could be accomplished. The feasibility of these methods have been proved by analysis on theoretical models and practical calculations.