Gabor有色反褶积

Colored Gabor deconvolution

Gabor反褶积假设反射系数是白噪,而实际反射系数序列具有蓝谱特征,如果把白谱作为期望输出,则将丢失薄层的细节信息。将Gabor反褶积与蓝色滤波两种技术融合,将蓝谱(根据实际反射系数序列得到)作为期望输出,实现地震记录的拓频处理。在实现过程中,采用变步长采样方式(即对双曲网格进行等面积剖分),对能量集中的低频区域加密采样,对能量较弱的高频区域稀疏采样,加快统计计算效率。模型仿真测试和实际资料应用表明,Gabor有色反褶积能较好地反映地层的细节信息,效果优于常规反褶积。

Gabor deconvolution assumes that the reflection coefficient is white noise,but real reflection coefficient sequences have blue spectral characteristics.If we take white spectrum as expected output,it is possible to lose details of thin layers.Therefore we propose here a colored Gabor deconvolution by fusion of Gabor deconvolution and blue filtering.Taking blue spectrum(based on the real reflection coefficient sequence)as expected output,we can achieve seismic frequency extension processing.In data processing period,we use variable-step sampling(equal area partition for hyperbolic grid)to improve statistical efficiency,because it conducts encrypted sampling in the area with low frequency and sparse sampling in the area with high frequency.Model tests and real data applications show that the proposed colored Gabor deconvolution can better delineate geological details than Gabor deconvolution.