Application of Non-uniform Coherent Noise Suppression Technique in MTZ Area

Hydrocarbon source rocks of MTZ survey deep Yingcheng formation in Shuangcheng depression in the eastern part of Songliao Basin are widely developed, and the combination of vertical upper reservoir cover has a good exploration prospect, which is an important replacement field for deep oil and gas resources in Daqing Oilfield. High-density 3D seismic acquisition was deployed in this survey in 2018, with serious variations in seismic acquisition observation system, high-energy multi-stage refraction wave development on single shot record with particularly low SNR, and the conventional linear noise attenuation technique is difficult to suppress the coherent noise of irregular sampling based on the assumption of regular observation system. Basing on the time-frequency analysis of noise classification and partition, we design a fan filter in F-X domain. The least square method is used to estimate the weighted coefficients of all the frequency components in each trace, to simulate the noise model of the weight value in reverse performance of the real position, and finally to suppress the coherent noise by adaptive matching subtraction, we refer to the above denoising method as non-uniform coherent noise suppression technique. For the actual seismic data processing in the MTZ survey, the coherent noise in the variable-observation system area can be suppressed accurately without the need of pre-stack data regularization, the processing results have good amplitude-preserving, high profile imaging accuracy, clear structural interface and wave formation characteristics. This technique provides high-quality seismic data for the optimal selection of oil and gas exploration targets for small faults on Songliao Basin.