Research And Application Of Desert Seismic Noise Properties Based On Kendall Rank Correlation Coefficient

Seismic exploration is an important mean for exploring oil,gas and mineral resources.With the increase of the country's demand for oil and gas resources and the improvement of exploration technology,the target of seismic exploration has been expanded to desert areas containing rich in oil and gas.Due to the large area,the changeable sand dunes,and the complex air pressure,the seemingly simple desert area produces random noise with complex properties.Affected by environmental differences in surrounding areas and collection equipment,desert random noise data is a low-frequency colored noise with non-Gaussian,non-stationary,and non-linear characteristics.The noise waveforms between adjacent traces look similar,even have weakly similar structural features to seismic signals in some areas.This kind of complex random noise obscures the information of the effective signal,and even seriously interferes with the signal identification and information extraction.Suppressing random noise in desert seismic data is very important for improving the signal-to-noise ratio of seismic data and identifying and extracting effective information about underground structures.Among them,understanding the characteristics of random noise is a prerequisite for improving the denoising technology.In order to analyze and measure the weak spatial similarity of desert background noise,a block-based spatial Kendall rank correlation coefficient(SKRCC)is proposed in this paper based on the Kendall correlation coefficient.This method calculates the rank correlation coefficient between adjacent blocks,and considers more time and space information in the rank correlation analysis.Therefore,the SKRCC is effective and powerful for the accurate measurement of desert background noise.In this paper,the spatial similarity of desert background noise is statistically analyzed using SKRCC,and it is found that the desert background noise has weak spatial similarity.In addition,this paper also studies the relationship between spatial-similar desert background noise,block size,and wind speed.The analysis of experimental data with different block sizes shows that the value of the SKRCC calculated from desert background noise is smaller than the value of the signal.It can be found that SKRCC based on a suitable block size can enhance the spatial difference between desert background noise and the signal.At the same time,the spatial similarity of desert noise at different wind speeds is also studied by using block scanning,and found that there is a directional characteristic of desert background noise with spatial similarity in the desert noise data.The study of noise properties is able to improve the noise reduction ability.An adaptive complex shock diffusion filtering algorithm based on SKRCC rank correlation is further proposed in this paper,which is applied to desert noise suppression.This method uses the SKRCC to measure the spatial correlation between the desert background noise and the signal.The diffusion coefficient based on the SKRCC is designed.In this way the diffusion is conducted along the texture structure of the seismic data,and different diffusion intensity is applied in the signal and desert background noise simultaneously.It can enhance the signal and increase the suppression intensity of desert background noise.In this paper,the effectiveness of the algorithm is verified by using the simulated and the real seismic data.The filtered results show that the complex shock diffusion filter based on SKRCC(CSKRCC)can not only achieve effective suppression of desert background noise,but also retain complex structural information in desert seismic data.