Study Of The Aeolian Environment System In Land Seismic Exploration

Seismic exploration is the main method of gas and oil exploration. However,in the actual seismic data acquisition process, the desired signal will be inevitablycontaminated by random noise, and the signal-to-noise ratio of seismic recordswill also be reduced by random noise. In order to suppress random noise inseismic exploration effectively, we need to understand the characteristics andgeneration mechanism of random noise. Seismic exploration ambient noise is themain component of random noise, and wind noise is the main component ofambient noise. Therefore, we need to study the generation mechanism of windnoise, namely, we need to study the Aeolian Environment System.We take the exploration area topography、surface conditions、ground surfacegeological conditions and the coupling degree between the detector and groundsurface as a system, namely, the Aeolian Environment System. By analyzing theAeolian Environment Systems of different regions, we can understand howtopography and ground surface conditions impact on random noise in seismicexploration. It can provide a basis to increase processing capacity of randomnoise. Therefore, we need to solve the Aeolian Environment System first. We useinverse filtering theory to solve the Aeolian Environment System and conduct apreliminary study in this paper.In the paper we study from two aspects: First, since the wind speed data isunknown when gathering seismic data, we use Fractional Brownian motionmethod to simulate the wind sequence in this paper. There are many commonmethods to achieve Fractional Brownian motion, such as: Passion Faulting,Successive Random Addition, Midpoint displacement and so on. Because themidpoint displacement method is the most basic and most commonly usedmethod and it has high computational efficiency, we use it to model the windsequence. The spectrum of the wind sequence simulated by Fractional Brownianmotion has a higher fitting degree with the theoretical spectrum Davenportspectrum, which verifies the wind reliability simulated by Fractional Brownian motion.Assuming that noise before primary arrival is entirely produced by the wind,and that the wind sequence is input of the Aeolian Environment System. Thususing deconvolution (inverse filtering) method can reverse the seismicexploration Aeolian Environment System. The so-called inverse filtering is theinverse process of a filtering process. We do forward modeling in this paper inorder to verify the feasibility of the Aeolian Environment System. First, for agiven system, fractal sequence as input, we can obtain output by convolving them.Then, we reverse system when Ricker wavelet and fractal sequence are used asinput respectively and the simulated output above is used as output. The result isthat the obtained system that fractal sequence is used as input has little differencewith the given system, but the obtained system that Ricker wavelet is used asinput has great difference with the given system. Thus the result verifies thevalidity of the model used. Finally, we use noise before primary arrival of theactual seismic record of different exploration areas as output, the simulated windsequence as input, and we obtain the actual seismic exploration AeolianEnvironment System of different areas by multiple iterations. We also analyzesystem characteristics. The result shows that the Aeolian Environment System hasa great relationship with terrain, and systems of forest area and desert areapproximately AR model, systems of loess plateau and mountain area areapproximately ARMA model.Through this study, it indicates that noise before primary arrival is relevantto the wind、topography and other comprehensive factors of the survey area, it isa beneficial discussion on generation mechanism of seismic exploration randomnoise, it is also a complement of existing understanding about seismic explorationnoise and provides a basis for noise cancellation technology.