Study On Several Problems Of Seismic Data Interpretation System

The interpretation of seismic data is a very important step in the work of geophysics as well as an important process to transform the seismic data into geological language. It relates to whether the seismic data could perform itself effectively. Nowadays, the interpretation of seismic data is not only confined to routine structure interpretation, but more depends on seismic information, using advanced interpretation technologies, developing further research on the comprehensive analysis of layer's characteristics, oil and gas's distributing rules, and so on. Based on this, in this paper, we study and improve sereval technologies and methods to make the interpretation of seismic data more accurate. The key points are as follows:First of all, this paper is a tentative study of the Coherence Cube Technology. Seismic Coherence Cube Technology is a remarkably simple yet powerful new technology of seismic attribute analysis which is developed in recent years. It is mainly used in the interpretation of geological structure and sedimentary environment, the exploration of hidden oil and gas reserves, and so on. This paper firstly researches and summarizes the three generation algorithms of Coherence Cube Technology, then compares and analyzes their effects. In order to solve the problem of low efficiency of C3, a new coherency algorithm based on Super trace technology and Rayleigh quotient acceleration method is introduced into the calculation of C3. Experiments demonstrate that this method can preserve the effect of C3 as well as improve the computing speed.Besides, this paper focuses on the filtering of coherence cube slice. This slice is often used to interpret the variation of faults and lithology. Its quality directly relates to the result of fault interpretation, so improving the quality of slice is very important. In this paper, several typical image denoising approaches are applied. In view of the features of coherence cube slice, we use the EPS method here on the foundation of several traditional methods, and then improve it by using a new analytical model. According to experiments, this method can not only filter the noise in the slice perfectly, but also preserve the edge of faults very well.Finally, it is concerned with the filtering of seismic section. In the seismic data section, identifying and tracing the seismic reflection in-phase axis are very important steps. In order to achieve these goals, high quality of the seismic section is very necessary, that is to say, high SNR and high resolution are needed. In this paper, according to the features of the seismic data section, filtering methods based on Partial Differential Equation are applied here. First, the filtering effects of P-M Model and J.Weickert Model are studied very carefully. Then, on the basis of the previous studies, an ameliorating method is put forward to solve the problem of the fake stripes which appear in the model of J.Weickert. After comparing the results, we draw the conclusion that this improved method can get better effects on filtering the fake stripes and preserving the character of textures.