Research Of Wavelet Denoising On Seismic Data Interpretation

In today’s large-scale industrialization, social demand for oil in anunprecedented intense, exploration and exploitation of oil resources， has alreadybeen mentioned an unprecedented importance. With the deepening of the explorationwork, all kinds of complicated geological conditions can met, it puts forward higherrequirements on exploration technology. A comparison to save exploration method isto adopt the way of artificial earthquake formation of reflection seismic data,through the interpretation of seismic data to the exploration of undergroundgeological structure. Fault structure for petroleum exploration in the stratum isof great significance, through the analysis of the fault, geological explorationpersonnel can be concluded that the underground reservoir.Coherence technique is the effective method of seismic data of faultinterpretation. It calculates the similarities of the adjacent several seismictraces, to detect the discontinuity of seismic wave in phase axis. The similaritycalculated values, according to their space, time arrangement, combination becomescoherence data. Coherence eliminates the amplitude fluctuation for faultidentification inconvenience highlights the discontinuous characteristics of faultbrings to the adjacent channel data. Using it to identify the fault is of great help.According to the characteristics of the seismic data, and the advantages of thecoherence technology, this article obtains from the point of the source data toimprove coherence, has carried on the related research work.Study the characteristics of seismic data, seismic data is the data which containpositive and negative value, and the mean amplitude wave of zero time. The seismicwave data in adjacent trace of similarity, decided to the size of the values ofcoherence. The difference values of adjacent points in one seismic trace, make upa new seismic trace which the phase of a wave with the original data of new datawave approximate difference of PI/2. By the data of wave data volume calculationof coherence can be similar to the original seismic data coherence, but can showmore details geological structure of coherence.Using a one-dimensional wavelet to wavelet decomposition of seismic data, toeliminate the high frequency part, get more smooth seismic trace data waveform. Theseismic data in the adjacent channel on the similarity of improved, so as to improvethe quality of the coherence. Have of decomposition of seismic data to eliminatethe high frequency part directly, rather than like using wavelet processing imagethreshold filtering, is considering the data, seismic data is volatility in the nearthe zero point fluctuation change fast, if the use effect of the threshold valuefiltering can produce a ladder. There is direct to eliminate the high frequency part,higher computing speed. To3d seismic data, wavelet filtering enhancement transverse similarity betweendata trace, eliminate the noise interference between the seismic trace. Researchof3d seismic data, wavelet filtering used two different filtering methods, foundthat each of the two methods has its advantages and disadvantages. One is, one inthree directions, respectively, using the wavelet filter, and reconstruction. Threedirections of filtering results are obtained, and then its average as the finalfilter results. Another method is that the traditional three dimensional waveletfiltering of separable variables. After stripping out three directions after gettingthe high frequency part of wavelet transform, then reconstruct, get the data afterfiltering. Both filter method for calculation on the basis of coherence andstrengthen again. The first method of seismic data with the original seismic datahaven’t changed much, so the smaller structure of geological reserves is better,but also keep the noise of the corresponding number. The second method improves thesimilarity of adjacent channel well, reserved for larger fault structure is alsovery good, very suitable for used to identify large fault.The application of different methods used in the different seismic data in thispaper. Seismic data A, suit to use one-dimensional wavelet filter method, and issuitable for use of3d seismic data volume B wavelet filter method. This articledoes not give a universal method, just gives some methods to be selectively.