Study On Data Processing And De-noising For Seismic Reflecting Wave Imaging Profiling In Water Domain

Seismic imaging profiling has proved to be one of an effective technique for geophysical investigation. This paper develops the seismic imaging profiling technique to be utilized in data processing and de-noising in water domain. Firstly, the theory of reflection waves imaging is analyzed and discussed thoroughly. By taking two methods to resolute the wave equations and the simplified Zoeppritz equation from the aspects that deduces and simulates the reflection characteristics of the reflection waves with varieties of the incidence angle in dissimilar geotechnic mediums, the optimal offset of observation system is obtained. Then, a reasonable system that collects and gauges water domain datum is proposed. Furthermore, a set of complete data-procession flows of seismic imaging profiling that differs from traditional common centre traces with utilizing common offset traces to analyze and process exactly the collected datum are designed. However, the signal-to-noise ration of the collected datum is lower because of the investigation methods are influenced by some factors such as exploding, receiving and instruments of prospecting etc. So, in order to improve the precision of data interpretation, some suitable methods and modules should be found to achieve the best signal-to-noise ration and data resolution in the course of processing data A large number of practical examples indicate that the signal-to-noise ration and data resolution will be improved to some extent by choosing some special modules of filtering and noise elimination to process seismic datum before and after taking the routine methods. According to the feature on water domain, this paper works out emphatically and develops five modules such as time-variant filtering, predictive filtering in f-x domain, multi-trace median filtering method, sharp impulse deconvolution and predictive deconvolution. By utilizing the wavelet of Richer to make the seismic model of the theory and calculating the seismic model of the theory and practical seismic datum repeatedly the optimal treatment parameter of each module is confirmed and by the principle and characteristic of each method is detailed analyzed. Influences that caused by some parameters recorded by seismic such as the signal-to-noise ration, frequency spectrum, amplitude of main purpose layers, and confidence that before and after dealing are discussed by analyzing fidelity of the result after processing the seismic model and actual seismic datum. Finally, some techniques and methods about predictive filtering in f-x domain and predictive deconvolution with strong capacity of eliminating noise and relative fidelity are worked out, and offered with optimal processing procedure to greatly improve the signal-to-noise ration and high resolution for the seismic datum. The main initiative of this paper is that a set of complete data collecting,treatment and interpretation method with high fidelity and high resolution is discussed and succeeded and the hot and hard problem that how to obtain valuable faint signals from a drastic noise background during the course of processing seismic data is also solved. Solving the problem will have a huge contribution and make huge economy benefit to the society.