| High resolution single channel seismic exploration is one of the essential methods to detect shallow stratum structures, active faults and also the potential geohazards under seabed. It is widely used in marine geological surveys and offshore site investigations. During the data acquisition stage of marine single channel seismic exploration, the seismic source and hydrophones are placed underwater and towed by the research vessel, seismic waves are produced by the source at fixed time interval and each seismic trace is recorded while the vessel navigating with a constant speed. Many kinds of interference may exist in single channel seismic record, such as multiple, ghost reflection, ambient noise, etc., and these interference will influence seismic data interpretation. Hence, understanding the causes and propagation rules of the interference will be helpful in improving seismic data processing.This thesis first introduces several methods of marine seismic exploration, sparker sources and hydrophone array usually used in marine single channel seismic exploration and also the familiar interference in single channel seismic record. And then based on the survey data of Ping-tan sea, spectral filtering according to the analysis result of background noise and extracting seismic wavelet from direct wave and correlation filtering utilized the seismic wavelet. Last but not least, partition seismic sequence according to the analysis result of nearby core sediment. Following are three main cognitions from this thesis:First of all, marine seismic exploration is based on research vessel, noise produced by vessel machineries(include propulsion motor, propeller, electric generator, etc.) can throughout the entire process of data acquisition. According to the analysis result of noise record getting without seismic source, the frequency of the noise produced by vessel machineries and also the environment is about 35-140Hz (the exact value is related to the research vessel and work area) while the basic frequency of sparker source is about 180-2000Hz, they can be separate in frequency domain. So we could choose the filtering parameters according to background noise record in order to retain the desired signal mostly.Secondly, the most fundamental requirement for seismic source is that the repeatability of the seismic wave should be remarkable. So we can consider that the waveform, the energy, and also the wavelength of the seismic waves produced by the sparker source each time is identical. Direct wave is the seismic wave spread through seawater from source to hydrophone. Since the absorption of seawater is weaken than sediments and rocks, the direct wave recorded in the raw seismic data is always containing the stronger energy while the less distortion. Therefore we can extract seismic wavelet from the direct wave. However, the influence of environmental factor must be eliminated while extracting seismic wavelet from direct wave. In practice, we can pick up the direct wave in each trace utilize correlation method and then average the direct waves and take the average result as seismic wavelet.Thirdly, the seismic wavelet extracting from direct wave should have a strong correlation with reflected wave while have a weak correlation with other interference. So it could be applied to correlation filtering. The processing result shows that correlation filtering utilize the seismic wavelet extracting from direct wave can improve the signal-to-noise ratio of the seismic record and the length of correlation wavelet must be shorter than seismic wavelet. In addition, band-pass filtering before correlation processing can eliminate the influence of water wave.Generally, seismic data interpretation must go ahead with the aid of geological drilling data or geophysical logging data. According to the analysis result of the core sediments near the survey line and set the average acoustic velocity as 1600m/s, transform the length to travel-time and then compare with the time profile and finally recognize four seismic interface. |
PDF Full Text Request