High-Precision Engineering Seismic Exploration

The engineering seismic exploration has several special characteristics, whichare different from oil seismic exploration, such as serious noise interference, complexterrain, varied underground structures, short observation, fold exile and so on. Thus,The data processing, which is built up from oil seismic exploration, will notapplicable to The engineering seismic exploration. Moreover, with the economydeveloping, The increasingly stringent requirements of structures, which under theconstruction, forces us to improve the accuracy of engineering exploration. Therefore,based on the characteristics of engineering seismic exploration, we propose ahigh-precision data processing which is suitable for engineering seismic theory. Thesemethods which include: CZT single-frequency interference suppression, F-K filter,coherent linear filter, F-X polynomial fitting filter, residual moveout correction, andthe fault detection with Coherent method.In the theoretical model data, we use ray tracing method to forward seismicsections, and then based on the characteristics of noise, our forward noise and addthem to the sections. First, we need to suppress single-frequency interferences, whichshapes like spikes on the spectrum. Previously, we used a single notch frequencyinterference suppression method, which hurt the pulse near the band, can not recover asingle frequency changes caused significant wave phase. Previously, we used a notchfilter to suppress it. But this way hurt the significant wave which is nearby frequencyinterference. At the same time, it can not restore the phase of significant wave, whichchange by single-frequency interference. Therefore, we use CZT to scan thefrequency and amplitude of the single-frequency interference, and then use themethod of least squares to inverse the single-frequency interference, In the way wecan avoid defects which is caused by notch filter. Secondly, what we need to do issuppressing the interference waves at low speed. We compare features between the F-K filter and linear coherence filter, then choose the appropriate speed filter. Finally,we want to suppress random interference. In the spectrum profile, The amplitude andfrequency of random interference are random, and it appears in a narrow band. Inorder to avoid the damage of significant wave, we will use the F-X polynomial filterto suppress random interference.After the suppression of the noise, we will study the impact of the superpositionwhich is caused by the residual moveout. And then we will to propose a new residualmoveout correction which is Suitable for engineering seismic exploration. Residualmoveouts include static residual moveout, NMO moveout, and dip NMO moveout.Based on their characteristics, We divided them to random residual moveout and bentresidual moveout. then puts forward a method for elimination by classifications andby steps, which make the Seismic events more smooth and broaden its frequency band.the fold exile makes the engineering seismic exploration have a poor statistics insuperposition. So we proposed to a way to increase the number of channels along thetendency which can strengthen the statistical effects on model tract. Finally, wewant to detect the faults and the fissures in the stacked section. We use the variancemethod and the improved coherent method to detection faults and fissures of section,and then compare their results to select more suitable method for the requirements ofhigh-precision detection.In the theoretical data, we completed the filtering, the residual moveoutcorrection with high precision, and the fracture detection with high precision or highSNR. From processing to interpretation, we achieve engineering seismic explorationwith high-precision.After the completion of the study on theoretical data, We processed the actualdata. And we verified the feasibility and practicality of the method mentioned earlier,but also put forward the corresponding recommendations and shortcomings.