A Study Of Quasi-Born Approximation On Migration In F-K Domain

When migrating, if we want to keep the amplitude and phase information included in the input, we can't use Ray-tracing-based migration method. The seismic-wave dynamics attribute can reflect abundant information of the elasticity in the underground structures, so the wave-equation migration method develops quickly. Prestack depth migration imaging in complex structures is a focus in geophysics field. Migration method based on wave-equation becomes mature incessantly and mainstream gradually. But all migration methods in the time field suppose that velocity is constant or changes a little horizontally or that structure doesn't complex. If this doesn't right, they can't make the location of target right. Prestack depth migration without these shortcomings becomes a focus in geophysics field with the powerful ability of computer. Prestack depth migration in frequency-domain improves efficiency by FFT (Fast Fourier Transform). But the common prestack depth migration in frequency-domain can't be applied when the velocity of medium isn't constant horizontally. In order to make prestack depth migration method be applied to the medium with horizontal velocity variety, many scholars improved on it these years, but all these reduce the efficiency of the calculation badly. In order to make prestack depth migration both image accurately and improve efficiency in complex structure, the thesis analyze some prestack depth migration methods in frequency-domain synthetically and improve the Quasi-Born method. The method improved includes high-level disturbance, can be not only applied in medium with various velocity horizontally, but also adjust the efficiency by the degree of the horizontal velocity variety, which has a better joint between efficiency and exactitude. The thesis proves that the method and theory is effective and feasible by forward modeling and migration of a series of models. The main idea of the thesis is making use of quasi-born approximation which decomposes the pressure wavefield into two parts, namely, the incident wavefield and scatter wavefield. While in the integrator, the scatter wavefield can be decomposed two terms, namely, the incident wavefield and other summation all terms, which is similarly to linear relation. This theory is called quasi-liner Born approximation in the theory of electromagnetic wavefield. The thesis extracts the idea of high-level approximation, and improves formula about extrapolate of wavefield based on quasi-born approximation following forerunners. We take into account of high-level disturbance, which accord with the fact of seismic wave's transmitting in the underground medium. And this also accord with the reality when extrapolating. Comparing to Born approximation, quasi-Born approximation algorithm need much more time in the same medium. So, the efficiency is very low making use of quasi-Born approximation algorithm than the other, especially when the medium isn't very complex. Born approximation can also get satisfying result. The thesis combines the Born approximation and the quasi-Born approximation algorithm, whose essential is that Born approximation is used when the medium isn't much complex and quasi-Born approximation is used when the medium is very complex in course of extrapolating. Thus the method combined has a satisfying joint between accuracy and efficiency. The thesis derivate the theory formula of extrapolate and forward algorithm, and applies these extrapolation formulas in forward. And applies formulas improved in wave extrapolation of the prestack depth migration imaging by correlative theory of prestack depth migration. Then, it tests the forward formulas and migration formulas using the numerical technical, and gains the corresponding synthetic seismic records of models. The synthetic records are migrated, and we gain the result of migration of corresponding models. From the maps gained, we see clearly that the maps gained are similar with the corresponding models, which proves that the migration improved can be applied both in poststack depth migration and in prestack depth migration and that the outcome of migration is satisfying. Finally, the migration results of the method before and after improved are written. From them, we can see easily that migration result by migration method improved is more accurate than method before improved, which prove that the migration method are successful. The result is satisfying, but the input data is man-made whose Signal-to-Noise is high. The impact need study more when the...