| Static correction estimation is an important technology in seismic data processing, especially the static correction under complex surface, which is hard and not resolved well at present. In this paper, this problem is further discussed, and some new progresses are made.After the field static correction, the residual static measurements can be estimated by the traditional residual static correction methods. When statics are larger than half of the period or the S/N ratio is too low, however, the traditional solutions are known to fail because of the cycle skips which grossly distort the apparent structure in seismic sections.The residual statics estimation is actually a nonlinear inverse problem. In this study, after analyzing the nonlinear algorithms including simulated annealing, genetic algorithm and particle swarm algorithm, we present a comprehensive nonlinear global optimization method for residual statics based on these algorithms and stack-power maximization. This comprehensive nonlinear optimization method has the advantage of the sample parameter selection and practical applicability. Results demonstrated on synthetic data show that this method is stable, its speed is acceptable and it can effectively solve the residual statics of low S/N ratio seismic data.In statics estimation using Monte Carlo technique, the result always accompanies with null space. In this study, we discuss the reason for null space occurrence and processes of removing it. Through the test, we find the self-adaptive spatial filtering method is efficient in removing null space. So we integrate the self-adaptive spatial filtering with comprehensive nonlinear optimization residual statics algorithm.We program the cross-platform code in c++ language and apply this program to filed data. The processed results of field data from Xinjiang and Shanxi show this algorithm is useful on seismic data with low S/N ration and it has been proved to be simple, economical, stable and easier to converge.
|
PDF Full Text Request