Multiple Elimination Research Based On Variable-depth Streamer Data

Multiple exists marine seismic data, especially some surface related multiples. the elimination is one of the most important processing task. Better eliminating for multiple, can laid a good foundation for other process and interpretation work. Surface Related Multiple Elimination method, through a prediction using raw data, then doing adaptive subtraction. After those steps, finally to achieve de-multiple. However, in shallow water,due to the absence of near offset data, the effective reflection is absent from the recording data, so, SRME method can not be used to process this data. Shallow Water De-multiple method can compensate the insufficient which exists in SRME’s processing step. The method building a model of shallow water environment, recovering and reconstructing a effective reflection. Through prediction for multiple, to achieving the purpose which is that eliminating multiple. variable-depth streamer is a broadband marine solution, compared with conventional streamer data, variable-depth streamer data can provide both high and low frequencies data. Due to the characteristic of variable-depth streamer, conventional processing method is non-compliant, taking into account the characteristics of variable-depth streamer, this paper conducted the following research:(1)Shallow Water De-multiple: research the characteristic of multiples under the shallow water environment. Using the wave field extrapolation method to predict the water layer multiple. In shallow water environment,with the offset increased, the reflected wave field reach a critical angle quickly, and then refracted. Marine seismic data, due to the lack of near-offset data, SRME method can not been applied in this moment. according to this reason, SWD can be used to process the marine data. Using recording data to reconstruct the missing water-bottom primary reflection, then pick the traveltime of the seafloor reflections. Utilize the traveltime information to model the Green’s function of the seafloor reflections. Convolve the Green’s function with the data to obtain the final multiple model.(2)Surface Related Multiple Elimination: starting from the feedback model, and deriving the 1D and 2D SRME method. using Berkhout and Verschuur ?s wave field expression, both modeling multiple and predicting multiple can achieve a good correspondence matching. both time domain and frequency domain doing the prediction, and compared calculating time. The results demonstrated that, two removal result are equal, but frequency domain’s calculating time is short.(3)Theoretical model verify the method: the model which is mainly been used, is Pluto model. Doing de-multiple processing to seismic data. And then get a stack section. Contrasting the event both in before and after processing section. The results prove that, SRME method can be well applied to the theoretical model. meanwhile, using the wave equation method doing forward modeling. Both SRME and SWD method been used to process the modeling data. Following this, the conclusion is that, both SRME and SWD method can eliminate multiple well.(4)Actual data test: using SRME and SWD method, to process variable-depth data and shallow water data. compared their spectral figure, auto-correlation figure. Coming into a conclusion is that, both SRME and SWD method can eliminate the multiples.The test on the model and actual data proves that the SRME and SWD method can be successfully used to remove the multiple. After de-multiple by these method, Besides, the spectra’s bandwidth is broad. auto-correlation is better.