Study on internal multiple elimination method on Land Seismic Data

Abstract

Multiple is a tough issue in recent years, especially the internal multiples in deep earth. In this paper we proposed to construct virtual events to predict internal multiples. Then adopt the mean value multi-channel adaptive subtraction method for matching the multiple model and seismic data. However, the results depend on the parameters closely. We propose to apply the dynamic time wrapping method in order to accomplish the precise matching. And the regression of multiples can effectively improve the internal multiple elimination results when internal multiples are interfered with effective signal.