Multimodel Adaptive Subtraction and Its Application to Multimeasurement Data Acquired in Shallow-water Surveys

Abstract

Many existing adaptive subtraction methods can be formulated as a parameter estimation problem and all of them, although fundamentally different, have a common restriction that the dimension of the unknown parameters must be determined in advance. We propose an adaptive subtraction framework, called multimodel adaptive subtraction (MMAS), that aims to relax this restriction as well as regularize the estimation of the parameters through a generalized information criterion (GIC). We show that MMAS can be applied to the popular least-squares adaptive subtraction (LSAS) method and call the resulting algorithm the multimodel least-squares adaptive subtraction (MMAS-LS). We further extend MMAS-LS to 3D and applied it to the multiple subtraction of a 3D data set from a multimeasurement shallow-water survey. We compare our proposed 3D MMAS-LS method with conventional 3D LSAS and observe that our proposed method is able to preserve the primary events better and achieve the same level of multiple attenuation compared to 3D LSAS, while using smaller or simpler filters.