In an earlier paper (Nichols, 1989) I proposed a method for separating the different wavetypes in a nine-component seismic dataset. My method is an approximate method that does not rely on knowledge of the elastic properties of the near surface. This contrasts with exact separation methods, such as that proposed by Herrmann et al. (1989), which require knowledge of the near surface properties.
In this paper I review the method that I presented in my previous paper. I use a first order approximation of the exact slowness dependent separation operators. I show that this approximation is valid for the range of slownesses acquired in a typical surface seismic experiment. The separation algorithm involves the estimation of several separation parameters. I discuss several methods of estimating these parameters and apply the separation scheme to a very simple synthetic seismic dataset.