A comparison of porosity estimates obtained using post‐, partial‐, and prestack seismic inversion methods: Marco Polo Field, Gulf of Mexico

Abstract

We compare porosity estimates obtained using post‐, partial‐, and prestack (i.e., full waveform) seismic inversion methods in unconsolidated Miocene sandstones. We invert a single 3D seismic data set using commercially available poststack and partialstack inversion algorithms and a prestack inversion algorithm that optimizes a quasi‐linear objective function via local optimization methods. We compare the inversion results with a priori structural and stratigraphic information and draw the following inferences: • Density is the most valuable elastic parameter for estimating porosity in the unconsolidated sandstones. P‐impedance‐to‐porosity transformations are unstable and magnify errors in the estimated elastic parameters. • Partialstack seismic inversion only resolves P‐impedance and S‐impedance within the study area. It cannot resolve density because the linearized forward model is insensitive to density over the seismic aperture and the NMO correction removes the low frequency velocity information that is necessary to decouple density and velocity within the P‐impedance. • Poststack and partialstack seismic inversion methods rely upon P‐impedance to estimate porosity and produce nonphysical porosity estimates. • Prestack seismic inversion resolves P‐velocity, S‐velocity, and density within the study area. • Only the prestack seismic inversion yields accurate porosity estimates within the study area.