Fast thin‐layer inversion (numerical and laboratory model examples)

Abstract

Summary Establishing layer thickness and layer properties from seismic data when the layer is smaller than an acoustic wavelength is a common problem in exploration geophysics. Many hydrocarbon reservoirs fall into this category, while estimates of their volume rely directly on our ability to resolve the thickness. Estimates of the world’s supply of methane hydrate are also based on our ability to measure the thickness of hydrate cemented sediments often in the 1-10 m range. In this abstract we demonstrate a very fast inversion method that solves for the layer thickness and layer properties of a thin layer from the seismic data. The signal from the thin layer is isolated (windowed) from the seismogram and converted to the reflection coefficient in the ω-p domain. The reflectivity based forward model calculates the reflection coefficient for a given layered earth model. The very fast simulated annealing inversion routine finds the layer properties by performing a global minimization of the L norm between the measured and modeled reflection coefficients. We present two demonstrations, a numerical example using synthesized data consistent with a DTAGS shot over a hydrate bearing sediment, and a laboratory physical modeling example where the model is a thin aluminum layer embedded in a Lucite background.