An empirical Bayesian approach to gamma-ray log deconvolution

The spatial resolution of gamma ray logs is generally defined by the length ℓ of the gamma-ray detector. In certain exploration areas, geologic formations of interest are composed of beds whose thickness is less than ℓ To resolve thin beds by using gamma ray logs, it is helpful to reduce the averaging effects of the tool. In this paper we present a Bayesian method for gammaray log deconvolution that relies on the correct accounting of gamma-ray measurement statistics as well as the incorporation of empirically estimated lithological parameters in the deconvolution process. Application of the method to simulated data and field measurements shows that the Bayesian method can be effective in suppressing high-frequency noise encountered in the deconvolution of gamma ray logs.