A Probabilistic Geologic Model of the Krafla Geothermal System Based on Bayesian Inversion of Gravimetric Data

Abstract

The quantitative connections between subsurface geologic structure and measured geophysical data allow 3D geologic models to be tested against measurements and geophysical anomalies to be interpreted in terms of geologic structure. However, geologic models are built on sparse and uncertain data, and the non-uniqueness of geophysical data implies that an infinite number of realizations of subsurface structure may be consistent with geophysical measurements. Bayesian inversion of gravity data accounts for the inherent uncertainty of gravity data by describing the inferred lithologic and mass density structure in terms of posterior probability distributions. Such methods require the specification of a prior model of subsurface structure and rock properties. It is challenging to se1ect the probabilistic weights and the structure of the prior model in such a way that the inversion process retains relevant geologic insights from the prior while also exploring the full range of plausible subsurface models. In this study, we investigate how the uncertainty of the prior controls the inferred lithologic and mass density structure obtained by probabilistic inversion of gravimetric data measured at the Krafla geothermal system. We combine a reference prior geologic model with statistics for rock properties (grain density and porosity) in a Bayesian inference framework implemented in the GeoModeller software package. The uncertainty of the reference prior geologic model controls exerts a strong control on the inferred lithologic structure, mass density distribution, and uncertainty quantification metrics. By assuming a moderate uncertainty of the prior, the posterior lithologic structure deviates from the reference prior model in areas where it may be most likely to be inconsistent with the observed gravity data. In the case of Krafla, we observe this deviation in the vicinity of Leirbotnar sub-field, and investigate whether the discrepancy may be resolved by including a low density intrusion to the reference geologic model or by accounting for the effects of alteration in the assumed prior properties of hyaloclastite. This study underscores the importance of reliable prior constraints on lithologic structure and rock bulk density during Bayesian inversion of gravity data.