Antarctica 3D Crustal Structure Investigation by Means of Gravity Inversion Algorithm: the Wilkes Land Case Study

Abstract:

Gravitational field observations can be inverted to estimate the density distribution within a given region with the aim to improve the geological knowledge. In recent years, geophysical methods based on gravity observations (as well as on others not-invasive techniques) have been widely used to investigate the lithosphere structure of the Antarctica continent which is still not completely unveiled. The extreme and complex conditions prevent direct in situ measurements thus preferring indirect investigation methods such as the gravity inversion and data interpretation.

In this study, we present preliminary results of a Bayesian inversion algorithm applied to the Antarctica continent, with focus on the Wilkes Land region. Here, we infer a model of the crustal structure and of the mass density distribution. Firstly, an initial 3D geological model (prior model), in terms of geological horizons (e.g. bedrock, sediments, crystalline crust, etc.) and densities, has been created based on available data from literature. In addition, the accuracy of each horizon and the variability of the density inside each layer are also compiled and used to formalize the accuracy of the prior model. The 3D model is discretized in a set of volumetric elements, each characterized by two variables: namely the mass density ρ_i and the “label” L_i (i.e. a discrete variable defining the geological material of the volumetric element itself). These two variables are considered as random variables and, within an iterative procedure based on Markov Chain Monte Carlo method, are adjusted to fit the observed gravity field. The tests performed show that the method is capable of retrieving a final 3D model consistent with the available prior information (model and accuracies) and fits the gravity observations (namely gravity disturbances and second radial derivatives), according to their accuracy. The results provide some insights into some of the major open questions about the Wilkes Subglacial Basin such as the thickening/thinning of the crust beneath the basin or the amount of sediment deposits in the area. A sensitivity analysis is also provided to determine how the final model is affected by changes in the prior information.