Viscoelasticity and Magma Reservoir Failure: Comparing Common Approaches to Ground Deformation Modelling and their Impact on Failure Timescales

Abstract:

As volcanic systems undergo unrest, understanding the conditions required for reservoir failure, and the associated timescales, are critical when evaluating the potential for eruption. The characteristics and dynamics of a deforming or pressurised magmatic system can be inferred from episodes of surface deformation, but this process is heavily reliant on the modelled crustal rheology. Viscoelastic rheologies can account for the subsurface thermomechanical heterogeneity induced by shallow or long-lived magmatic systems, a limitation of elastic models. Here we consider two principal methods for employing viscoelasticity: i) a Maxwell viscoelastic shell surrounding the magmatic source and embedded within an elastic medium (“MaxSh”); and ii) a temperature-dependent viscosity structure together with a Standard Linear Solid viscoelastic configuration (“TDSLS”). Given the differences between these two modelling approaches, it is important to understand how the chosen methodology affects the model outputs; such as the predicted ground displacement patterns, inferences of the underlying source processes (i.e. time-varying overpressure), and the projected timescales of reservoir failure. Recognising these impacts are paramount when assessing the present-state of a volcanic system and the likelihood of eruption. In this study we present a multi-parametric assessment of the two modelling methods, exploring the influence of key model characteristics (e.g. shell thickness and viscosity for MaxSh, geothermal gradient and magmatic temperature for TDSLS), coupled with a linearly increasing overpressure load, on surface deformation patterns and timescales of failure. We determine the mechanical stability of the magmatic system by evaluating the occurrence of tensile failure and Mohr-Coulomb failure in the surrounding crustal rocks. Preliminary results highlight significant differences between the MaxSh and TDSLS configurations, both in terms of the predicted ground displacements and the timescales for the apparent eruptibility and subsequent through-going failure of the magmatic system. These results suggest a major disparity in model “equivalence” when reproducing observed deformation patterns, reinforcing the necessity to consider different modelling viewpoints during analysis.