Exploring the impact of thermally-controlled crustal viscosity on volcanic ground deformation

This is the final version. Available on open access from the American Geophysical Union via the DOI in this recordNote - there is another record for this publication in ORE: http://hdl.handle.net/10871/126736Volcanoes undergoing unrest often produce displacements at the ground surface, providing an important window to interpret the dynamics of the underlying magmatic system. The thermomechanical properties of the surrounding host rock are expected to be highly heterogeneous, with key physical parameters having strong temperature dependence. Deformation models that incorporate non-elastic rheological behaviours are therefore heavily reliant on the assumed thermal conditions, and so it is critical to understand how the thermomechanical crustal structure affects the observed deformation field. Here, we use a series of thermo-viscoelastic Finite Element models to explore how variations in thermal constraints (i.e., reservoir temperature and background geothermal gradient) affect surface displacement patterns when using the Maxwell and Standard Linear Solid viscoelastic configurations. Our results demonstrate a strong variability in the viscoelastic deformation response when changing the imposed thermal constraints, caused by the partitioning of deformation and the dissipation of induced stresses. When using the Standard Linear Solid rheology, we identify that cumulative long-term displacements can vary by over 20%, relative to a reference model with reservoir temperature of 900 °C and background geothermal gradient of 30 K km-1. The relative change increases to a maximum of 35% when thermal weakening of the Young’s modulus is also considered. Contrastingly, the deformation patterns of the Maxwell rheology are governed by unbounded displacements and complete stress relaxation. Ultimately, we outline that uncertainties in the thermal constraints can have a significant impact on best-fit source parameters (e.g., size and depth) and overpressure/volume-change loading histories inferred from thermo-viscoelastic models.GNS Scienc