Mechanical layering and layer inclination controls on magma chamber induced stress

Abstract

The distribution of stress surrounding a shallow magma chamber presents a first order control on both the position of magma chamber rupture and the resulting direction of any dyke, sill or inclined sheet that is formed. As such, knowledge relating to the stress field surrounding magma chambers and at the near surface are vital for correctly assessing the likely location of magma injection and eventual eruption position. The vast majority of models used to estimate the crustal stress field during periods of magma inflation either assume that the crust is homogeneous, an elastic half-space, or heterogeneous but simplified to a series of horizontal layers with often contrasting mechanical properties. Using finite element method (FEM) analysis we expand on those assumptions so as to consider a magma chamber hosting crustal segment composed of heterogeneous layers of contrasting mechanical properties and with variable dip angles. Two magma chamber geometries are considered, the first an idealised circle (sphere in three dimensions) and the second, a sill like ellipse (ellipsoid in three dimensions). The main boundary condition considered is an overpressure of 10 MPa in the chambers which is used to generate the resulting stress field. Layer inclinations (dip) alter between horizontal to vertical with model simulations performed at 20-degree inclination variations. The layer heterogeneity is governed by differences (contrasts) in each layer’s elastic moduli (stiffness). We tested stiffness ratios of 100:1, 10:1, 1:1 (homogeneous case), 1:10 and 1:100. Results confirm previous insight from heterogeneous crustal modelling indicating high stiffness ratios generate significantly perturbed stresses within different layers and can lead to local stress rotations at the contact between two mismatched or contrasting layers. The effect of the layer inclination further perturbs the stress field and shift the location of maximum stress concentration by several hundreds of metres to kilometres preferentially in the direction of the layer dip. Our results confirm the need to accurately characterize both subsurface magma chamber geometries and the mechanical properties and attitude of the layers that comprise a volcano. This is especially important in the Andes where volcanoes are often built atop heterogeneous crust that has been deformed and folded, generating highly non-horizontal stratification.