Magma Intrusions and Surface Deformation; A Perspective from Discrete Element Models

As geodetic instrumentation has improved, our ability to accurately observe the surface deformation caused by shallow volcanic activity has also advanced. Methods to ensure that these precise observations are utilized in our interpretations of the driving process have also moved to keep pace with advances in the instrumentation. For analytic models it is now considered standard practice to include first order topographic corrections, with higher order corrections becoming the norm, layered structures are favored where justified, and analytic solutions beyond spheres and rectangular dikes are commonly used (e.g. pipes, ellipsoids, irregular dikes). Finite element models have gone further employing lateral variations in material properties and including visco-elastic materials. A fundamental assumption in all of these methods is that the geometry of the model forcing body represents a real region and structure beneath the volcano. For some cases this may well be the case, e.g., dikes determined by geodesy have been independently corroborated by seismicity patterns. For many cases confirmation is difficult or not possible and the assumptions for chamber like sources (e.g., spheres, ellipsoids) are often thought to be incorrect (intrusions do not form spheres). I use 3D discrete element models to investigate intrusion geometries and the subsequent effects on surface deformation. Discrete element models evolve through local contact laws, allowing particles to respond to the local stress field and material properties which results in simulated intrusions that reflect realistic geometries. Examination of the simulated surface deformation patterns and the relationship to the simulated intrusion volumes and shapes can reveal biases in current analysis techniques used in volcano geodesy. Understanding the biases in our models will improve the interpretations that we make from those models, and ultimately will drive improvements in modeling techniques.