Summary: | Electronic Thesis or Dissertation The Askja volcanic complex and rift segment of Iceland's Northern Volcanic Zone has been continuously subsiding at an usually high rate for more than two decades. InSAR data compiled over the last decade reveal two patterns of deformation: (1) a radially symmetric pattern of subsidence local to Askja's caldera and (2) an elongated pattern of subsidence tracking the rift segment. Microgravity data suggest a mass loss from a shallow reservoir and seismicity data reveal a relatively shallow brittle-ductile transition. A simple model combining two vertically-aligned and contracting Mogi sources, one shallow (~3 km) and one deep (~20 km), in an elastic half space generally predicts the observed InSAR deformation. Subsidence along the Askja fissure swarm has also been attributed to effects of plate spreading across rheologically weak fissure swarms. The shallow contracting Mogi source and microgravity data are consistent with magma migration out of the shallow reservoir. Interpretations of the deep contracting source are more uncertain. We present an alternative model that combines magma extraction from a shallow, fluid-filled cavity with a plate spreading model having rheologic partitioning expected for the rift segment. This 3D finite element model (FEM) simulates an elastic upper crust and viscoelastic lower crust. Inspired by a model configuration presented by Pedersen et al. (2009), the simulated brittle-ductile transition shallows beneath the rift, in accord with seismicity data. The FEM is driven by plate spreading at a constant rate and specified mass flux from the shallow cavity. The magnitude of flux is a calibration parameter estimated from InSAR data via inverse methods. Preliminary results suggest this alternative model generally predicts the both deformation patterns. However, the simulated shallow brittle ductile transition, combined with kinematic loading of plate spreading, accounts for much of the regional deformation originally attributed to magma migration out of ...
|