The mantle plume beneath Iceland and its interaction with the North-Atlantic Ridge: A seismological investigation

Richard M. Allen

Ph.D. Thesis, Princeton University, Princeton, NJ, USA.

Abstract This thesis presents a range of studies designed to reveal the detailed structure of the plume beneath Iceland as it rises through the upper mantle and interacts with the lithosphere and the Mid-Atlantic Ridge. These studies are part of the HOTSPOT project, which was a collaborative effort to collect data from a network of thirty PASSCAL seismometers deployed across Iceland from July 1996 to August 1998.

The crustal model presented is derived from local Love wave observations, Sn travel-time measurements and point observations of crustal thickness from previous studies. The low S-velocity structure reveals a vast plumbing system which feeds melt from the plume vertically through the lower crust into the upper crust where it travels laterally along the mid-ocean rift. The lowest velocity anomalies are found at ~10 km depth beneath the two most active volcanic complexes on Iceland, Hekla and Bardarbunga-Grimsfjall. The crustal thickness varies from 15 km around the edges of Iceland and beneath the Snaefellsnes rift zone to its thickest, 46 km, above the current center of the mantle plume. The thickness is a record of plume activity and indicates that since the Snaefellsnes rift zone was active the plume has been increasing in intensity to its current maximum today.

Two independent approaches are used to investigate the mantle structure. Firstly, a new technique is developed to sample the plume conduit at depth (250-400 km) in which the frequency dependent amplitude variations across Iceland are measured, and the plume geometry which best satisfies the patterns of observed amplitudes is determined through forward modeling. This technique avoids the drawbacks of ray theory which reduce the amplitude of the imaged anomaly through the process of wavefront healing. In the best fit model the plume conduit has a peak S-velocity anomaly of -12% and is 200 km in diameter. Secondly, the largest teleseismic body wave travel-time dataset compiled to date for Iceland is inverted to find lateral velocity variations to a depth of 400 km, the maximum depth of resolution for such data sets. Resolution is achieved in the uppermost mantle (above 100 km) for the first time by removing crustal travel-time anomalies using the 3-D crustal S-velocity model and Moho map developed here. The mantle velocity structure recovered reveals a 200 km diameter plume conduit from 400 to ~250 km depth. The size of the velocity anomalies are reduced as expected (due to the application of ray theory) from -12% to -3.8% for S-velocity, and the maximum P-velocity anomaly is -2.1%. The addition of teleseismic surface wave phase-velocity measurements provides absolute S-velocity information in the uppermost mantle. Both the body wave and surface wave data suggest the existence of a horizontal low velocity anomaly from the Moho to ~250 km depth which is interpreted as plume material flowing away from the hotspot, i.e. a plume head. In addition, both datasets reveal the presence of a high velocity anomaly above the core of the mantle plume beneath central Iceland.