Tomographic Evidence for a Whole Mantle Plume below Iceland

Harmen Bijwaard and Wim Spakman

Vening Meinesz School of Geodynamics, Utrecht University, The Netherlands
E-mail: bijwaard@geo.uu.nl

Iceland has long been known to be a center of both hotspot and Mid-Atlantic Ridge activity. Hotspots are generally thought to be the surface expression of mantle plumes: relatively narrow channels of hot mantle upwellings. Numerical modeling and laboratory experiments have helped to gain insight into the dynamics of plumes, but on several aspects no complete consensus exists. Especially the origin of plumes (the upper-to-lower mantle discontinuity or the D" layer), the apparent stationarity of hotspots, and the size and excess temperature of plumes are subject of ongoing debate. Further constraints on these issues may come from seismological investigations. Both regional and global seismological studies have recently provided evidence for the low seismic wavespeeds and deflected mantle discontinuities below Iceland, which are expected for a deep rooted mantle plume. From this evidence estimates of size and excess temperature have been derived that agree reasonably well with those from numerical modeling. Furthermore, the observed deflections point toward a lower mantle (and hence D") plume origin.
We will show results for the entire mantle below Iceland from recent global tomography research (Bijwaard et al., JGR, 1998). The main advantages of this global study over previous work are the exploitation of a reprocessed global seismic data set and the implementation of an irregular cell parameterization. This has allowed us to construct a detailed solution in well-sampled mantle volumes. Although many of the well-known hotspots are located above mantle volumes that are poorly sampled by seismic rays and hence are not very well resolved, one of the few hotspots below which we have reasonably good resolution for the entire mantle volume seems to be Iceland.
In the mantle below Iceland we observe a wavy, plume-shaped low wavespeed anomaly with a broad (1000 km wide) root zone, an approximately 500 km wide central part and a large (1200 km) 'plume head'. These measures may be somewhat overestimated since the horizontal resolution is of the order of 300/500 km in the upper/lower mantle. At several depths the plume-shaped anomaly seems to be connected to other low velocity features. For example, at approximately 1300 km depth it is part of a semi-circular ring of low velocities below Europe and northern Africa and seems to connect several hotspot regions (Eifel, Massif Central, Hoggar, Canary Islands). Apart from that, the plume-shaped anomaly appears to be deflected with a root zone stretching west toward Greenland and a central part east from Iceland, approximately 1000 km below the Faeroe Islands. This wavy shape may contradict the concept of stationary hotspots. The vertical continuity of the (deflected) structure is well resolved (on the order of 120-200 km) which suggests a D" origin for the Iceland hotspot. The amplitude of the inferred seismic anomaly can be translated into an excess temperature of 200-300 K for the entire mantle which agrees well with previous and independent estimates.