The Iceland Plume is an upwelling of anomalously hot rock in the Earth's mantle beneath Iceland whose origin probably lies at the boundary between the core and the mantle at ca. 2880 km depth. It is generally thought to be the cause of the formation of Iceland and its volcanism, which characterizes the island to the present day, according to the plume theory of W. Jason Morgan.
The plume, which bears the name of Iceland and lies roughly beneath the center of the island, is considerably older than Iceland. Volcanic rocks related to it are found to both sides of the coast of southern Greenland and had their ages determined to lie between 58 and 64 million years; this coincides with the opening of the north Atlantic in the late Paleocene and early Eocene. It is generally thought that the volcanism was caused by the flow of hot material from the plume head into regions beneath the lithosphere which had previously been thinned by rifting and produced large amounts of melt there. The exact position of the plume at that time is controversial, but was probably beneath central Greenland; it is also not entirely clear whether the plume had ascended from the deep mantle only at that time or whether it is much older and also responsible for the old volcanism in northern Greenland, the Ellesmere Island Volcanics on Ellesmere Island, and in the Arctic Ocean (Alpha Ridge). All of these volcanics are part of the High Arctic Large Igneous Province.
As the northern Atlantic opened to the east of Greenland during the Eocene, North America and Eurasia drifted apart; the Mid-Atlantic Ridge formed as an oceanic spreading center and a part of the submarine volcanic system of mid-oceanic ridges. In the course of these plate motions Greenland moved above the Iceland plume; different investigations located the plume beneath the coast of southeastern Greenland (Scoresby Sound) or slightly eastward from it 40 million years ago and related it to the North Atlantic Igneous Province. Upon further opening of the ocean and plate drift, the plume and the mid-Atlantic Ridge approached each other, and finally a part of the plume head reached the region of thinned lithosphere at the ridge, leading to increased generation of melt and crust; these processes grew stronger the more both structures converged. The Greenland-Iceland Ridge and the Faroe-Iceland Ridge, both regions of greatly thickened oceanic crust, are traces of this stage of the convergence preceding the formation of Iceland.
The oldest crust of Iceland itself is more than 20 million years old and was formed at an old, now extinct oceanic spreading center in the western fjord (Vestfirðir) region. The westward movement of the plates and the ridge above the plume and the strong thermal anomaly of the latter caused this old spreading center to starve 15 million years ago and lead to the formation of a new one in the area of today's peninsulas Skagi and Snæfellsnes; in the latter there is still some activity in the form of the Snæfellsjökull volcano. The spreading center, and hence the main activity, have shifted eastward again 7–9 million years ago, however, and formed the current volcanic zones in the southwest (WVZ; Reykjanes–Hofsjökull–Vatnajökull) and northeast (NVZ; Vatnajökull–Tjörnes). Presently, a slow decrease of the activity in the WVZ takes place, while the volcanic zone in the southeast (Katla–Vatnajökull), which was initiated 3 million years ago, develops.
In addition to the formation of Iceland the plume has also influenced the generation of crust at the adjacent segments of the mid-oceanic ridge, especially at the Reykjanes Ridge southwest of the island. In this region, a significant thickening of the crust and an anomalous uplift of the seafloor are observed, which are explained by a hot mantle current emerging from the plume and flowing along the bottom of the thin lithosphere at the ridge; the variations in crustal thickness, which form a chevron-like pattern, show that this current has not been constant over time. Among the different possibilities to explain this pattern are interactions of the shift of the spreading center with the plume head or pulses of the plume itself.
Information about the structure of Earth's deep interior can be acquired only indirectly by geophysical and geochemical methods. For the investigation of the Iceland Plume as well as of other plumes, gravimetric, geoid and in particular seismological methods along with geochemical analyses of erupted lavas have proven especially useful. Numerical models of the geodynamical processes attempt to merge these observations into a consistent general picture.
An important method for imaging large-scale structures in Earth's interior is seismic tomography, by which the area under consideration is “illuminated” from all sides with seismic waves from earthquakes from as many different directions as possible; these waves are recorded with a network of seismometers. The size of the network is crucial for the extent of the region which can be imaged reliably. For the investigation of the Iceland Plume both global and regional tomography have been used; in the former, the whole mantle is imaged at relatively low resolution using data from stations all over the world, whereas in the latter, a denser network only on Iceland images the mantle down to 400–450 km depth with higher resolution.
Regional studies from the 1990s (ICEMELT, HOTSPOT) show unambiguously that there is a roughly cylindrical structure with a radius of 100–150 km beneath Iceland down to at least 400 km depth, in which the velocities of seismic waves are reduced by up to 3% (P waves) and more than 4% (S waves), respectively, compared to the reference model; the most recent analyses point towards an even stronger reduction. If these values are converted into a temperature anomaly using rock-physical models, it is found that the mantle there is 150–250°C hotter than normal. Some uncertainty is introduced into the models by the limitations in spatial resolution of seismic tomography: it is difficult to distinguish a hot, thin plume from a less hot, broader one with this method.
Global tomography confirms that there is a strong anomaly with clearly reduced seismic velocities in the upper mantle beneath Iceland. For the lower mantle (below 660 km depth), the picture is more contradictory. In all investigations the anomaly weakens considerably there and has a more irregular shape; in some depth intervals it even seems to disappear, although those depths are not the same in the different studies. At the core-mantle boundary below Iceland, an anomalously hot region has also been found with other seismological methods , and the structure of seismic discontinuities at 410 and 660 km depth below Iceland also indicates that the temperatures are elevated there. Therefore the majority of scientists thinks that the weaker signature of the plume in the lower mantle can be explained with possible temporal variability of the plume and/or change of physical properties of the mantle with depth as well as with the limitations of the method and the available data, but that the plume does indeed reach down to the base of the mantle.
Numerous studies have addressed the geochemical signature of the lavas present on Iceland and in the north Atlantic. The resulting picture is extraordinarily complex and partly self-contradicting, but nonetheless consistent in several important respects. For instance, it is not debated that the source of the volcanism in the mantle is chemically and petrologically heterogeneous: not only the normal peridotite, but also eclogite contribute to the melts. The origin of the latter is assumed to be metamorphosed, very old oceanic crust which had sunk into the mantle several hundreds of millions of years ago during the subduction of an ocean. Moreover, isotopic ratios of noble gases yield evidence that there is also a contribution from rock from the lower mantle .
The variations in the concentrations of trace elements like helium, lead, strontium, neodymium, and others show clearly that Iceland is an anomaly also with regard to geochemistry in comparison to the rest of the north Atlantic. For instance, the ratio of He-3 and He-4 has a pronounced maximum on Iceland, which correlates well with geophysical anomalies, and the decrease of this and other geochemical signatures with increasing distance from the plume allows to estimate that the influence of the plume reaches about 1500 km along the Reykjanes Ridge and at least 300 km along the Kolbeinsey Ridge. Depending on which elements are considered and how large the area covered is, one can identify up to six different mantle sources, which, however, are not all present in any single location.
Furthermore, some studies show that the amount of water dissolved in mantle minerals is two to six times higher in the Iceland region as compared to undisturbed parts of the mid-oceanic ridges, where it is regarded to lie at about 150 ppm.
The north Atlantic is characterized by strong, large-scale anomalies of the gravity field and the geoid, with Iceland lying in their center. The geoid rises up to 70 m above the geodetic reference ellipsoid in an approximately circular area with a diameter of several hundred kilometers; this is explained by the dynamic effect of the upwelling plume which bulges up the surface of the Earth. Furthermore, the plume and the thickened crust cause a positive gravity anomaly of about 60 mGal (=0.0006 m/s²) (free-air).
Since the mid-1990s several attempts have been made to explain the observations with numerical geodynamical models of mantle convection. The purpose of these calculations was, among other things, to resolve the paradox that a broad plume with a relatively low temperature anomaly is in better agreement with observed crustal thickness, topography, and gravity, whereas a thin hot plume matches seismological and geochemical observations better. The most recent models indicate that the plume is probably 180–200°C hotter than the surrounding mantle and that its stem has a radius of ca. 100 km, i.e. the seismological findings are confirmed; crustal thickness, topography, and gravity can be explained with such a model if one takes into account that the loss upon melting of water which was dissolved in mantle rock massively alters the fluid dynamical behaviour of the plume, so that the corresponding anomalies become broader and melt production decreases. Thus far, the models do not or not fully take into account petrological heterogeneity, however.
The chevron structures of the Reykjanes Ridge mentioned in the section on the geological history are explained by geodynamical models as pulsations of the plume, i.e. as variations in the mass flux through the plume stem.
As mentioned in the beginning, the plume model is the commonly accepted concept for explaining the formation of Iceland and its volcanism. However, in particular the weak visibility of the plume in tomographic images of the lower mantle and the geochemical hints on eclogite in the mantle source have raised doubts among some scientists such as Don L. Anderson and Gillian Foulger that the plume model is really valid. As an alternative, they propose processes which are restricted to the upper mantle.
According to one of those models, a large chunk of the subducted plate of a former ocean has survived in the uppermost mantle for several hundred million years, and its oceanic crust now causes excessive melt generation and the observed volcanism. This model, however, is not backed by dynamical calculations, nor is it exclusively required by the data, and it also leaves unanswered questions concerning the dynamical and chemical stability of such a body over that long period or the thermal effect of such massive melting.
Another model proposes that the upwelling in the Iceland region is driven by lateral temperature gradients between the suboceanic mantle and the neighbouring Greenland craton and therefore also restricted to the upper 200–300 km of the mantle. However, this convection mechanism is probably not strong enough under the conditions prevailing in the north Atlantic, e.g. with respect to the spreading rate, and it does not offer a simple explanation for the observed geoid anomaly.