Science 2009 (The Bent Hawaiian-Emperor Hotspot Track: Inheriting the Mantle Wind, by John Tarduno, Hans-Peter Bunge, Norman Sleep and Ulrich Hansen)
A number of volcanically active areas, such as Hawaii or Iceland, are thought by geoscientists to originate from localized regions of high temperature within the Earth's mantle. The term Hotspot is now commonly employed to describe such regions. One key element of Hotspots is that they are assumed to be stationary. In other words, they move little over geologic time, especially when compared to the Earth's tectonic plates. When a plate moves over a Hotspot, a chain of volcanic islands is left behind. This makes Hotspot chains a convenient marker to track the motion of plates back in time. They form a so-called Reference System for plate motion. The figure below shows the track the Hawaiian Hotspot left on the Pacific plate. The distinct directional change of the island chain is commonly thought to correspond to a change of motion of the Pacific plate. The plate, presumably, went first northward. Then it changed direction rapidly to an eastward motion some 40 million years ago, thus causing the bent in the chain.
1: Present-day track of the Hawaiian-Emperor Seamount Chain, with ages highlighted.
Paleomagnetic studies now show that the Hawaiian Hotspot was probably not stationary. Rather it drifted rapidly to the South some 80 to 40 million years ago. The Hotspot then slowed down, thus causing the distinct bent. Consequently, it seems likely that the bent does not reflect a change in plate motion, and that the Pacific plate moved eastward for most of the past 80 million years.
Wiithout the inherent Hotspot motion a rather different track would have resulted (see Figure 2 below)
2: Northern Pacific Ocean and the Hawaiian-Emperor chain (top) and a view with the volcanic track that would have been produced had the Hawaiian hotspot been fixed in the deep mantle (bottom). The difference between the predicted track and the actual track is a record of hotspot motion which can be explained by mantle plume tilt beginning a mid-mantle levels toward a ridge (see Figure 3 below).
The new result fits in neatly with what one would expect theoretically, for a simple reason: the Earth's mantle - though made up of solid rock - behaves like a fluid over geologic time. In other words, it can move and flow. One should, therefore, be prepared to find evidence for rapid Hotspot motion at some times, even when they are near stationary at other times. Computer models can now be used to simulate the flow of the mantle in some detail. These models confirm that Hotspots may get tilted and moved around from time to time by large scale mantle flow. Put differently, the Hotspots move in the mantle wind (see Figure 3 below).
Figure below:
3. Computer model of the mantle flow showing the (Superadiabatic) temperatures. In this particular simulation there is an increase of the viscosity of the mantle with depth. One also includes some heating (15 % of the total) coming from the Earth's core. This heating from the core gives rise to plumes (localized hot upwelling regions) in the model. The color scale is linear and denotes Hot=red, Blue=cold. The upper 100 km of the model are not shown, so that one has a better view of the convective geometry (planform) of up- and downwellings. Note the mantle plume on the right, which is clearly tilted toward a near-surface upwelling, beginning at 1500-1200 km depth. (Figure taken from an earlier publication, Bunge et al., Journal of Geophysical Research, 1997)
There are now very high-resolution computer simulations that give a glimpse at how the mantle may move. Such models are termed Mantle Circulation Models (MCMs), in analogy to the circulation models of the ocean and atmosphere. These models will help researchers to gain a better understanding of how Hotspots move.
One circulation model is shown below.
4. Animation (click on the white triangle to start the movie) of high resolution MCM. In this model there are a number of strong plumes, reflecting the fact that about 30% of total model heat flow enters the mantle from core in this simulation. Colors denote temperature variations (hot = red, cold = blue) from the radial mean. The linear color scale is saturated at -400 K and +400 K, respectively. Continents with color coded topography and plate boundaries (cyan lines) are overlain for geographic reference. Isosurfaces of temperature are displayed for -600 K and +400 K. The +400 K isosurface is clipped in the uppermost 500 km of the mantle to allow views into the mantle underneath the mid-ocean ridge system, which spans large parts of the oceanic upper mantle. Figure from Schuberth et al. [2009].