The best science happens when ideas compete. It happened in the 1900s when geologists were won over to plate tectonics, and it's happening today between two theories about hotspots.
The idea that the Earth's crust moves through cycles of formation and destruction—plate tectonics—came to be accepted in the mid-1960s. It competed with an older theory that instead the whole Earth is expanding. The predictions of each theory were put to the test, and expansion theory withered away.
The Hotspot Problem
Next came the hotspot problem. Plate tectonics explains why chains of volcanoes occur where crustal plates are being subducted into the mantle (see Arc Volcanism in a Nutshell). But there's a leftover, scattered set of volcanoes that occur far from the edges of plates. These were early given the name of hotspots.
The first theory to explain hotspots was that they are the mark of long-lived plumes of hot material rising from fixed sources deep in the mantle, perhaps from its very base. As the crustal plate moves across it, the plume top pushes up strings of volcanoes the way a candle flame burns holes in a sheet of paper passed over it. The type example of such a chain is the Hawaiian Islands and Emperor Seamounts, a string of volcanoes that runs all the way to the plate's subducting edge off Siberia.
Soon afterward the plume model was extended, holding that new plumes build up a large body of magma under a plate, then burst through. Thus hotspots begin with huge eruptions of basalt rock, creating what geologists know as Large Igneous Provinces or LIPs. After that, the plume settles down to a steady output of lava. The Yellowstone hotspot is a well-known example.
This is a fine theory.
- It's easy to teach and it accounts for a wide set of observations.
- Hotspots seem to have a deep source, beneath the wayward motions of the surface plates.
- Some of their magma chemistry suggests a different origin from the rocks of the upper mantle and crust.
- Plumes seem like an upward-moving counterpart of the downward motion of subducted rock into the deepest mantle.
- Lab experiments using suitable fluids, like glycerine and corn syrup, produce similar-looking plumes, as do numerical experiments.
- And plumes suggest ideas for further, exploratory research.
For many years plume theory didn't get much criticism. Partly this was because no one posed persuasive alternatives. Partly it was once the basics of plumes were assumed, researchable questions arose in many different specialties. The assumption was too productive to discard.
Seismic tomography (the technique that yields "ultrasound baby pictures" of the deep Earth) shows shapes much like those predicted by plume theory, but only in the shallow mantle. The deep mantle is harder to image, so the relative lack of direct support for deep plumes didn't trouble the plume theorists.
The Hotspot Alternative
However, there is a new body of theory explaining hotspots not as plumes from the deep, but as stirrings in the upper mantle—its top 400 kilometers—caused from above by the movements of the crustal plates and surface-based cooling. Hotspots are given the more neutral name of "melting anomalies."
Earth scientist extraordinaire Don L. Anderson has been a key player in this alternative hotspot hypothesis. He argues that the original hotspot theorists made assumptions about the upper mantle that new facts have overturned—the whole upper mantle is hotter and more fluid than once thought. Making magma happens easily in the upper mantle, he says, simply by releasing pressure, for example during continental breakups. There is no reason to call for hot magmas from any deeper in the Earth, he believes.
He's not alone. Marine geophysicist Marcia McNutt argued in 2000 that features of the Marquesas hotspot usually explained by hot plumes are of shallow origin, due to relatively buoyant rocks just beneath the crust. A team led by seismologist Eugene Humphreys looked at the Yellowstone hotspot and reported, "Where once a plume origin seemed natural, we now consider a nonplume explanation to be at least as attractive."
Seismic imaging around some major, classic hotspots has not found clear evidence of magma conduits below the transition zone (400 to 660 km deep). At the Yellowstone hotspot, Humphreys' team found that the transition zone is cool, not hot. At the Iceland hotspot, a team led by Gillian Foulger and Bruce Julian reported that the magma conduit appears to bottom out around 400 km.
Chemical evidence of plumes has also fallen short. For instance, Anders Meibom explained in Science in 2008 that a purported signal of material from the Earth's core is discredited. The mantle is complex enough to produce these signals by itself, he points out.
The Battle for the Mantle
Starting around 2000, Foulger joined Anderson in escalating the argument against plumes. For instance, her 2002 article in Astronomy and Geophysics asks starkly, "Plumes, or plate tectonic processes?" coming down firmly for the latter. Since then I have noticed a distinct acknowledgment in the literature that plumes are a hypothesis, not a fact. This is good for science.
The best home page to watch the anti-plumers is at www.mantleplumes.org, a remarkable example of what I call the clear literature. And a blockbuster book published in 2005 by the Geological Society of America, "Plumes, Plates, and Paradigms," has pushed this debate firmly into the mainstream. Its 2007 followup, Plumes, Plates, and Planetary Processes, continues the process, as does Foulger's subsequent Plates vs. Plumes. Mantle studies have become very interesting for everyone.
PS: Another scientific dispute—about the cause of the Cretaceous mass extinction—ended with both sides winning. While the comet impact theory is the consensus explanation for how the dinosaurs died, the volcanic theory is recognized as accounting for many other mass extinctions. The long argument between the two camps created a lot of supporting evidence for both sides, and it drove significant advances in research methods.