Then, in 1971, a Princeton University geologist named Jason Morgan offered a bold explanation: suppose there's a hotspot on the surface of the Earth's core. Why, we don't know. Maybe it has something to do with the way heat escapes from the core's interior, or perhaps it's simply due to a pocket of radioactive materials. Whatever the cause, the spot acts as a geological hotplate, producing a thermal plume that rises through the Earth's mantle until it eventually punches through the surface and emerges as—hurrah!—a big volcano. Since hotspots can presumably be anywhere on the core, hotspot volcanoes can be anywhere on Earth.
By the end of the twentieth century, this once-revolutionary theory had become conventional wisdom. But is it true? Sometimes, universally accepted theories are the ones that least stand up to challenge: consider the fate of the Ptolemaic view of the Solar System or the pre-Einstein view that the speed of light was fixed, relative to a stationary ether.[1] Now, a small group of geophysicists are unleashing a firestorm of controversy by arguing that the hotspot/ mantle-plume theory might not be as correct as everyone has presumed.
The birth and death of such theories is simply part of the cycle of science. But many geophysicists are rooting for the old theory to hold correct. Many years ago, Jules Verne's classic novel, A Journey to the Center of the Earth, postulated an expedition to the Earth's core in which the explorers descended the mouth of Iceland's Snaefellsjökull volcano, and kept going downward for a long, long way. In the (scientifically dreadful) movie The Core, modern explorers took a faster route via an extraordinary burrowing machine.
Real scientists have no hope of such adventures. But if the hotspot theory is correct, they don't need them. That's because the core—or at least material that has been in contact with it—may have come to them via hotspot-derived plumes. “If there are no plumes, it will be a great disappointment,” says Donald J. DePaolo of the University of California, Berkeley. “The romance of plumes is that they are messengers from the bottom of the mantle, and they may be our only chance to get direct information on what's down there."
* * * *
Peeling the Onion
To understand why the theory is under attack, we must begin with a primer on the Earth's interior. It is comprised of three basic layers. The uppermost is the lithosphere (also called the crust), which is where we live, drill for oil, dig mines, and collect samples for laboratory analysis.
The lithosphere is comprised of two basic subdivisions: seabeds, which are made of relatively dense basalt and are only about ten kilometersthick; and continents, which are made of lighter-weight granite and which can be up to sixty or seventy kilometers thick. Both, of course, contain numerous impurities, ranging from sedimentary rocks and precious minerals to organic matter such as oysters, tree roots, and Analog readers.[2]
Beneath the lithosphere is the mantle. It appears to have two distinct sub-layers (called the upper mantle and the lower mantle) with a boundary between them. This is important because the boundary may impede the movement of plumes rising from the lower mantle, and it definitely impedes our ability to figure out what's going on below it.[3]
The mantle is denser than the crust, so the crust floats on top of it. Although the mantle's average temperature of about 2000 °C is hot enough to melt rock, the rock is under so much pressure it remains solid. (Unlike ice, rock expands when it melts, which means that pressure prevents it from doing so.) However, the mantle rocks are very slightly plastic, which means that they can flow—very, very slowly.
Because it is heated from below, the mantle contains convection currents, like a very thick, slow-boiling stew. “Slow” is the operative word. Most scientists believe that plumes are the racehorses of the mantle world, and are galloping along at a pretty good rate if they hit speeds on the order of a foot per year.[4]
It is only when a hot portion of the mantle comes close to the surface that the pressure drops below the critical threshold at which it finally melts. Geophysicists refer to the resulting liquid as “the melt.” Most people call it magma.
The mantle is about 2,800 km thick. Beneath it lies the core, which also appears to have two layers. The core is the least-understood part of the Earth, but all that matters for us is that it's hotter yet (about 3000 °C), and that there is a distinct boundary between it and the mantle. Just as the burner of a stove provides the heat that causes your dinner to boil, heat from the core drives the mantle's convection currents. Hotspots, if they exist, simply provide unusually strong focal points for the same process.
When the mantle's convection currents reach the lithosphere, they buffet it like bubbles floating in a hot tub. The result is that rather than being a single eggshell layer coating the entire planet, the lithosphere is broken into about thirty plates, each moving independently in a process called plate tectonics. Sometimes plates bash into each other, producing enormous mountains, such as the Himalayas. Sometimes they grate against each other, producing zones such as California's San Andreas Fault. Sometimes, they rift apart, allowing magma to upwell from below, and on still other occasions, one plate dives beneath another, melts, and comes back to the surface a hundred miles or so away as a chain of volcanoes, such as Oregon's Cascade Range.
Plate tectonics works perfectly well without mantle plumes. But the standard theory doesn't give you Hawaii, because Hawaii is smack-dab in middle of a plate.
If a mantle plume were to erupt beneath a stationary plate, the result would be a single, enormous volcano. This, in fact, is one explanation for why Mars, whose crust is one large, unmoving plate, sports enormous volcanic cones such as Olympus Mons. But if a plume erupts beneath a moving plate, what you'll get is a chain of volcanoes, as the continuously upwelling magma punches through first one weak spot and then another, as the crust drifts by. The young, active volcano (or cluster of volcanoes) will be at the site of the plume. Farther away, extinct volcanoes will form a nice, nearly straight line, marking the direction in which the plate is moving.
In Hawaii, this is precisely what you see. The Pacific Plate is moving northwestward at about seven centimeters per year, or seventy kilometersper million years, and the islands run in exactly the right direction. Farther out, beyond the last island, the line continues in a string of underwater mountains known as the Emperor Seamounts.
As you move down the chain, the islands (and seamounts) become increasingly old—exactly as they must if the theory is correct, because the farther you go, the longer it's been since they were atop the hotspot. The islands also get smaller, but this doesn't mean the hotspot has changed intensity; rather, the inactive volcanoes have been eroded until nothing remains but stubs.
There is one anomaly, but it is easily explained. About 4,000 kilometers from Honolulu, the Emperor chain shows a distinct kink—a 60-degree bend to the northward. How can a hotspot track produce a kink? Easily, it turns out. Crustal plates are always bumping into each other, and the collisions can cause them to change course. Presumably, something deflected the Pacific Plate at the time these volcanoes were active, causing it to start tracking at a new angle. The kink is one of those anomalies that reinforces a theory because, in retrospect, it should have been predictable.
* * * *
Hot LIPs
The mantle-plume theory is the type of insight about which most scientists can only dream. In one simple stroke, it explained all the puzzles about Hawaii. Needless to say, geologists soon began applying it to other anomalous volcanoes. Today, there are at least 73 suspected plumes, according to a list compiled by Don L. Anderson of the Seismological Laboratory at California Institute of Technology.[5]
The most obvious candidates are island chains: the Galapagos, the Canaries, Tahiti (and associated seamounts), the Azores, and many more. Another candidate is Iceland, although it's a bit different because instead of being a chain of islands, it's a single big one: more like Olympus Mons than Hawaii. The explanation is that Iceland straddles the rift between the North American and Eurasian plates, so that while plate tectonics are ri
pping the island apart, its center stays put, atop the hotspot. This allows the hotspot to keep pumping magma upward, continuously filling the gap.
In principle, there is no need to postulate a hotspot to explain Iceland: something similar occurs at any place where crustal plates are pulling apart. But Iceland is unique because of its size. Something is pumping up a lot more magma there than elsewhere, and hotspot hunters believe they know the answer.
So far, we've only looked at plumes that erupt beneath the sea. What would happen if one encountered a continent? Rather than a chain of islands, you should get a swath of lava flows. The best known of these (presumed) continental hotspots is the Yellowstone Hotspot, which appears to have originated beneath eastern Oregon, Washington, and California. From there, its track moved eastward (as the continent moved westward), creating the Snake River Plain. Now, the plume provides heat for Yellowstone's geysers.
The Yellowstone Hotspot also illustrates what hotspot believers expect to happen when the core develops a new hotspot. At a rising rate of a foot per year, it takes several million years for the plume to reach the surface. En route, traditional theory says it should develop a mushroom shape, with a broad, flat head, and a thinner stem.[6]
When the plume hits the surface, that broad head should produce mammoth “flood” eruptions spread across enormous areas—and this is precisely what we see in the inland Pacific Northwest. The region is dominated by flood basalts, up to 15,000 feet thick, sprawling across an enormous chunk of Oregon and Washington, plus pieces of California and Nevada.[7]
After the plume head comes the stem, which can continue to fuel eruptions for as long as the hotspot persists, possibly for many millions of years. But it's narrower than the head, so downstream from the flood basalts, the plume track should narrow to about the width of the stem, typically believed to be on the order of 100 kilometers. Again, that's what you see with the Yellowstone Hotspot, as the track crosses Idaho to its present location.
Once scientists started looking for locations where plume heads might have erupted, they found many. Two of the most famous are the Siberian Traps (one of the world's largest volcanic provinces) and the Deccan Traps (in India).[8] Similar regions can also be found under the sea, where they are called large igneous provinces or LIPs, and often mark the start of plume-related island chains.
* * * *
How Hot is Hot?
Scientists are human. They love nifty theories that explain lots of seemingly unrelated things. So it's not surprising that the hotspot theory was popular. For years, nobody doubted it. And then, heresy. Shortly after its thirtieth anniversary, the theory came under attack.
I was at one of the scientific conferences where the battle was first joined. Unfortunately, the symposium title didn't say anything about fisticuffs, so I went to sessions that sounded more Analog-appropriate, like “Wet Mars,” or “Astrobiology.” A fellow journalist later told me that what I missed had all the flavor of a fifth-grade playground “debate."
“Yes, it is a mantle plume!"
“No, it isn't!!"
“Yes, it is!!!"
“No, it isn't!!!!"
Or words to that effect. The tone, undoubtedly, was civil (there were no actual fisticuffs), but the rebels were saying that one of the underlying theories of modern geology was hokum, and believers in a theory rarely take kindly to such suggestions the first time they hear them.
By 2004, the debate had gathered momentum but also become less polarized. Even hotspot believers were realizing that there were important questions to be addressed.
One has to do with how hot a plume needs to be before it can rise through the mantle and create an eruption. Until recently, nobody appears even to have asked. Then one of the leading hotspot opponents, Gillian R. Foulger of the University of Durham, England, calculated that for a mantle plume to rise properly, it must be at least 200 °C to 300 °C hotter than the surrounding rock.[9]
Foulger's calculation is interesting, but it's merely a tidbit of computer-modeling information until it's married to the geophysical study known as petrology.
Despite the similarity in words, petrology has nothing to do with petroleum. Rather, it's the study of how magmas crystallize into rocks.[10] Geologists have long known that different volcanoes produce subtly different lavas, thanks to differences in the temperature and chemical composition of the magmas that feed them. Petrology uses these differences either to look backward—from the rocks to the melt from which they were derived—or forward, from the melt to the rocks.
When petrologists examine the rocks erupted from presumed hotspot volcanoes, the results are mixed. Data presented at the Fall 2004 meeting American Geophysical Union indicate that the lava from Hawaii's currently active Kilauea volcano probably comes from a source hot enough to be a mantle plume. In fact, petrologist Paul Asimow of California Institute of Technology went so far as to say that he had never seen a plausible model for a “cold” Hawaii.
But Iceland is a different matter. Isotope ratios indicate that its magma probably originates deep in the mantle (although not necessarily at the bottom), but the petrology shows it to be no more than 100 °C hotter than its surroundings. This means that while there's a hot lump beneath Iceland, it's not hot enough to be a plume. Petrological studies of older parts of Iceland indicate that the magma source may be growing cooler with time. If so, this might mean that what's beneath the island now is the remnant of a plume that's in the process of petering out. Or maybe it's something entirely different. At the moment, nobody knows for sure.
* * * *
CAT-Scanning the Globe
Raffaella Montelli is part of a team from Princeton University that's trying to look directly at plumes, using a technique called “seismic tomography."
Tomography is simply imaging. When doctors do it to the human body, it's called computer-assisted tomography (CAT) and uses low-intensity X-rays to build three-dimensional images by scanning the body from multiple angles. Seismic tomography attempts to use seismic waves to do the same for the interior of the Earth.
When an earthquake occurs, seismic waves travel to seismometers all across the globe, moving not only through the crust, but also through the mantle. The hotter the rock, the slower they move. This means that once you know the exact time and place of any given earthquake, you can figure out how fast the signal traveled to each station. Add in a great many earthquakes from different epicenters and some heavy-duty computing, and you can start mapping the “slow” spots.
The problem is resolution. “We don't have a big choice of where there are earthquakes,” says Norman H. Sleep of Stanford University (who was not part of Montelli's team). “So we have problems the doctor doesn't have.” Also, he adds, “X-rays go in nice straight lines. With seismic waves it's like watching ocean waves veer after they hit a pier."
Montelli's group generated instant controversy when it published its first set of scans in Science, in early 2004.[11] Sleep said that looking at the images was about like using someone else's eyeglasses to watch a press conference or panel discussion. “You'd probably be able to tell that there are people up here,” he said, “but you wouldn't be able to identify us. The tomography is like that—just tantalizing."
Foulger was less charitable. Very few seismologists believe that the tomographic images had enough resolution to actually show anything, she said. “I think the only people who put any faith in them are non-experts,” she said, “or the people who do the models."[12]
In December 2004, Montelli was back. This time, her team expanded the analysis to include two types of seismic waves, called S-waves and P-waves, improving the resolution. What she had, though, was just as tantalizing as ever. In some places, plumes appeared to descend all the way to the core. Others were decidedly odd: broken off, disconnected at the boundary between the upper and lower mantle, or bending sideways as though swaying under the influence of other currents.
Based on her new tomography, Montelli claimed to have found �
�well-resolved” plumes in six oceanic locations (Ascension Island, the Azores, the Canaries, Easter Island, Samoa, and Tahiti) plus “robust” findings in three others (Cape Verde, Cook Island, and Kerguelen Island in the South Pacific). More interestingly, she found an incipient plume—not yet reaching the surface—beneath the Coral Sea.
Other plumes, however, including Iceland, appeared to be confined to the upper or middle depths of the mantle, indicating that they may either have shallow sources or be tapering out.[13]
Skeptics are still not convinced that she is seeing actual plumes or plume fragments. That's because temperature isn't the only factor that can create seismic “slow” spots. Other possibilities include unusual rock composition, melting, and relatively high water content—any of which could occur beneath volcanically active areas, with or without a plume.
* * * *
Bow Waves and Foraminifera
Without better tomography, the plume debate is branching in new directions as scientists seek innovative new ways to test the plume theory.
One test comes from Norman Sleep. Sleep, who was one of the early skeptics, set out to disprove the theory at its source: Hawaii. (Like many geophysicists, he figured that if Hawaii isn't a hotspot, nothing is.)
Sleep was intrigued by the fact that the seabed near Hawaii has a large bulge, about 1,000 meters high and 1,000 kilometers wide. Something similar shows up along the track of the putative Yellowstone Hotspot. Hotspot theory viewed these bulges as uplift zones, created as the crustal plates plow across the upwelling plumes. But the swells extend upstream of the hotspot—well ahead of where the plate is supposedly tracking across the plume. Sleep didn't believe a plume could create something that looked so much like a bow wave, so he modeled it, and got a surprise: his model predicted exactly the features seen in Hawaii and Yellowstone. This unexpected result convinced him that Hawaii and Yellowstone probably are the result of plumes. But, he is quick to point out, this doesn't mean that everything that's ever been catalogued as a hotspot really is one.
Analog SFF, July-August 2006 Page 7
