A DAY just isn鈥檛 what it used to be. Since the moment the Moon was formed it has played a gravitational tug of war with the Earth, generating friction, sapping energy and gradually slowing our planet鈥檚 rotation. For the trilobites, 530 million years ago, one year contained about 420 days and each day lasted about 21 hours. Now we get a mere 365 days every year and our days last for 24 hours. And as time goes by, days and nights will continue to stretch. The Earth is winding down.
So why, around 180 million years ago, did the Earth鈥檚 spin suddenly increase? And not for the first time. The same thing seems to have happened once before around 400 million years ago. It鈥檚 a mystery geologists have struggled to solve. Until now.
We know that the Earth鈥檚 spin has changed over the millennia thanks to geological and fossil records. In coastal areas, for example, layers of sediment are deposited on the seabed as the tide goes in and out. And creatures such as corals and molluscs add new layers of calcium carbonate to their body each day, just as trees add a ring for each year. Since the thickness of these layers varies with the seasons, and with the cycle of low and high tides that occur during the lunar month, scientists can use them to work out the number of days in the year at a particular time in the geological past.
Advertisement
These ancient clocks seem to indicate that the Earth is winding down, like a spinning top gradually running out of energy. Except for two puzzling anomalies. The first appears to occur between 420 and 360 million years ago. Cross sections through sediments, corals and molluscs from these periods don鈥檛 seem to fit the winding-down model. Instead 鈥� in an idea first proposed in 1974 by Ken Creer, a geophysicist at the University of Edinburgh 鈥� the discrepancies suggest that the Earth could have put on a spurt of acceleration lasting a few tens of millions of years, making the days and nights slightly shorter than expected. The records seem to indicate that the same thing happened again about 200 million years later.
There are only two ways to increase the Earth鈥檚 rate of rotation. You could give it more energy, but that would require some external force. A collision with an asteroid or small planet might do the job, but a collision that provided exactly the right amount of energy at precisely the right spot would seem unlikely. The other option is that the Earth somehow changed the distribution of its mass, shifting weight inwards towards the planet鈥檚 centre. Like an ice skater pulling in their arms, this would decrease the moment of inertia and make the planet spin faster.
However, while scientists accept that material does shift around in the Earth鈥檚 mantle, most would view a shifting of a large mass towards the centre of the Earth as about as unlikely as a perfectly timed asteroid collision. And since the anomalies in the fossil and geological record remained controversial, many researchers simply gave up on any attempt at explaining them.
Certainly Philippe Machetel, Director of the Institute of Earth Sciences at the University of Montpellier in France, had no interest in solving the mystery when he began studying the Earth鈥檚 interior in 1990. Yet 12 years later he has reason to believe that geologists should no longer ignore the anomalies. His research has uncovered a catastrophic geological phenomenon that could explain the mysterious data 鈥� the first real evidence that the fossil clocks are right.
Three years ago, Machetel and his student Emilie Thomassot developed a new computer simulation that uses fluid dynamics to model the heating and behaviour of rock inside the Earth. There are two distinct layers beneath the Earth鈥檚 crust: the core and the mantle. Within these layers there are variations in rock density and chemistry and geologists subdivide the mantle into three more layers: the lower mantle, a transition zone and the upper mantle (see Diagram). These layers are all defined by abrupt changes in density which can be seen in the reflections and refractions of seismic waves travelling through the Earth.
Deep beneath our feet, rock is constantly on the move. As it flows between the upper mantle and the transition zone, around 400 kilometres down, it changes gradually, melting as it slides downwards or solidifying as it rises towards the surface. But it鈥檚 a very different situation at a depth of around 670 kilometres, at the boundary between the transition zone and the lower mantle.
This boundary marks a sudden transition between two phases of molten rock, a dense gloopy rock above and the hotter, softer rock beneath. Above the boundary heat is continually lost as it convects through the Earth and out into space. But heat generated deep within the lower mantle is trapped at the boundary. 鈥淚t鈥檚 like a pressure cooker,鈥� explains Machetel. 鈥淏eneath the boundary the temperature just keeps rising and rising.鈥� Since the boundary is so abrupt, any movement of material from one side to the other requires considerable energy.
To find out how rock in this boundary zone moves as the temperature and pressure in the region rise, Machetel and Thomassot designed a computer simulation that could compress time, modelling the geological history of the planet in days rather than millennia. When they ran their simulation, they discovered that as the lower mantle heats up, the transition zone becomes unstable and less able to support the weight above it. Eventually conditions reach a critical point at which weak spots in the zone fail, allowing a cold, dense chunk of the upper mantle to break away and sink into the lower-density magma below.
Like a giant avalanche this chunk of mantle gathers momentum as it goes: by the time it reaches the boundary layer with the lower mantle it is travelling at high speed and with enough energy to punch through the boundary completely. And it carries on going, sliding through the lower mantle like a stone dropped into a pond. Finally, when it reaches the edge of the core it spreads out, pushing the hot material beneath out of the way and forcing plumes of hotter rock to burst upwards. In just 10 million years 鈥� the blink of an eye in geological time 鈥� a massive blob of rock the size of the Moon has shifted towards the centre of the Earth.
Machetel and Thomassot began to wonder about the consequences of such an event. One possibility is that a mantle avalanche would suck all the continental plates together on the Earth鈥檚 surface. 鈥淭he continents float on the mantle, like leaves on a river,鈥� says Machetel. 鈥淎n avalanche in the mantle would probably make them gather above the avalanche point, creating a supercontinent.鈥�
But soon they also realised that such an avalanche of dense rock moving towards the core would decrease the Earth鈥檚 moment of inertia and speed up its rotation. 鈥淵ou can compare it to a spinning ice skater,鈥� says Machetel.
Since the results of almost any simulation depends on a number of variables and starting conditions, the researchers began to hunt for other independent evidence that might support their results 鈥� and that led them to the anomalies in the fossil clocks.
When they examined data from sediments, coral and shell growth, Machetel found that the most recent increase in the speed of the Earth鈥檚 rotation occurred around 180 million years ago, which coincides with the date of the mantle avalanche in their simulation. 鈥淲hen I found the fossil evidence for acceleration I was very happy,鈥� he recalls. And, says Machetel, the fossil clocks are not the only evidence that the avalanches occurred. There are two other independent indications that something strange was going on deep within the Earth at around this time.
First, there is geophysical evidence that the Earth鈥檚 tectonic plates started moving in an unusual manner around 170 million years ago. Whenever molten lava solidifies, for example, a signature of the Earth鈥檚 magnetic field at that moment is frozen into it. As the rock starts to cool, the magnetic minerals within it align themselves with the Earth鈥檚 magnetic field and when the rock solidifies, the minerals are fixed in place. The Earth鈥檚 field varies according to latitude and if the rock is part of a tectonic plate that then starts to move, the rock鈥檚 鈥渘orth鈥� will cease to match up with true magnetic north. So measuring the orientation of magnetic north in rocks of different ages provides a measure of how the plates have moved over the millennia.
This palaeomagnetic record suggests that for most of Earth鈥檚 history the plates have simply jostled against each other. But 170 million years ago all the plates apparently moved together, rather quickly, in the same direction at the same time.
This could certainly happen if Earth suffered a sudden change in its distribution of mass (New 杏吧原创, 18 August 2001, p 34). Imagine fixing lead weights close to the top and bottom of a basketball and then attempting to spin it on your finger. It would wobble crazily as the weights unbalanced it. But if you then moved the weights so that they were evenly distributed around the ball鈥檚 equator, it would spin happily. Similarly, a dense blob of mantle crashing towards the core would change the mass distribution inside the Earth and the easiest way to compensate for this would be for the crust, mantle and core to shift, realigning the distribution of mass and stabilising the Earth. The period of synchronised tectonic movement appears to have started very soon after Machetel鈥檚 proposed mantle avalanche and, as he would expect, the geological evidence shows that the heaviest parts 鈥� mountain ranges, for example 鈥� did move towards the equator.
The second piece of evidence to support Machetel鈥檚 theory occurred some 40 million years after the tectonic movement. It concerns the reversal of Earth鈥檚 magnetic poles.
Every now and then, Earth鈥檚 magnetic field flips direction 鈥� north becomes south and south becomes north. This usually occurs at an interval of between a few tens of thousand and a few million years. However, around 130 million years ago, the Earth鈥檚 magnetic field seemed to get stuck. It didn鈥檛 reverse direction for another 40 million years. This is an exceptionally long period without a reversal and suggests that something had tampered with the Earth鈥檚 interior magnet.
It鈥檚 impossible to say exactly what could cause such an anomaly because scientists are still struggling to understand how the Earth generates its magnetic field, and why it should flip (New 杏吧原创, 30 March 1996, p 24). What they do know, however, is that somehow the convection of charged iron particles in the Earth鈥檚 core creates the field. And in 1999, Gary Glatzmaier of the University of California, Santa Cruz, showed that large changes in temperature at the boundary between the core and the mantle could prevent the magnetic field from changing direction (Nature, vol 401, p 885). Machetel and Thomassot think that the quenching effect of a cold mantle avalanche enveloping the Earth鈥檚 core could have upset the flow of charged particles and prevented the magnetic field from flipping direction for many millions of years. The researchers recently published their results in Earth and Planetary Science Letters (vol 202, p 379).
Although all the strands of evidence seem to tie together, not everyone is convinced by Machetel and Thomassot鈥檚 theory. 鈥淚t鈥檚 an intriguing concept, which could work in principle,鈥� says Walter Kiefer, of the Lunar and Planetary Institute in Houston, Texas, 鈥渁lthough I鈥檓 still a little sceptical.鈥� And even the clock anomalies remain controversial: counting patterns in sediment and fossil records can be tricky. 鈥淲e can鈥檛 be absolutely sure that an acceleration happened, but it is certainly a possibility,鈥� says Colin Scrutton of Durham University, one of the researchers who first interpreted the fossil clock record.
Paul Tackley, a planetary scientist at the University of California at Los Angeles, has a more specific criticism: the size of the avalanches deduced from the simulations wouldn鈥檛 be large enough to cause the observed changes in day length of day. 鈥淭he model produces mantle avalanches that are about 100 times too small to explain the observations,鈥� he says.
But Machetel points out that there could be several effects not taken into account in their model that could amplify the effect of the avalanche. 鈥淲e don鈥檛 model the changing shape of the core-mantle boundary, and this could be important,鈥� he says. And the sinking convection currents created by the avalanche would probably draw down another slab of high density magma with them, which would further decrease the planet鈥檚 moment of inertia, he says. But it is also possible that inaccuracies in the fossil clocks have exaggerated the acceleration of the Earth. He and Thomassot are now working to improve the accuracy of their simulation.
Machetel suspects that an avalanche inside the Earth could also have caused the other earlier anomalous acceleration shown up in the fossil clocks. It鈥檚 more difficult to get supporting geological evidence from this period, as it was a fairly tempestuous, volcanic era, during which the continents were splitting apart. But Machetel鈥檚 model indicates that these deep-Earth avalanches will occur every few hundred million years, depending upon the heating from the Earth鈥檚 core. It鈥檚 hard to be precise about dates because the heating is a chaotic process, but there really seems to be a link between the accelerations and the avalanches, he believes.
Vincent Courtillot, a geologist at the University of Paris 7, agrees. 鈥淚 believe avalanches are a potentially important concept in geodynamics,鈥� he says. The case for mantle avalanches is far from proven, but Machetel is still piecing the evidence together. More powerful computer models could soon put his theory on a firmer footing, he believes.
If he鈥檚 right, of course, the Earth could change gear and speed up again in the future. So next time you complain that there are not enough hours in the day, just be thankful for those you have.