ABSTRACT
Maybe the continents did not grow quickly then stabilize, neither growing nor shrinking over time. Perhaps they not grow by a steady output of mantle material over time. Instead, it is possible that they grew in spurts, caused by occurrences of massive outpourings of mantle lava. In this essay, a scientist shows a curious correlation between basalt helium isotope ratios and zircon ages indicating that perhaps the latter is the correct model. Interestingly, it is Jupiter that helps strengthen his argument.
INTRODUCTION
Like those invisible structures called atoms that make up matter or just as the cores wherein stars create their energy are utterly hidden from view, so is the past locked away from our sight for ever, invisible and directly unobservable. However, if we turn our eyes toward space and look out across the galaxies, then, using some plausible, sensible and agreed upon assumptions coupled with some well known facts, we accept that we are looking over to something like our past. Thus, when we see a sun-like star, ten, twenty or one hundred million light years away, so we know that we are seeing what our own sun looked like some ten, twenty or one hundred million years ago.
In this essay, I wish to look at an unobservable past on earth which has been locked away by time. Just as the invisible atom and the core of a star leave their finger prints for us to observe and so make sense of them, so does the past leave its fingerprints on the present for that is how the present got to be. And similar to those finger prints of atoms and stellar cores, these finger prints from eons ago can inform us as to what may have happened way back then.
Here am discussing the article Helium isotopic evidence for episodic mantle melting and crustal growth, which appeared in an April issue of Nature (1). The article deals with evidence for several pulses of magma deposition from the mantle onto continental crust, over a time period stretching from about 1 billion to 4 billion years ago. How the evidence is used provides insight into the methods by which science is done.
The story is fascinating. It has competing theories at the cutting edge of scientific research. It deals with data, their uncertainties, and the testing of ideas. It shows how theories point the way for future research by opening up new questions and how the facts, Facts, FACTS of scientific research constrain model construction and explanation. That is, those Facts narrow down scientific theorizing, confining it to the plausible.
HEALTH WARNING
This article is based on my understanding of research reported in a professional journal. Please do not accuse the relevant researcher of incompetence should you find errors or concepts in this essay that make no sense. This is my work but I would not count on it to pass an exam. Nevertheless, hopefully it will give you an idea as to how real scientific research is done, whether it be in medicine, materials, macro evolution, cosmology, physics, or geology.
This is all about the stuff of great ideas, plus the FACTS to support them.
BACKGROUND
It is known that the earth (radius about 6,500 km) is made up of three general components. From the inside of the earth to the outside they are the core, the mantle and the crust.
The core, about 3,500 km in radius, is made up of two parts, a solid inner and a liquid outer. The mantle, a thickness of some 3,000 km, is also made up of two parts, a solid lower and a semi molten upper. The crust is a mere few tens of kilometers thick. However it also has two components, a relatively dense basalt part which underlies the ocean floor and a less dense continental part which constitutes the continents. Accordingly the continents behave like a scum, bobbing on top of the underlying mantle. Given its density, continental crust can never sink back into the mantle whereas the ocean crust can.
It seems that volcanism is the mechanism forming the crust. At the mid ocean ridges, molten material from the mantle rises to the surface, erupting in a relatively gentle manner as lava, which on cooling, becomes the basalt. These eruptions, occurring all along the mid ocean ridges, produce a basalt ocean floor which spreads away from the ridge. At the edges of many continents, this dense basalt crust plunges back into the mantle underneath the continent, in many cases dragging overlying sediment (generally eroded continental crust from the land we live on) and slabs of continental edge with it. The associated friction due to the plunging oceanic crust as well as extreme pressures at depth, leads to a partial melting of the rock allowing less dense molten materials to rise to the surface either near a continent or actually on a continent. This magma contains a lot of water as steam, (from its previous association with the ocean), and so it erupts more violently than that at the mid ocean ridges. The bulk of material these continental volcanoes deposit is given the name andesite and this rock is very similar in chemical composition to the rest of the continental crust.
Generally away from either of these places are ocean islands such as Hawaii. These islands are volcanic and basaltic. That is, they are composed of the same kind of material as that erupting from the mid ocean ridges mantle material.
All of this can be seen in the first attachment.
OTHER THINGS YOU SHOULD KNOW
1) While isotopes are used by geologists to date rocks, they are also used to ascertain the history of rocks. Isotopes are different kinds of atoms of the same atom. What you may ask? Any one particular element has a unique set of chemical properties defined by the number of negatively charged electrons it has whizzing around its nucleus, for it is these electrons which do certain things allowing atoms to bind together, thereby producing molecules, compounds and so on. For an atom to be neutral, that is, to carry no net electric charge, the negative electrons on the outside must be balanced by an exact number of positive protons in the nucleus. Take helium as an example. A helium atom will have two orbiting electrons and two protons in its nucleus. However, the nucleus of an atom can also contain neutrons which have no electric charge. It is these differing numbers of neutrons in the nucleus of helium atoms which give rise to isotopes. He3 has two electrons in orbitals and two protons (to balance the electrons), and one neutron in its nucleus. The two protons plus one neutron make He3. He4 still has two electrons and two protons, but it also has two neutrons. Thus the two protons and two neutrons make for He4.
2) Helium is an atom which has two isotopes and two methods of production. Helium 3 (He3) is primordial. It was created at the time of the Big Bang. Helium 4 (He 4) is a wee bit heavier than Helium 3. It is not primordial. It is created as a by product of the radio active decay of uranium to lead.
Uranium itself comes from the explosion of stars. Like helium, uranium and lead have isotopes. No isotope of uranium is stable. Three isotopes of lead (206, 207 and 208) are stable. They do not decay. Lead 206 is produced from the decay of uranium 238. When uranium 238 decays it gives rise to other unstable elements (including unstable isotopes of lead) which then decay. This happens in a long series of steps, at the end of which is lead 206. Importantly, during this decay process, He4 is produced. To see all the isotopes of lead and uranium known so far, check out reference 5.
Thus we have:-
KABOOM ----> H2 (lots of) + He3 (some) (NB The Big Bang is not really an explosion)
U ----> Thorium ------> Other unstable isotopes through a series of steps ------> Pb206 + He4
At this stage you may not be able to see how isotopes can give clues to the history of rocks. However, if my essay is any good, then it will help you understand this bit about isotope ratios and rock history particularly with respect to helium.
3) Partial melting. Generally rocks are made of many different crystals and even bits of other rock. These various components will have differing melting temperatures. At increasing depths within the earths mantle, temperatures will be reached at which the various components of rocks will melt. This differential melting is called partial melting. By it, lighter elements (or those with lower melting points than the parent rock) can leave their parent rock and reach the surface as molten lava. The lava, on cooling, forms new rocks which have different element abundances and densities to those of the parent rock from which they were derived.
HERE WE GO
Imagine that both the upper and lower parts of the mantle started with the same amounts of helium 3 and uranium (U). Over time the U would give rise to the same amount of He4 in both the upper and lower mantle such that, for example only ,upper mantle (He4/He3) = lower mantle (He4/He3). Should any component of mantle rock get hot enough (because of depth of burial) and become magma, then it will be degassed of most of its helium on reaching the surface. (Like rising hot air, magma rises with respect to the colder denser rock around it). It will still contain its heavy uranium such that after cooling back into rock and over time, He4 will build up, leading to a higher ratio of He4/He3 relative to that rock which is still trapped within the mantle. If this continental rock gets recycled back into the mantle, and the upper and lower mantles are isolated and there is no mixing, then part of the mantle can, over time, show a signature of depleted He3. For pictures of this, see the third attachment.
Now for some more FACTs and some theory. On average, ocean island basalts (OIB) have a lower He4/He3 ratio that mid ocean ridge basalts (MORB). Ocean islands in this context are those islands like Hawaii which formed from the same kind of material which wells up from the mantle at the mid ocean ridges. However given this disparity in the helium isotope ratio, it was believed that the OIB came from the lower mantle where the initial helium incorporated at the earths origin has been retained, while the MORB came from the upper mantle where, due to escape into the earths crust followed by recycling back into the mantle, the He3 content had decreased, thereby giving rise to a higher He4/He3 ratio. This model therefore required that the two mantle layers be isolated such that helium disparities could not be smoothed out over time due to mixing or some other process.
However, over recent years, the concept of isolated upper and lower mantle reservoirs has been called into question. And it has been determined that helium and uranium/thorium may behave in melts differently to what had been thought before, with consequences for rock histories derived from this information. If this turns out to be the case that the previous (reasonable assumptions) are indeed incorrect then it kind of throws some models of continental crust formation into doubt. Most researchers agree that the continental crust comes mainly from the mantle and constraining parameters for associated models are now called into question, given that they relied on an incorrect model of the mantle and/or incorrect data relating to the behavior of isotopes.
[Aside: Reference (2) is an example of an experiment that calls old models into question. In this particular case, laboratory tests, using new experimental techniques, have been applied to natural and synthetic olivines (a mineral important in the formation of rocks). These have shown helium to be more soluble in molten rock, relative to uranium and thorium than previously thought.]
Parman argues that it is time for models relying on the evolution of helium isotopes to be considered with more seriousness. One reason for this plea is that the above mentioned experiments which throw the old models into doubt show that rather than MORB and OIB pointing to an isolation between the upper and lower mantles, they could instead be pointing to times of mantle melt depletion.
That is, the very FACTS that are calling the old models into doubt could be showing that the helium ratios of the OIB indicate the times at which the mantle underwent depletion of its molten component. Where did this mantle depletion go? Obviously upwards to make new continental crust. (And associated with this would be changes in the ratios of isotopes of various elements in this particular case, helium 4 and helium 3.)
So Parmans letter to Nature is his exploration of this idea, using the facts, Facts, FACTS at hand.
He notes that because OIB and MORB come from the mantle, they provide the primary constraints on the He isotropic composition of the mantle. Parman then goes on to discuss the results of measuring the helium isotope values in rocks from various volcanic islands and from mid ocean ridges around the world. Taking into account the amount of sampling done at various locations around the world which can affect both the strength of any isotope signal peaks (the height of a peak) or the range of signal peaks, he presents a series of graphs that demonstrate:-
1) on average MORBS have consistently higher He4/He3 ratios than OIBs and these MORB ratios have a dominating peak (in terms of probability distribution) at 90x10^3. That is, you are much more likely to find a MORB with 90,000 times more He4 than He3 than you are a ratio a bit less than this or a bit more than this. There is another MORB peak at 85x10^3 but it is the 90,000 peak plus four others that are important to this story.
2) the OIBs have helium isotope ratios ranging from 18x10^3 to 120x10^3 and peaking at roughly (20, 30, 36, 42, 50, 56, 64, 69,86 and 90)x10^3. For various reasons some OIBs show only a few peaks. The reasons appear to be because of limited sampling at particular islands more than anything else. (On islands with a high sampling number, the spread of isotope values is large and the peaks tend to be more distinct.)
3) where the OIB and MORB data overlap, the peaks still show up at matching places.
See the fourth attachment.
It is the He4/He3 ratio peaks at 30000, 42000, 56000, and 69000 which are important, for Parman demonstrates that those peaks can be shown to match ages from particular zircons which some researchers use to argue that the continental crust has grown in massive episodic spurts (reference 3). This is unlike other models in which the continents formed quickly and early and from then on continued to recycle themselves, neither growing nor shrinking and unlike other models in which the continents continually and steadily grew.
The problem is in dating the samples at which the helium isotope peaks occur.
One method of dating would be to return to all the islands, collect hundreds of samples and attempt to date them. Clearly this would be an impossibility very expensive in both time and money. (Parmans research was done from existing data sets.)
Another way would be to use a model for the evolution of helium within the earth as a basis for calculation. However, given the large number of models, it would be impossible to select any one model and use it to calculate a reliable age for these helium peaks.
A third way would be to see if there is some kind of correlation between the helium peaks and ages for the supposed melting events that caused the spurts in crustal growth.
Parman chose this latter course, noting that the zircon studies have shown the main peaks in zircon ages (and by implication bursts in continental growth) to be at 1.2, 1.9, 2.7 and 3.3 thousand million years ago (Gya). There is a complicating factor though, for as he points out, the peaks in occurrence of zircons at these ages could well be an indication, not of spurts in continental growth but rather of pieces of continental crust that have randomly escaped tectonic recycling. In tectonic recycling, continental crust with its zircons erodes and ends up on the ocean floor, where it gets placed back into the mantle at the ocean floor subduction zones. The zircons melt and have their radioactive clocks reset. Magma forming through partial melting rises to the surface and erupts. New zircons form but with much younger ages. If this is what really happened to produce the clusters in zircon ages, then a correlation with the helium isotope ratios should be most unlikely.
A point to begin correlating both sets of data is the fact that each set contains a distinctive drop in peak production. Thus, there are a large number of zircon peaks up till 1.2 Gya and relatively few after that. Similarly the He4/He3 peaks show a large number up to a ratio of 70x10^3 and relatively few after that. So Parman used the gap as a starting basis for correlating the two sets of data. He found the following:-
1) Given that both sets of data are monotonically increasing, other ways of aligning them exit. However, it is this one which uses the gap which shows the highest correlation with a correlation coefficient of 0.9986.
2) Importantly, the slope of the line is not random. If one projects the correlation line to the present day, it matches the value of the highest MORB of approx 90,000 for He4/He3. It is reasonable to suppose that the MORB trap into mantle which has undergone the most He3 depletion and so one should expect that the correlation line would intercept this value for time = 0 since mid ocean ridge basalts are very young. (These ridges are where the sea floor is continually being created.) And if one projects the line back 4.6 Gya (to the time of the earths formation) one gets a ratio of 5600 which is very close to the estimate for Jupiter of 6000 (see reference 4). And Jupiter, it is believed, retains the original solar system helium 3 content, simply because of its massive gravity. Helium can easily escape from the earth, but not Jupiter. (Given that these He4/He3 ratios are from natural systems, then other natural systems of helium isotopes should fall onto the same correlation line as that obtained from the OIB and zircon age data.)
The importance of these two endpoint numbers is that MORB record the mantle as it is now with its current helium 3 depletion. Jupiter records the original solar system helium 3 content and hence the isotope ratio the earth began with. That the experimental correlation line connects to these points indicates very strongly that the intervening peaks could well be recording mantle depletion events and associated spurts in the growth of continental crust.
See the fifth attachment.
Parman argues that it is the independent prediction of these two fundamental values by the helium-zircon correlation line, that indicates the plausibility of a casual relationship between the helium peaks and those of the zircons. As Parman notes, if the correlation was spurious, or the peaks artifacts from over-sampling, then the plots would result from three highly improbable coincidences a fluky correlation in peaks from island to island, the accidental matching of the zircon age pattern, and the fluky prediction of both the Jovian (initial Earth) and present day MORB values. The slope of the line appears to be robust in that it is not controlled by one point. If the four main helium peaks are used, then a line with an even stronger correlation coefficient is produced.
Two helium peaks do not have matching zircon peaks. Parman suggest that further studies may well reveal zircon peaks at these points.
So Parman concludes that this correlation is significant in that it adds weight to the idea that the continents did grow in spurts and that these spurts are recorded in both the mantle (from the OIB and MORB data since both kinds of rocks come from the mantle), and in the continental crust (in the zircons).
If this correlation is real, then Parman notes that it throws open many issues including:-
1) How and why did these massive building events occur, and what were they like?
2) How does the helium isotope signature remain in the various mantle domains, against mixing in the mantle due to the slow convection of rock occurring there?
3) The constraints on other models of helium evolution in the mantle and crust. Parman writes that there are numerous models of helium evolution (some relying on a separate upper and lower mantle that are isolated from each other, others relying on un-isolated mantle components etc.)
So has Parman uncovered strong evidence that bolsters the idea that the continents grew in a series of massive spurts over billions of years?
SUMMARY
New data on how molten rocks retain (or lose) their component elements of helium, uranium and thorium, have thrown into doubt some theories about how the continents grew.
Parman used this same data to argue that helium isotopes may provide a better method of modeling some aspects of continent growth.
Using helium isotope ratio data from various mid ocean ridges and many ocean islands that were made of basalts from magma which came from the mantle, Parman showed that there were several peaks in the ratios of helium isotopes. These peaks indicated that one was more likely to find rocks with these helium ratios than rocks with other ratios.
Given the way in which helium is formed and how its isotopes behave during the creation and cycling of rocks, Parman reasoned that this could tell him something about how the continental crust formed. He noted that these helium ratio peaks could be strongly correlated with evidence from zircons which indicated that continents may have grown from several massive inputs of lava from the mantle.
The associated correlation line predicted the initial helium isotope ratio of the earth and the current helium isotope ratio from mid ocean ridge basalts. This lent credence to the idea that the helium spikes were indeed indicative of episodes of mantle helium depletion and hence continental growth.
CONCLUSION
The argument from Parmans facts, Facts, FACTS looks pretty convincing, particularly when you look at the correlation, its robustness as well as the prediction of those two important values, the current MORB and the initial earth helium isotope ratios.
But nothing in science is simple.
So stay tuned. Better still, go out, do some learning and lots of reading. Then next time you spot an article related to this, you will not be able to ignore it, for you will have more to ponder. If these isotopes and zircons are telling us what some think, can you imagine what those times may have been like?
Regards, Roland
REFERENCES and NOTES
(1) S.W. Parman, Helium isotopic evidence for episodic mantle melting and crustal growth, Nature 446, 19 April 2007, pages 900-903.
(2) Stephen W. Parman, Mark D. Kurtz, Stanley R Hart & Timothy L. Grove, Helium solubility in olivine and its implications for high He3/He4 in ocean island basalts, Nature 437 20 October 2005, pages 1140 1143.
(3) A. I. S. Kemp, C. J. Hawkesworth, B.A. Paterson & P. D. Kinny, Episodic growth of the Gondwana supercontinent from hafnium and oxygen isotopes in zircon, Nature 439 2-February-2006, pages 580 583.
(4) The Nature article actually publishes values such as (92.2+/-0.7)x10^3 (+/-s.e.). I have written this as approximately 90000 for the sake of making things simpler.
(5) Isotopes of lead:-
http://ie.lbl.gov/education/parent/Pb_iso.htm
The half-life column tells whether the isotopes are stable or not. Seconds = s, milliseconds = ms, years = y and so on.
Isotopes of uranium:-
http://ie.lbl.gov/education/parent/U_iso.htm
Maybe the continents did not grow quickly then stabilize, neither growing nor shrinking over time. Perhaps they not grow by a steady output of mantle material over time. Instead, it is possible that they grew in spurts, caused by occurrences of massive outpourings of mantle lava. In this essay, a scientist shows a curious correlation between basalt helium isotope ratios and zircon ages indicating that perhaps the latter is the correct model. Interestingly, it is Jupiter that helps strengthen his argument.
INTRODUCTION
Like those invisible structures called atoms that make up matter or just as the cores wherein stars create their energy are utterly hidden from view, so is the past locked away from our sight for ever, invisible and directly unobservable. However, if we turn our eyes toward space and look out across the galaxies, then, using some plausible, sensible and agreed upon assumptions coupled with some well known facts, we accept that we are looking over to something like our past. Thus, when we see a sun-like star, ten, twenty or one hundred million light years away, so we know that we are seeing what our own sun looked like some ten, twenty or one hundred million years ago.
In this essay, I wish to look at an unobservable past on earth which has been locked away by time. Just as the invisible atom and the core of a star leave their finger prints for us to observe and so make sense of them, so does the past leave its fingerprints on the present for that is how the present got to be. And similar to those finger prints of atoms and stellar cores, these finger prints from eons ago can inform us as to what may have happened way back then.
Here am discussing the article Helium isotopic evidence for episodic mantle melting and crustal growth, which appeared in an April issue of Nature (1). The article deals with evidence for several pulses of magma deposition from the mantle onto continental crust, over a time period stretching from about 1 billion to 4 billion years ago. How the evidence is used provides insight into the methods by which science is done.
The story is fascinating. It has competing theories at the cutting edge of scientific research. It deals with data, their uncertainties, and the testing of ideas. It shows how theories point the way for future research by opening up new questions and how the facts, Facts, FACTS of scientific research constrain model construction and explanation. That is, those Facts narrow down scientific theorizing, confining it to the plausible.
HEALTH WARNING
This article is based on my understanding of research reported in a professional journal. Please do not accuse the relevant researcher of incompetence should you find errors or concepts in this essay that make no sense. This is my work but I would not count on it to pass an exam. Nevertheless, hopefully it will give you an idea as to how real scientific research is done, whether it be in medicine, materials, macro evolution, cosmology, physics, or geology.
This is all about the stuff of great ideas, plus the FACTS to support them.
BACKGROUND
It is known that the earth (radius about 6,500 km) is made up of three general components. From the inside of the earth to the outside they are the core, the mantle and the crust.
The core, about 3,500 km in radius, is made up of two parts, a solid inner and a liquid outer. The mantle, a thickness of some 3,000 km, is also made up of two parts, a solid lower and a semi molten upper. The crust is a mere few tens of kilometers thick. However it also has two components, a relatively dense basalt part which underlies the ocean floor and a less dense continental part which constitutes the continents. Accordingly the continents behave like a scum, bobbing on top of the underlying mantle. Given its density, continental crust can never sink back into the mantle whereas the ocean crust can.
It seems that volcanism is the mechanism forming the crust. At the mid ocean ridges, molten material from the mantle rises to the surface, erupting in a relatively gentle manner as lava, which on cooling, becomes the basalt. These eruptions, occurring all along the mid ocean ridges, produce a basalt ocean floor which spreads away from the ridge. At the edges of many continents, this dense basalt crust plunges back into the mantle underneath the continent, in many cases dragging overlying sediment (generally eroded continental crust from the land we live on) and slabs of continental edge with it. The associated friction due to the plunging oceanic crust as well as extreme pressures at depth, leads to a partial melting of the rock allowing less dense molten materials to rise to the surface either near a continent or actually on a continent. This magma contains a lot of water as steam, (from its previous association with the ocean), and so it erupts more violently than that at the mid ocean ridges. The bulk of material these continental volcanoes deposit is given the name andesite and this rock is very similar in chemical composition to the rest of the continental crust.
Generally away from either of these places are ocean islands such as Hawaii. These islands are volcanic and basaltic. That is, they are composed of the same kind of material as that erupting from the mid ocean ridges mantle material.
All of this can be seen in the first attachment.
OTHER THINGS YOU SHOULD KNOW
1) While isotopes are used by geologists to date rocks, they are also used to ascertain the history of rocks. Isotopes are different kinds of atoms of the same atom. What you may ask? Any one particular element has a unique set of chemical properties defined by the number of negatively charged electrons it has whizzing around its nucleus, for it is these electrons which do certain things allowing atoms to bind together, thereby producing molecules, compounds and so on. For an atom to be neutral, that is, to carry no net electric charge, the negative electrons on the outside must be balanced by an exact number of positive protons in the nucleus. Take helium as an example. A helium atom will have two orbiting electrons and two protons in its nucleus. However, the nucleus of an atom can also contain neutrons which have no electric charge. It is these differing numbers of neutrons in the nucleus of helium atoms which give rise to isotopes. He3 has two electrons in orbitals and two protons (to balance the electrons), and one neutron in its nucleus. The two protons plus one neutron make He3. He4 still has two electrons and two protons, but it also has two neutrons. Thus the two protons and two neutrons make for He4.
2) Helium is an atom which has two isotopes and two methods of production. Helium 3 (He3) is primordial. It was created at the time of the Big Bang. Helium 4 (He 4) is a wee bit heavier than Helium 3. It is not primordial. It is created as a by product of the radio active decay of uranium to lead.
Uranium itself comes from the explosion of stars. Like helium, uranium and lead have isotopes. No isotope of uranium is stable. Three isotopes of lead (206, 207 and 208) are stable. They do not decay. Lead 206 is produced from the decay of uranium 238. When uranium 238 decays it gives rise to other unstable elements (including unstable isotopes of lead) which then decay. This happens in a long series of steps, at the end of which is lead 206. Importantly, during this decay process, He4 is produced. To see all the isotopes of lead and uranium known so far, check out reference 5.
Thus we have:-
KABOOM ----> H2 (lots of) + He3 (some) (NB The Big Bang is not really an explosion)
U ----> Thorium ------> Other unstable isotopes through a series of steps ------> Pb206 + He4
At this stage you may not be able to see how isotopes can give clues to the history of rocks. However, if my essay is any good, then it will help you understand this bit about isotope ratios and rock history particularly with respect to helium.
3) Partial melting. Generally rocks are made of many different crystals and even bits of other rock. These various components will have differing melting temperatures. At increasing depths within the earths mantle, temperatures will be reached at which the various components of rocks will melt. This differential melting is called partial melting. By it, lighter elements (or those with lower melting points than the parent rock) can leave their parent rock and reach the surface as molten lava. The lava, on cooling, forms new rocks which have different element abundances and densities to those of the parent rock from which they were derived.
HERE WE GO
Imagine that both the upper and lower parts of the mantle started with the same amounts of helium 3 and uranium (U). Over time the U would give rise to the same amount of He4 in both the upper and lower mantle such that, for example only ,upper mantle (He4/He3) = lower mantle (He4/He3). Should any component of mantle rock get hot enough (because of depth of burial) and become magma, then it will be degassed of most of its helium on reaching the surface. (Like rising hot air, magma rises with respect to the colder denser rock around it). It will still contain its heavy uranium such that after cooling back into rock and over time, He4 will build up, leading to a higher ratio of He4/He3 relative to that rock which is still trapped within the mantle. If this continental rock gets recycled back into the mantle, and the upper and lower mantles are isolated and there is no mixing, then part of the mantle can, over time, show a signature of depleted He3. For pictures of this, see the third attachment.
Now for some more FACTs and some theory. On average, ocean island basalts (OIB) have a lower He4/He3 ratio that mid ocean ridge basalts (MORB). Ocean islands in this context are those islands like Hawaii which formed from the same kind of material which wells up from the mantle at the mid ocean ridges. However given this disparity in the helium isotope ratio, it was believed that the OIB came from the lower mantle where the initial helium incorporated at the earths origin has been retained, while the MORB came from the upper mantle where, due to escape into the earths crust followed by recycling back into the mantle, the He3 content had decreased, thereby giving rise to a higher He4/He3 ratio. This model therefore required that the two mantle layers be isolated such that helium disparities could not be smoothed out over time due to mixing or some other process.
However, over recent years, the concept of isolated upper and lower mantle reservoirs has been called into question. And it has been determined that helium and uranium/thorium may behave in melts differently to what had been thought before, with consequences for rock histories derived from this information. If this turns out to be the case that the previous (reasonable assumptions) are indeed incorrect then it kind of throws some models of continental crust formation into doubt. Most researchers agree that the continental crust comes mainly from the mantle and constraining parameters for associated models are now called into question, given that they relied on an incorrect model of the mantle and/or incorrect data relating to the behavior of isotopes.
[Aside: Reference (2) is an example of an experiment that calls old models into question. In this particular case, laboratory tests, using new experimental techniques, have been applied to natural and synthetic olivines (a mineral important in the formation of rocks). These have shown helium to be more soluble in molten rock, relative to uranium and thorium than previously thought.]
Parman argues that it is time for models relying on the evolution of helium isotopes to be considered with more seriousness. One reason for this plea is that the above mentioned experiments which throw the old models into doubt show that rather than MORB and OIB pointing to an isolation between the upper and lower mantles, they could instead be pointing to times of mantle melt depletion.
That is, the very FACTS that are calling the old models into doubt could be showing that the helium ratios of the OIB indicate the times at which the mantle underwent depletion of its molten component. Where did this mantle depletion go? Obviously upwards to make new continental crust. (And associated with this would be changes in the ratios of isotopes of various elements in this particular case, helium 4 and helium 3.)
So Parmans letter to Nature is his exploration of this idea, using the facts, Facts, FACTS at hand.
He notes that because OIB and MORB come from the mantle, they provide the primary constraints on the He isotropic composition of the mantle. Parman then goes on to discuss the results of measuring the helium isotope values in rocks from various volcanic islands and from mid ocean ridges around the world. Taking into account the amount of sampling done at various locations around the world which can affect both the strength of any isotope signal peaks (the height of a peak) or the range of signal peaks, he presents a series of graphs that demonstrate:-
1) on average MORBS have consistently higher He4/He3 ratios than OIBs and these MORB ratios have a dominating peak (in terms of probability distribution) at 90x10^3. That is, you are much more likely to find a MORB with 90,000 times more He4 than He3 than you are a ratio a bit less than this or a bit more than this. There is another MORB peak at 85x10^3 but it is the 90,000 peak plus four others that are important to this story.
2) the OIBs have helium isotope ratios ranging from 18x10^3 to 120x10^3 and peaking at roughly (20, 30, 36, 42, 50, 56, 64, 69,86 and 90)x10^3. For various reasons some OIBs show only a few peaks. The reasons appear to be because of limited sampling at particular islands more than anything else. (On islands with a high sampling number, the spread of isotope values is large and the peaks tend to be more distinct.)
3) where the OIB and MORB data overlap, the peaks still show up at matching places.
See the fourth attachment.
It is the He4/He3 ratio peaks at 30000, 42000, 56000, and 69000 which are important, for Parman demonstrates that those peaks can be shown to match ages from particular zircons which some researchers use to argue that the continental crust has grown in massive episodic spurts (reference 3). This is unlike other models in which the continents formed quickly and early and from then on continued to recycle themselves, neither growing nor shrinking and unlike other models in which the continents continually and steadily grew.
The problem is in dating the samples at which the helium isotope peaks occur.
One method of dating would be to return to all the islands, collect hundreds of samples and attempt to date them. Clearly this would be an impossibility very expensive in both time and money. (Parmans research was done from existing data sets.)
Another way would be to use a model for the evolution of helium within the earth as a basis for calculation. However, given the large number of models, it would be impossible to select any one model and use it to calculate a reliable age for these helium peaks.
A third way would be to see if there is some kind of correlation between the helium peaks and ages for the supposed melting events that caused the spurts in crustal growth.
Parman chose this latter course, noting that the zircon studies have shown the main peaks in zircon ages (and by implication bursts in continental growth) to be at 1.2, 1.9, 2.7 and 3.3 thousand million years ago (Gya). There is a complicating factor though, for as he points out, the peaks in occurrence of zircons at these ages could well be an indication, not of spurts in continental growth but rather of pieces of continental crust that have randomly escaped tectonic recycling. In tectonic recycling, continental crust with its zircons erodes and ends up on the ocean floor, where it gets placed back into the mantle at the ocean floor subduction zones. The zircons melt and have their radioactive clocks reset. Magma forming through partial melting rises to the surface and erupts. New zircons form but with much younger ages. If this is what really happened to produce the clusters in zircon ages, then a correlation with the helium isotope ratios should be most unlikely.
A point to begin correlating both sets of data is the fact that each set contains a distinctive drop in peak production. Thus, there are a large number of zircon peaks up till 1.2 Gya and relatively few after that. Similarly the He4/He3 peaks show a large number up to a ratio of 70x10^3 and relatively few after that. So Parman used the gap as a starting basis for correlating the two sets of data. He found the following:-
1) Given that both sets of data are monotonically increasing, other ways of aligning them exit. However, it is this one which uses the gap which shows the highest correlation with a correlation coefficient of 0.9986.
2) Importantly, the slope of the line is not random. If one projects the correlation line to the present day, it matches the value of the highest MORB of approx 90,000 for He4/He3. It is reasonable to suppose that the MORB trap into mantle which has undergone the most He3 depletion and so one should expect that the correlation line would intercept this value for time = 0 since mid ocean ridge basalts are very young. (These ridges are where the sea floor is continually being created.) And if one projects the line back 4.6 Gya (to the time of the earths formation) one gets a ratio of 5600 which is very close to the estimate for Jupiter of 6000 (see reference 4). And Jupiter, it is believed, retains the original solar system helium 3 content, simply because of its massive gravity. Helium can easily escape from the earth, but not Jupiter. (Given that these He4/He3 ratios are from natural systems, then other natural systems of helium isotopes should fall onto the same correlation line as that obtained from the OIB and zircon age data.)
The importance of these two endpoint numbers is that MORB record the mantle as it is now with its current helium 3 depletion. Jupiter records the original solar system helium 3 content and hence the isotope ratio the earth began with. That the experimental correlation line connects to these points indicates very strongly that the intervening peaks could well be recording mantle depletion events and associated spurts in the growth of continental crust.
See the fifth attachment.
Parman argues that it is the independent prediction of these two fundamental values by the helium-zircon correlation line, that indicates the plausibility of a casual relationship between the helium peaks and those of the zircons. As Parman notes, if the correlation was spurious, or the peaks artifacts from over-sampling, then the plots would result from three highly improbable coincidences a fluky correlation in peaks from island to island, the accidental matching of the zircon age pattern, and the fluky prediction of both the Jovian (initial Earth) and present day MORB values. The slope of the line appears to be robust in that it is not controlled by one point. If the four main helium peaks are used, then a line with an even stronger correlation coefficient is produced.
Two helium peaks do not have matching zircon peaks. Parman suggest that further studies may well reveal zircon peaks at these points.
So Parman concludes that this correlation is significant in that it adds weight to the idea that the continents did grow in spurts and that these spurts are recorded in both the mantle (from the OIB and MORB data since both kinds of rocks come from the mantle), and in the continental crust (in the zircons).
If this correlation is real, then Parman notes that it throws open many issues including:-
1) How and why did these massive building events occur, and what were they like?
2) How does the helium isotope signature remain in the various mantle domains, against mixing in the mantle due to the slow convection of rock occurring there?
3) The constraints on other models of helium evolution in the mantle and crust. Parman writes that there are numerous models of helium evolution (some relying on a separate upper and lower mantle that are isolated from each other, others relying on un-isolated mantle components etc.)
So has Parman uncovered strong evidence that bolsters the idea that the continents grew in a series of massive spurts over billions of years?
SUMMARY
New data on how molten rocks retain (or lose) their component elements of helium, uranium and thorium, have thrown into doubt some theories about how the continents grew.
Parman used this same data to argue that helium isotopes may provide a better method of modeling some aspects of continent growth.
Using helium isotope ratio data from various mid ocean ridges and many ocean islands that were made of basalts from magma which came from the mantle, Parman showed that there were several peaks in the ratios of helium isotopes. These peaks indicated that one was more likely to find rocks with these helium ratios than rocks with other ratios.
Given the way in which helium is formed and how its isotopes behave during the creation and cycling of rocks, Parman reasoned that this could tell him something about how the continental crust formed. He noted that these helium ratio peaks could be strongly correlated with evidence from zircons which indicated that continents may have grown from several massive inputs of lava from the mantle.
The associated correlation line predicted the initial helium isotope ratio of the earth and the current helium isotope ratio from mid ocean ridge basalts. This lent credence to the idea that the helium spikes were indeed indicative of episodes of mantle helium depletion and hence continental growth.
CONCLUSION
The argument from Parmans facts, Facts, FACTS looks pretty convincing, particularly when you look at the correlation, its robustness as well as the prediction of those two important values, the current MORB and the initial earth helium isotope ratios.
But nothing in science is simple.
So stay tuned. Better still, go out, do some learning and lots of reading. Then next time you spot an article related to this, you will not be able to ignore it, for you will have more to ponder. If these isotopes and zircons are telling us what some think, can you imagine what those times may have been like?
Regards, Roland
REFERENCES and NOTES
(1) S.W. Parman, Helium isotopic evidence for episodic mantle melting and crustal growth, Nature 446, 19 April 2007, pages 900-903.
(2) Stephen W. Parman, Mark D. Kurtz, Stanley R Hart & Timothy L. Grove, Helium solubility in olivine and its implications for high He3/He4 in ocean island basalts, Nature 437 20 October 2005, pages 1140 1143.
(3) A. I. S. Kemp, C. J. Hawkesworth, B.A. Paterson & P. D. Kinny, Episodic growth of the Gondwana supercontinent from hafnium and oxygen isotopes in zircon, Nature 439 2-February-2006, pages 580 583.
(4) The Nature article actually publishes values such as (92.2+/-0.7)x10^3 (+/-s.e.). I have written this as approximately 90000 for the sake of making things simpler.
(5) Isotopes of lead:-
http://ie.lbl.gov/education/parent/Pb_iso.htm
The half-life column tells whether the isotopes are stable or not. Seconds = s, milliseconds = ms, years = y and so on.
Isotopes of uranium:-
http://ie.lbl.gov/education/parent/U_iso.htm
