[ RadSafe ] Radioactive decay under your feet!!

Fri May 12 14:29:59 CDT 2006

>From Nature 437, 485-486 (22 September 2005)

On-line article at
http://www.nature.com/nature/journal/v437/n7058/full/437485a.html

Earth science: Unleaded high-performance
Tim Elliott1

Abstract: Previous measurements of uranium-series
isotopes have implied uncomfortably fast speeds of
melt movement through the mantle. Yet the latest
results suggest such velocities were serious
underestimates.

Most volcanism on Earth occurs unseen, at submarine
volcanic ridges that form in response to the sedate
spreading of the oceanic plates. As the plates pull
apart at a genteel rate of a few centimetres per year,
underlying mantle viscously rises at a similar rate to
fill the space. As a result of this upwelling and
decompression, the mantle melts, producing the magma
that feeds mid-ocean-ridge volcanoes.

On page 534 of this issue1, Rubin et al. provide a
dramatic illustration that magma rises to the surface
with unexpected haste, in stark contrast to the
stately movements of the solid from which it is
derived. Melt velocities of up to a few metres per
year, about 100 times faster than plate spreading
rates, have been rationalized from simple physical
models2. But Rubin and colleagues' measurements
suggest that melts beneath oceanic ridges may move up
to three orders of magnitude faster than that —
raising questions about our understanding of
permeability and fluid flow in the mantle3.

The authors present a geochemical study of very
recently erupted mid-ocean-ridge magmas. Investigation
of the rates of magma production and transport exploit
the uranium-series nuclides, which include isotopes of
thorium, radium, radon and lead. These short-lived
daughter nuclides occur in the decay chain between the
long-lived radioactive parent 238U (half-life 4.5109
years) and its ultimate, stable daughter, 206Pb. Given
time, these intermediate daughter nuclides will
establish a steady-state decay chain, in which all
nuclides decay at the same rate. This steady state is
termed secular equilibrium. Various processes, such as
mantle melting, may disturb secular equilibrium, but
equilibrium is re-established between any nuclide pair
in the uranium-series chain within around five
half-lives of the shorter-lived nuclide. Any
disequilibrium between a nuclide pair records an event
more recent than this. Within the uranium series,
nuclide half-lives range from 2.5105 years to 1.610-4
seconds, providing an ample choice of chronometer.

Compared to many geological processes, the timescales
documented by even the longer-lived uranium-series
nuclides, such as 230Th (half-life 76,000 years) and
226Ra (half-life 1,600 years), are rapid. Yet previous
studies of mid-ocean-ridge magmas had already revealed
disequilibrium in both 230Th–238U and 226Ra–230Th
nuclide pairs4, 5. Rubin et al.1 have upped the ante
and analysed 210Pb, which has a half life of only 23
years. After disturbance, the 210Pb–226Ra nuclide pair
will return to equilibrium on a timescale of about 100
years, dizzyingly fast for most processes in the
Earth's interior.

A major hurdle is to find samples from the seabed that
are so young that any 210Pb–226Ra disequilibrium
present at eruption has not significantly decayed. It
is troublesome to detect, let alone sample, eruptions
that occur some 2,500 metres beneath the ocean
surface, and so it is a remarkable achievement to
obtain lavas to test for initial 210Pb–226Ra
disequilibrium. The magmas erupted at mid-ocean ridges
also have notably low abundances of uranium and its
daughter nuclides, making accurate analysis
challenging.

Rubin et al.1 have overcome these problems, and make
the striking observation that many of their samples
have 210Pb–226Ra deficits — that is, less 210Pb than
would be expected relative to 226Ra at the steady
state, secular equilibrium. It is both reasonable and
conceptually appealing to attribute disequilibrium to
the very melting process that produces the magmas. Yet
it is also possible that contamination of magma in the
crust just before eruption, or degassing of the
volatile intermediate 222Rn (or even of 210Pb itself),
produces 210Pb deficits. Rubin et al., however,
present convincing arguments that these secondary
processes do not significantly influence 210Pb–226Ra
disequilibrium.

If the 210Pb–226Ra disequilibrium is then a result of
melting, rates of melting and melt movement to the sea
floor can be inferred. It is first necessary, however,
to assess what part of the melting process the
disequilibrium is timing. Disequilibrium records the
fractionation of parent from daughter nuclide. This
occurs during melting because elements have different
affinities for melt relative to the melting solid. For
example, 210Pb enters the melt less readily than
226Ra, giving rise to a melt with a 210Pb deficit. But
many of the uranium-series nuclides, including 226Ra
and 210Pb, favour the melt over the solid so strongly
that the differences in their behaviour are apparent
only when very small amounts of melt are present. In
simple models of melting, this means that the
production of 210Pb–226Ra disequilibrium can occur
only when melt is first produced (Fig. 1). The
presence of any disequilibrium in erupted lavas then
provides a constraint on the time taken for melt to
travel from the very bottom of the melting region to
the top.

[[[[[Figure 1: Melt pathways and possible sites for
generation of uranium-series nuclide disequilibrium.

Nuclides are fractionated at the onset of melting
because of their different affinity for melt relative
to the solid. The 210Pb–226Ra disequilibrium in magmas
measured by Rubin et al.1 potentially records this
event and so the transit time to eruption. Yet
continued equilibration between upwelling melt and
solid leads to different velocities for the nuclides
through the mantle7, generating further
disequilibrium. Generation of disequilibrium by such
an 'ingrowth' process is also only effective where the
proportion of the melt is small compared to the solid.
That is unlikely to be the case in the main melt
conduits, but tributaries to the main channels may
contribute ingrown nuclides even high in the melting
column. Finally, disequilibrium may be caused by
contamination or degassing in the crust, but Rubin et
al. make a good case against this.

High resolution image and legend (71K)]]]]]]

The previous speed limit for this process was clocked
in 1988 — also by Rubin, who, together with J. D.
Macdougall5, used the 226Ra–230Th pair, which returns
to equilibrium in about 8,000 years. The time
constraints of this earlier study were thus some two
orders of magnitude less stringent than those of the
new observations, but at the time they came as a big
surprise. In response to the perceived difficulty of
moving melt so fast to the surface6, melting models
were developed that relieved some of the need for
speed7. However, the less glamorous but quite
plausible alternative of crustal contamination has
also continually raised its head (Fig. 1).

Importantly, the new study1 not only requires faster
melt transport than before but also provides evidence
against some of the increasingly sophisticated
scenarios of contamination8. The effects of
contamination on the 226Ra–230Th pair are strikingly
different from those on 210Pb–226Ra. Thus, models
constructed to explain previously observed 226Ra–230Th
excesses by contamination seem unlikely to be able to
account for the new 210Pb–226Ra deficits. On the other
hand, coupled 210Pb–226Ra deficits and 226Ra–230Th
excesses are expected for most melting processes.
Rubin et al. demonstrate that a simple model can
reasonably account for their observations.

Clearly, a more comprehensive exploration of the new
data using refined models9, 10 will follow. But now,
even more emphatically than before, it seems that you
can't keep a good melt down.

References
1Rubin, K. H., van der Zander, I., Smith, M. C. &
Bergmanis, E. C. Nature 437, 534–538 (2005).
Kelemen, P. B., Hirth, G., Shimizu, N., Spiegelman, M.
& Dick, H. Phil. Trans. R. Soc. Lond. A 355, 283–318
(1997). | Article |

2Phipps Morgan, J. & Holtzman, B. K. Geochem. Geophys.
Geosyst. 6, Q08002; doi:10.1029/2004GC000818 (2005). |
Article |

3Condomines, M., Morand, P. & Allègre, C. J. Earth
Planet. Sci. Lett. 55, 247–256 (1981). | Article |
Rubin, K. H. & Macdougall, J. D. Nature 335, 158–161
(1988). | Article |

4Faul, U. Nature 410, 920–923 (2001). | Article |
Spiegelman, M. & Elliott, T. Earth Planet. Sci. Lett.
118, 1–20 (1993). | Article |

5Saal, A. E. & van Orman, J. A. Geochem. Geophys.
Geosyst. 5, Q02008; doi:10.1029/2003GC000620 (2004). |
Article |

6Lundstrom, C. Phys. Earth Planet. Inter. 121, 189–204
(2000). | Article |

7Jull, M., Kelemen, P. B. & Sims, K. Geochim.
Cosmochim. Acta 66, 4133–4148 (2002). | Article |

+++++++++++++++++++
"People will be shocked to see how safe it is to live in New York City."
ANDREW KARMEN, a sociology professor at John Jay College of Criminal Justice, on murder trends in the city.

-- John
John Jacobus, MS
Certified Health Physicist
e-mail:  crispy_bird at yahoo.com

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 