[meteorite-list] Chips Off an Old Lava Flow (Lunar Meteorite Kalahari 009)

From: Ron Baalke <baalke_at_meteoritecentral.com>
Date: Wed, 19 Dec 2007 17:39:14 -0800 (PST)
Message-ID: <200712200139.RAA18996_at_zagami.jpl.nasa.gov>

http://www.psrd.hawaii.edu/Dec07/cryptomareSample.html

Chips Off an Old Lava Flow
Planetary Science Research Discoveries
December 19, 2007

--- Lunar meteorite Kalahari 009 contains fragments of basalt about 4.35
billion years old, a record-breaking old age for mare basalt.

Written by G. Jeffrey Taylor
Hawai'i Institute of Geophysics and Planetology

Photogeologic and remote sensing studies of the Moon show that many
light-colored, smooth areas in the highlands contain craters surrounded
by dark piles of excavated debris. The dark deposits resemble the dark
basalts that make up the lunar maria. They contain the same diagnostic
minerals (especially high-calcium pyroxene) and chemical compositions
(high iron oxide) as do mare basalts. The deposits formed when vast
amounts of material ejected during the formation of giant impact basins
covered pre-existing lava plains. Since the smooth plains are older than
the youngest impact basin (about 3.8 billion years old), the lavas must
have erupted before formation of the visible maria. In fact, they were
visible maria for a while eons ago, but were buried by ejecta when the
basins formed.

We have samples of these ancient mare basalts. They reside in breccias
collected from the lunar highlands. Age dating indicates that the chips
have ages of 3.9 billion years and older. The oldest dated mare basalt
in the Apollo collection is 4.23 billion years. Now Kentaro Terada
(Hiroshima University, Japan), Mahesh Anand (Open University, UK), Anna
Sokol and Addi Bischoff (Institute for Planetology, Muenster, Germany),
and Yuji Sano (The University of Tokyo, Japan) have determined the age
of pieces of an ancient lava flow in a lunar meteorite, Kalahari 009,
found in Botswana in 1999. The team dated this very low-titanium mare
basalt by using an ion microprobe to measure the isotopic composition of
lead and uranium in phosphate minerals. They found that the basalt
fragments in the rock have an age of about 4.35 (plus or minus 0.15)
billion years. This overlaps with the ages of chemically-distinct
igneous rocks from the highlands, indicating that diverse magmas were
being produced early in the history of the Moon.

Reference:

    * Terada, K., Anand, M., Sokol, A. K., Bischoff, A., and Sano, Y.
      (2007) Cryptomare Magmatism 4.35 Billion Years Ago Recorded in
      Lunar Meteorite Kalahari 009. Nature, v. 450, p. 849-853.

PSRDpresents: Chips Off an Old Lava Flow --Short Slide Summary
<http://www.psrd.hawaii.edu/Dec07/PSRD-cryptomareSample.ppt> (with accompanying notes).
------------------------------------------------------------------------

Visible and Hidden Lava Plains

>From the moment Galileo peered at the Moon through his homemade
telescope, he recognized two main areas on the Moon (see photograph):
the rugged, light-colored highlands
(which he named "terra") and the smoother, darker areas ("maria").
Maria is the Latin word for sea, which Galileo
figured them might be. We did not know for sure what they were until
Apollo 11 astronauts retrieved samples from Mare Tranquillitatis. They
are basalts--ancient lava flows that flooded low areas, many the
interiors of large impact basins.

Analyses of samples and remote sensing measurements show that the maria
are dark because the lavas contain a lot more FeO (iron oxide) than do
highland rocks. FeO inside a mineral such as pyroxene makes it darker.
In addition, the mare basalts contain less plagioclase feldspar,
a light-colored mineral. Hence, the maria are dark. They are smooth in
part because they are so much younger
than the highlands and so did not accumulate as many craters. However,
they are also smoother because lava flows fill up low areas, tending to
produce smooth plains.

Many areas in the highlands are smooth plains, but they are light
colored, hence low in FeO. Remote sensing shows that they are not
composed of mare basalts. However, many light plains deposits are
decorated with impact craters surrounded by dark piles of ejecta,
nicknamed "dark-haloed craters." These curious features were debated for
years. Finally, Pete Schultz (now at Brown University) and Paul Spudis
(now at the Lunar and Planetary Institute, Houston) assembled all the
available evidence to make a good case that the dark-haloed craters
formed when mare basalt lava flows were covered with ejecta from large
impact craters and basins, and then small craters punctured through the
ejecta to toss out mare basalt. Detailed studies during the 1980s by B.
Ray Hawke and Jeff Bell (University of Hawaii), and investigators
elsewhere, provided further evidence that many light plains in the
highlands are underlain by dark basaltic rock. In 1992, Jim Head (Brown
University) and Lionel Wilson (Lancaster University, UK) named these
widespread deposits "cryptomaria, meaning hidden maria.

Clementine images of dark-haloed craters on the Moon

LEFT: Seventeen dark-haloed craters are indicated by numbers on this
image mosaic from Clementine 750 nm remote sensing data of the
Lomonosov-Fleming region. TOP RIGHT: Clementine 750 nm image of crater
#7. This crater is 8 kilometers in diameter. The arrows show where
additional remote sensing spectra were obtained. BOTTOM RIGHT: FeO map
of crater #7.

Drawings showing the formation of cryptomaria and dark-haloed craters on
the Moon

The four panels in this diagram show cutaway, side views into the Moon
to describe the basic formation of a cryptomare. Cryptomaria formed when
low -lying areas in the heavily -cratered ancient highlands (TOP panel)
were flooded by mare basalt lava flows (black deposit, SECOND panel). A
subsequent basin-forming impact, hundreds to thousands of kilometers
away, deposited ejecta on top of the surface (THIRD panel). During
deposition, some of the mare basalt was mixed with ejecta from the
basin, forming a deposit consisting of broken and mixed highland rocks
with fragments of mare basalt (grey deposit). Eventually, smaller
craters punctured through the ejecta, (BOTTOM panel) depositing debris
from the underlying mare basalts around them, creating dark -haloed
craters and revealing the existence of the buried mare basalt. The
bottom panel shows two dark -haloed craters on the left side of the
panel. The small crater on the right side does not have a dark halo.

------------------------------------------------------------------------

Ancient Samples

While geologic studies of dark haloed craters were leading to the idea
of mare-like volcanism that pre-dates the visible maria, cosmochemists
were getting hints of ancient mare volcanism from samples of the lunar
highlands. One of the first studies (in 1976) was by Graham Ryder (then
at the Smithsonian Astrophysical Observatory) and me (then at Washington
University in St. Louis). While studying impact breccias formed in
ancient, pre-mare events, we noticed small rock fragments that looked
just like mare basalts (high FeO, small amount of plagioclase feldspar).
The way the minerals were intergrown and their compositions indicated
that the rock fragments were volcanic. We proposed that they were pieces
of mare-like lava flows that formed before the visible maria. Aside from
some debate (which Graham and I referred to as whining) about whether it
was proper to call them "mare" basalt when they did not come from a
visible maria, the work was well received, and promptly ignored.

Part of the reason for ignoring the idea of mare-like volcanism before
3.8 billion years ago was that individual basalts had not been dated.
That problem was solved when the age daters started to extract samples
from highland breccias and date them by a variety of isotopic
techniques. Sure enough, the ages started coming in older than 3.8
billion years. Many were 3.9 to 3.95 billion years, and one, dated by
Larry Taylor (University of Tennessee) and his colleagues was 4.23
billion years old. The rock was mineralogically and compositionally a
mare basalt. It was the oldest mare basalt (or pre-mare mare basalt!)
dated...until now.

Photomicrograph of Apollo sample 14305
Photomicrograph in cross polarized light of the mare basalt fragment in
lunar highlands breccia 14305, dated by Lawrence Taylor and his
colleagues. The view is dominated by pyroxene, olivine, and chromium and
titanium oxides. The rock's age is 4.23 billion years.

------------------------------------------------------------------------

The Oldest Mare Basalt (so far)

Photo of lunar meteorite Kalahari 009. Click for more information.
<http://meteorites.wustl.edu/lunar/stones/kalahari008.htm> Kalahari 009
was found in September 1999 near the small village of Kuke in the
Kalahari Desert, Botswana. It was first described in a paper by Anna
Sokol and Addi Bischoff (Institute for Planetology, Muenster, Germany).
Sokol and Bischoff show that the rock is composed of fragments of lunar
mare basalt and minerals derived from the basalt. Their data indicate
that the rock is a very-low-Ti mare basalt. An important observation is
that the minerals making up the rock show clear evidence for
high-pressure shock, such as feldspar converted partly to glass (called
maskelynite). Thus, the rock's Ar-Ar age might have been partially or
completely reset by an impact event.

images of Kalahari 009

LEFT: Photomicrograph of basalt fragment in Kalahari 009, taken in cross
polarized light. Medium to bright gray is plagioclase feldspar, while
blue and darkest gray is pyroxene. The texture is typical of
crystallized lava, but the minerals have been damaged by shock caused by
a meteorite impact on the Moon. RIGHT: Backscattered electron image of
the same area as in the photomicrograph. Brightest mineral is pyroxene,
dark is plagioclase. The vein along the lower left is calcite (calcium
carbonate), a terrestrial contaminant.

The age of Kalahari 009 was first determined by Vera Fernandes
(University of California, Berkeley), using the argon-argon technique.
This is the modern way geochronologists determine the potassium-argon
age for a rock (radioactive potassium-40 decays to argon-40). Its
advantage is that it gives information about gas loss caused by a
reheating event after initial crystallization of a rock. Fernandes'
analyses indicated a well-defined age of 1.70 ?? 0.04 billion years. An
important question is whether that age represented the time the basalt
formed, or a subsequent event that produced the shock effects observed
in the rock.

To get a better idea of the rock's age, Kentaro Terada and his
colleagues concentrated on the phosphate minerals (apatite and
merrillite) that occur between the abundant plagioclase and pyroxene
crystals. Phosphates are the main repositories of uranium in lunar
basalts, making them efficient recorders of uranium-lead ages.
(Uranium-238 decays to lead-206, uranium-235 decays to lead-207. In
general, the higher the uranium concentration, the higher the precision
of the ages determined.) Phosphates also have the great virtue of being
relatively resistant to lead redistribution by heating, so are less
likely to have been affected by the shock event the Kalahari 009 basalt
experienced.

phosphate grains in Kalahari 009

These are scanning electron microscope images of two of the phosphate
grains (labeled phos 5 and phos 3) used for determining the age of
basalt in Kalahari 009. The crystals are too small to separate
efficiently, so their uranium and lead isotopes were measured using an
ion microprobe at the Hiroshima University. Fe-px is iron pyroxene, fa
is fayalite, and ilm is ilmenite.

The phosphate crystals are not abundant in the rock, and they are too
small to separate physically, so Kentaro Terada used an ion microprobe
called SHRIMP to measure the isotopic abundances of uranium and lead.
(For more information about ion microprobe analysis, see
PSRD-Instruments of Cosmochemistry article: Ion Microprobe
<http://www.psrd.hawaii.edu/Feb06/PSRD-ion_microprobe.html>.) The
laboratory at the University
of Hiroshima is experienced in making such measurements on a range of
terrestrial and extraterrestrial materials, including accurately dating
shark teeth hundreds of millions of years old.

The results can be plotted on graphs in a number of ways that
geochronologists understand well, but baffle those of us who do not use
them regularly. The diagrams are complicated because there are two
parent isotopes (uranium-235 and uranium-238) and two daughter isotopes
(lead-206 and lead-207), plus the isotope lead-204 that does not form by
decay. Cosmochemists use plots of different isotope ratios to estimate
an age, but often, as in this case, the ages differ somewhat (see
diagram below). To plot all the isotopes at once you need a
three-dimensional diagram, which is what Terada and his colleagues used
to make their best estimate of the age of the basalt in Kalahari 009.
The diagram is not really three dimensional. Instead, it is a projection
from three dimensions onto a two-dimensional plot [more information
<moreInfo.html>]. This results in a best estimate of 4.35 ?? 0.15 billion
years for the age of the basalt in Kalahari 009. Maybe not the most
precise age ever determined, but it clearly shows that the rock is
ancient and almost certainly older than the basalt fragment dated by
Larry Taylor and his colleagues.

isotope rations of phosphate grains in Kalahari 009

Graphs of isotope ratios in phosphate minerals measured by ion
microprobe in a basalt fragment in Kalahari 009. The relatively
well-defined lines (called isochrons) in graphs a and b yield ages for a
particular set of isotopes. The diagram on the right (c) shows a
projection from three-dimensions, so uses all isotopic data, and gives
an age of 4.35 ?? 0.15 billion years, which Terada and his colleagues
suggest is the best estimate of the age of the lava flows represented in
Kalahari 009. The uncertainty in the age is based on the uncertainties
in the individual measurements (see error bars) and is at the 95%
confidence level. (MSWD is the mean square weighted deviate, a measure
of the scatter in the data.)

We now know that mare-like basalts erupted very early in lunar history
and that lunar cryptomaria range in age from 4.35 to about 3.9 billion
years. The cryptomaria and lunar impact breccias returned by Apollo and
found as lunar meteorites record an extensive record of pre-mare
mare-like volcanism. The early start to lunar volcanism, as noted by
Terada and co-workers and by Larry Taylor and his co-workers, indicates
that many types of magmas were produced early in the Moon's history.

------------------------------------------------------------------------

Magma Magma Everywhere

In broad terms there are three groups of lunar rocks: (1) Anorthosite
(nicknamed FAS), which formed by floatation in an ocean of magma
surrounding the Moon soon after it formed. Determining their ages is
tricky, but they likely formed by 4.46 billion years ago (see PSRD
article: The Oldest Moon Rocks <http://www.psrd.hawaii.edu/Dec07/lunarAnorthosites.html>). (2)
Highlands magnesian-suite (Mg-suite), which formed soon after the FAS
crust formed and continued until about 3.9 billion years ago. It is
composed of several rock types, including intrusive igneous rocks (norites
and troctolites) and lava flows called KREEP basalts. (3) Mare basalts, which we
now know began to form by 4.35 billion years ago and continued for over 2
billion years (on the basis of crater counts, the youngest are possibly
as young as 1 billion years old).

Photos of Apollo samples of anorthosite, Mg-suite rock troctolite, and
mare basalt.

For a long time we had a comfortable, somewhat simple picture of magma
formation in the Moon. The FAS suite formed first, followed by assorted
Mg-suite magmas, which was followed by mare basalts. Simple. Nice crisp
divisions. One thing, then another. Early indications of mare basalts
older than 3.9 billion years showed that mare and Mg-suite volcanism
overlapped, but the overlap was concentrated near the end of Mg-suite
volcanism. Now, with two mare basalt samples substantially older than
3.9 billion years, the picture is more complicated. Mare basalt
volcanism continued after Mg-suite magmas were no longer being produced,
but there was a time when both were operating robustly.

The interesting thing about such compositionally-diverse magmatism
operating at the same time is that the two suites of magma must have
formed in compositionally-diverse regions inside the Moon. These
distinctive regions located deep inside the Moon had to partially melt
to produce magma. The melting could have been triggered by a variety of
mechanisms. The most common might have been the rise of plumes of solid
mantle rock triggered by density differences caused by the way the magma
ocean crystallized. When magma solidifies, the first crystals of olivine
and pyroxene to form have high ratios of magnesium to iron, and end up
forming a pile at the bottom of the magma pool. The ratio decreases as
crystallization proceeds. Because density increases as magnesium/iron
decreases, the final column has denser (hence heavier) rock on top and
less dense rock below. This is unstable, so it overturns. The solid base
rises and, as it does so, the lower pressure allows the rock to
partially melt, producing magma.

Procellarum KREEP Terrane is characterized by high Th

Procellarum KREEP Terrane is characterized by high thorium (and the
other heat producing elements potassium and uranium). Heating by
radioactive decay might have caused melting and overturn of the mantle,
producing widespread mare basaltic volcanism.

The second mechanism is simply heating of portions of the interior by
the decay of radioactive elements. This happens over a long time,
accounting for the long duration of mare basalt magmatism, and is aided
by high concentrations of radioactive elements (potassium, uranium, and
thorium). Those elements and others like them (called large-ion
lithophile elements) concentrate in the last dregs of the magma ocean,
making up a material called KREEP. Terada and his colleagues suggest
that high concentrations of radioactive elements in the mantle could not
be the source of the mare basalt in Kalahari 009 because it contains
only small concentrations of large-ion lithophile elements. However,
Mark Wieczorek (now at the Institut de Physique du Globe, Paris) and
colleagues at Washington University in St. Louis have shown that an
upper mantle and crustal concentration of radioactive elements in the
Procellarum-KREEP Terrane (see PSRD article: A New Moon for the
Twenty-First Century <http://www.psrd.hawaii.edu/Aug00/newMoon.html>)
sent a heat pulse into the
mantle, causing substantial overturn and basalt production, including
melting of regions of the mantle low in radioactive and other large-ion
lithophile elements.

Terada and colleagues point out an intriguing alternative hypothesized
by Linda Elkins-Tanton, Brad Hagar, and Tim Grove (Massachusetts
Institute of Technology). They suggest that basin-forming impacts could
have triggered substantial amounts of magmatism, some immediately, some
years to 10,000 years after a huge impact, and more up to 350 million
years after an impact. The instantaneous melting happens because of
rapid decrease in the pressure at depth because of the removal of a
large mass of rock. Their calculations indicate that subsequent
upwelling of the upper mantle could cause convection to take place for
up to 350 million years, perhaps producing ancient cryptomaria. Terada
and his team suggest that the 4.35-billion-year-old basalt represented
by Kalahari 009 could have formed as the result of an impact of
approximately that age.

If this fascinating interpretation is correct (and it is far from
proven), it suggests that not all large impacts occurred during a narrow
interval from 3.95 to 3.85 billion years ago, the time of the
hypothetical lunar cataclysm. The idea of the cataclysm is that there
was a dramatic spike in the impact rate about 3.9 billion years ago,
lasting only 100 million years. (See PSRD articles: Lunar Meteorites and
the Lunar Cataclysm <http://www.psrd.hawaii.edu//Jan01/lunarCataclysm.html>,
Wandering Gas Giants and Lunar Bombardment
<http://www.psrd.hawaii.edu//Aug06/cataclysmDynamics.html>, Uranus,
Neptune, and the Mountains of the Moon
<http://www.psrd.hawaii.edu/Aug01/bombardment.html>.) While not a
clear-cut case with only two very old basalt samples, it shows that
ancient cryptomaria would be worthy targets of sample return missions
from which we would learn more about not only when mare basaltic
magmatism began, but perhaps about the bombardment history of the Moon
as well.

------------------------------------------------------------------------

Samples and Remote Sensing

The research on cryptomaria shows how remote sensing, geologic studies,
and sample analysis combine to allow us to understand the volcanic
history of the Moon. Remote sensing gives us an overview and geologic
context. Sample analysis gives us detailed rock textures, mineralogy and
chemical composition, and precise ages. In the future, continued remote
sensing studies will reveal ideal targets for sample return missions by
robots and people, leading to an improved understanding of the Moon's
geologic history for comparison with Earth, Mars, and other rocky
objects in the inner solar system.

Sample analyses and remote sensing are all necessary to understand the
geologic history of the Moon.
Sample analyses and remote sensing are all necessary to understand the
geologic history of the Moon.

------------------------------------------------------------------------

ADDITIONAL RESOURCES

    * PSRDpresents: Chips Off an Old Lava Flow--Short Slide Summary
      <http://www.psrd.hawaii.edu/Dec07/PSRD-cryptomareSample.ppt> (with accompanying notes).

    * Cohen, B. A. (2001) Lunar Meteorites and the Lunar Cataclysm.
      Planetary Science Research Discoveries.
      http://www.psrd.hawaii.edu/Jan01/lunarCataclysm.html
    * Elkins-Tanton, L. T., Hager, B. H., and Grove, T. L. (2004)
      Magmatic Effects of the Lunar Late Heavy Bombardment. Earth
      Planet. Sci. Lett., v. 222, p. 17-27.
    * Giguere, T. A., Hawke, B. R., Blewett, D. T., Bussey, D. B. J.,
      Lucey, P. G., Smith, G. A., Spudis, P. D., and Taylor, G. J.
      (2003) Remote Sensing Studies of the Lomonosov-Fleming Region of
      the Moon. Journal of Geophysical Research, v. 108,
      doi:10.1029/2003JE002069.
    * Jolliff, B. L., Gillis, J. J., Haskin, L. A., Korotev, R. L., and
      Wieczorek, M. A. (2000) Major Lunar Crustal Terranes: Surface
      Expressions and Crust-Mantle Origins. Journal of Geophysical
      Research, vol. 105, p. 4197-4216.
    * Kaguya (Selene) Mission to the Moon Homepage
      <http://www.selene.jaxa.jp/index_e.htm>, from Japan Aerospace
      Exploration Agency (JAXA).
    * Lunar Meteorites
      <http://meteorites.wustl.edu/lunar/moon_meteorites.htm>,
      comprehensive site from Randy Korotev, Washington University in
      St. Louis.
    * Norman, M. (2004) The Oldest Moon Rocks. Planetary Science
      Research Discoveries.
      http://www.psrd.hawaii.edu/April04/lunarAnorthosites.html
      <../April04/lunarAnorthosites.html>.
    * Ryder, G. and Taylor, G. J. (1976) Did Mare-Type Volcanism
      Commence Early in Lunar History? Proc. Lunar Sci. Conf. 7th, p.
      1741-1755.
    * Sokol, A. and Bischoff, A. (2005) Meteorites from Botswana.
      Meteoritics and Planetary Science, v. 40, p. A177-A184.
    * Taylor, G. J. (2000) A New Moon for the Twenty-First Century.
      Planetary Science Research Discoveries.
      http://www.psrd.hawaii.edu/Aug00/newMoon.html <../Aug00/newMoon.html>
    * Taylor, G. J. (2001) Uranus, Neptune, and the Mountains of the
      Moon. Planetary Science Research Discoveries.
      http://www.psrd.hawaii.edu/Aug01/bombardment.html
      <../Aug01/bombardment.html>
    * Taylor, G. J. (2006) Wandering Gas Giants and Lunar Bombardment.
      Planetary Science Research Discoveries.
      http://www.psrd.hawaii.edu/Aug06/cataclysmDynamics.html
      <../Aug06/cataclysmDynamics.html>.
    * Taylor, L. A., Shervais, J. W., Hunter, R. H., Shih, C.-Y.,
      Bansal, B. M., Wooden, J., Nyquist, L. E., and Laul, L. C. (1983)
      Pre-4.2 AE Mare-Basalt Volcanism in the Lunar Highlands. Earth and
      Planetary Science Letters, v. 66, p. 33-47.
    * Terada, K., Anand, M., Sokol, A. K., Bischoff, A., and Sano, Y.
      (2007) Cryptomare Magmatism 4.35 Billion Years Ago Recorded in
      Lunar Meteorite Kalahari 009. Nature, v. 450, p. 849-853.
    * Wieczorek, M. A. and Phillips, R. J. (2000) The "Procellarum KREEP
      Terrane": Implications for Mare Volcanism and Lunar Evolution.
      Journal of Geophysical Research, v. 105, p. 20,417-20,430.
Received on Wed 19 Dec 2007 08:39:14 PM PST


Help support this free mailing list:



StumbleUpon
del.icio.us
reddit
Yahoo MyWeb