Transcript Continental Formation - Department of Geosciences

Making the Earth’s crust - The
generation and destruction of
crustal masses through time
Mihai Ducea
University of Arizona,
Department of Geosciences
Tucson, AZ, 85721
Outline
•
•
•
•
•
•
The paradox of continental formation
Making basalts, the oceanic crust
Plate tectonics, now and in the geologic past;
Continental arcs- the possible solution
The need for a recycling mechanism
Future challenges
Continental crustal paradox
• Continental crust is being generated by
several melting stages of the Earth;s mantle;
it is essentially “distilled” from the mantle
• All mantle melts are basaltic
• Continents are granitic on average
• WE DON’T QUITE UNDERSTAND HOW
CONTINENTS FORM(ED)
Granites, granodiorites, tonalites= continental crust
Basalt and gabbro density-3 g/cm3
Granite-2.7 g/cm3
Oceanic lithosphere- different from continental:
Basaltic crust, relatively thin (5-7 km), leads to
topographic lows
Heat engine- very efficient
Earth differentiation- primarily by magmatism
Mantle
convectionMostly solid
state
Melting shallow
by adiabatic
decompression
Lithosphere- the
cold
lid at the top
Chondrites- bulk Earth
Iron meteorites=core-like
Mantle xenolithsPERIDOTITES
CONVECTION
T - scale ~ plate motions
Length scales - appear much more complicated than
the ridge-trench systems
Convection cells penetrate to the coremantle boundary
Making the oceanic crust
Young------- Old
The oceanic crust
• Young (< 200 Ma)
• Forms at mid-ocean ridges, the oceanic
crust and lithosphere cools away from ridge
• Made entirely of basalt - which is what we
expect to form by partial melting of the
mantle
• All other solar system “crusts” are basaltic
In addition- experiments carried out under
any possible conditions of mantle melting
have demonstrated unambiguously that
basaltic melts should be what we get out of
the mantle.
Basaltic melts are ~ 50 SiO2, granites are
65% SiO2. If anything (e.g. higher degrees of
melting) we’d expect lower than 50% SiO2
melts.
It is instructive to check if this
applies to another product of
mantle melting - “mantle plumes”
or “hot spots” like Hawaii. The
answer: YES, they’re also made of
basalts.
What happens at subduction
zones?
• Oceanic plates are young; older oceanic
crust has been subducted
• Obviously it had to be recycled- otherwise
the Earth should be increasing its volume
• At subduction zones, magmatism tends to
follow the subduction margin - the resulting
product is a magmatic “arc”
Island arcs - classic
example- the Aleutians arc
On average - the
composition is also basaltic
But then there are the continents
• When did they form?
• How did they form?
• Answers: we think they form gradually over
the 4.5 Ga history of the Earth
• We also think they form by remelting
basalts
• That requires an additional “residual”
reservoir that we don’t see anywhere
Evidence for PT goes back to the Archean. Faster motions,
more melt, smaller continents (the continental nuclei known
as cratons or “croutons”)
Continents-succession of orogenic events
Ages of continental rocks
• Young ages are determined
biostratigraphically
• Older rocks are being determined
geochronologically
Zircons - as old as 40- 4.2 Ga; evidence for continental crust
Adding
mass to a
continent
Materials get accreted to the
continental margin and
become terranes, “exotic
terranes”; I.e. blocks of
unrelated origin that were
once far apart but got
assembled by accretion onto
a continental margin
E.g. the Pacific
Classic example: the North American Cordillera
But a factory that transforms
basalts into grantioids has to exist
• The only reasonable place- the continental
arcs
• If not- we simply don’t have a good enough
environment today (or in the recent
geologic past) that makes continents
• The key place where all may can be
explained: the Sierra Nevada of California
North American batholiths
Coast batholith
Idaho batholith
Sierra Nevada
batholith
Peninsular Ranges
batholith
5-10 km
Igneous Intrusions
100-1000000 km3
????
Typically shallow
exposures and no deep
“windows” into the
arc
cl
1

c0 D  F(1  D)
melt s
solids
Source region
Part ial melt ing
Melt
Residue
Magma chamber
Fract ional
cryst allizat ion
Melt
Cumulat e
1 part melt
requires 1 to 2
parts residue
Sierra granitic thickness
is 30-35 km, much
larger than previously
thought (8-10 km)
three deep
crustal
exposures (3035 km)
BIG SUR (Santa Lucia Mts)
Granulites (feldspar +garnet +pyroxenes)
Miocene
Pliocene
Quat ernary
Xenoliths
CSNB-Central Sierra Nevada batholith
Gar
Cpx
Minerals:
clinopyroxene
garnet
orthopyroxene
amphibole
rutile, apatite
Textures
1. Large grain sizes
2. Cumulate textures
=IGNEOUS textures
Key new observations
• The arc is much thicker than thought
• Requires about 1:1 or 2;1 residue to melt
ratio
• Much of the residue is not granulitic as all
textbooks argue, but it is “ECLOGITIC”
ECLOGITE = garnet + pyroxene
Garnet- much denser than any other crustal mineral
Arc root density vs. temperature
Mafic lower crust
Magmatism intimately linked
to thursting of the lithosphere:
magmatic flare-ups follow
thrusting after 15-25 my
Convective removal of roots
• Don’t see thick roots in N-American arcs.
They ought to be removed somehow.
• Most likely process- foundering, or
convective removal.
• Buoyancy-driven process that appears to be
triggered only in continental arcs and only if
the arc column is thick enough to generate
eclogitic residues.
Tectonic MOHO
Arc keels- do they founder?
0
4 g/cm 3
0 km
Granitoids
=2.6-2.7 g/cm3
remnant crust
vp=6-6.3 km/s
Granulites
50 km
Eclogites
? ??
=3.5-3.6 g/cm3
vp=8.4 km/s
Mixed zone
??
100 km
Peridotites
?
150km
?
Moho
=3.25-3.3 g/cm3
vp=7.7-7.9 km/s
foundering
upwelling
Eclogite: convectively unstable, will founder?
In which tectonic environments
Working hypothesis for
convective removal of root
• Should take place in thick arcs (25-30 km
thick batholiths) that leave behind an
eclogite facies residue
• Should not take place when most residue is
in granulite facies;
• Follows periods of flare-ups and might be
triggered by extension
So what do we know?
• In the Sierras, we are confident that pre-existing
crustal materials have been involved in a second
stage melting process that made the granitic
“batholith”
• We know now that some 35 km of the crust is
granitic
• Since at least that much residue is required, and it
had to be located at 35-70 (or more) km depth,
much of it had to be dense “eclogitic” residue
• We see these eclogites in 12 Ma volcanic rocks (as
xenoliths)- we have direct evidence that they
existed!
• They are currently lacking beneath the Sierras;
there is strong evidence that this root has
foundered (sank!) in the less dense mantle
• If true, we had found an elegant mechanism to
making continental crust
Key issues to remember
• Oceanic crust is mafic/basaltic and young older oceanic materials are subducted
• Continental crust has the average of a
granitoid, appears unsubductable
• Continental masses grow over time, some
continents have older cratons in their cores
separated by mobile belts
More
• Making continents is difficult to explain
unless a “residual” mass equivalent to that
of the continents is present under the
continents;
• We think that these residues have detached
and sank in the mantle