![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
kaberettF SARA RUSSELL -- THE COMPOSITION OF THE LUNAR MAGMA OCEAN - EVIDENCE FOR HETEROGENEITY?
or: Was there a lunar magma ocean at all?
Magma ocean concept:
- First proposed by Wood et al (1970): A lunar structural model is proposed in which a 25km anorthositic crust, produced by magmatic fractionation, floats on a denser gabbro" - visionary
How much can we trust Apollo rocks?
- it's cool, we've got lunar meteorites! 164 named meteorites, 84% found since year 2000
- which *may* represent a random sampling of the lunar crust
- lunar meteorites are typically highly impact-processed breccias but may contain relict grains that have been unaltered since it crystallised as an igneous rock
- relict anorthites may represent the original composition of anorthite crystallised from the LMO
- can be recognised by large crystal sizes, good crystalliniy, fracturing, lack of metal
- ooh, been exploring using "CL" (cathodoluminescence?) vs BSE - get much clearer images of plag clasts + reaction rims
-- CL can also give an indication of how pristine a clast is - where it's been melted, etc the clast gets brighter (this is a technique the NHM are developing)
Lunar highlands rock suites:
- often only get relict anorthite grains by themselves - no coexisting mafic minerals can be sure are cogenetic
-- difficult to make a Warren diagram!
- plag chemistry in its own right can also be an indicator of magmatic history - they say!
- earlier work on this used Mg#: however, both Fe and Mg are incompatible in plag, and partitioning is v v complex (melting, post-crystallisation metamorphic processes, etc)
- Raaaaaaaare Earth Elements. Trivalent REE can also be effected by impat shock effects and mtamorphic intermineral subsolidus equilibration, where they partition into co-existing mafic impact melt or px
-- however, metamorphic equilibration effect is thought to be far less significant in lunar anorthosites compared to terrestrial ones, and can be monitored in lunar plag by assessing chondrite-normalised LREE to HREE slopes in pristine unequilibriated plagioclase
-- Eu and Sm are useful diagnostics of the main Apollo rock suits
Lunar meteorite plagioclase:
- major trace element data reported from rock & ineral fragment
- and data also from various other bits and bobs...
Plagioclase REE in lunar rocks:
- normalised to chondrite. HUGE Europium spike - because it can exchange for calcium in plag structure, so sucks up any Eu in melt
- new feldspathic lunar meteorite REE data looks quite similar to FANS
- plag in different Apollo rock types can be distinguished by comparing EU-anomaly (chondrite-normalised Eu/Sm) to Sm values (proxy for LREE)
- Apollo FAN trend distinguished from Apollo mafic ferroan FAN sub-group: suggests ptrological variability within anorthositic plag
- lunar meteorite plag define different trend - comparatively lower Eu-anomalies for REE abundance
- suggests further petrological range in lunar meteorites compared with Apollo samples - LMO models have to account for this range (primary [FAN] vs secondary [KREEP-bearing] crust)
--> heterogeneity in source rocks...
[oh wow she has not timed this talk well]
Other evidence for lack of a homogeneous source for anorthosites:
- Gross, Treiman and Mercer - "Sinking the magma ocean: new evidence from lunar meteorites and the return of serial magmatism" - LPSC 2012
- Sm-Nd: differences in Nd-isotopics in A16 FAN and LM anorthosites; some samples from LREE enriched sources and some from HREE enriched sources, suggests 'lunar anorthosites do not derive from a single source, i.e., they are not all products of the LMO' (Nyquist et al 2010) - okay, I am so unconvinced, how does this play with e.g. the mantle zoo? what kind of heterogeneities do we see in Sept-Iles (i.e. across a much smaller volume/surface)? Flotation from multiple cycles of large-scale heterogeneities (does the silica content remain low enough for long enough)?
- ages of lunar anorthosites: variable, ish, with huge error bars
Implications:
- perhaps see any rocks from LMO, but instead see rocks taht formed from more complex geological processing, for example:
-- crust formed quickly, which was then processed by a hot, partially melted mantle; anorthositic blebs rose to the surface by serial magmatism; lots of diapirism
-- preserves initial heterogeneity, causes wide range in trace and major element chemistry and gives rise to a range of ages for anorthosite
- not arguing that LMO NEVER existed, just that we don't see ROCKS from it (so far) (see e.g. Longhi & Aswal, 1985)
Discussion:
- what would ratio of initial crust to later magmatism be? -- she reckons we see very little of the very earliest crust
- Moon asymmetry - far side mafics more magnesian?
- primary claim that anorthosite did not predominantly form within 1Ma of morn formation
or: Was there a lunar magma ocean at all?
Magma ocean concept:
- First proposed by Wood et al (1970): A lunar structural model is proposed in which a 25km anorthositic crust, produced by magmatic fractionation, floats on a denser gabbro" - visionary
How much can we trust Apollo rocks?
- it's cool, we've got lunar meteorites! 164 named meteorites, 84% found since year 2000
- which *may* represent a random sampling of the lunar crust
- lunar meteorites are typically highly impact-processed breccias but may contain relict grains that have been unaltered since it crystallised as an igneous rock
- relict anorthites may represent the original composition of anorthite crystallised from the LMO
- can be recognised by large crystal sizes, good crystalliniy, fracturing, lack of metal
- ooh, been exploring using "CL" (cathodoluminescence?) vs BSE - get much clearer images of plag clasts + reaction rims
-- CL can also give an indication of how pristine a clast is - where it's been melted, etc the clast gets brighter (this is a technique the NHM are developing)
Lunar highlands rock suites:
- often only get relict anorthite grains by themselves - no coexisting mafic minerals can be sure are cogenetic
-- difficult to make a Warren diagram!
- plag chemistry in its own right can also be an indicator of magmatic history - they say!
- earlier work on this used Mg#: however, both Fe and Mg are incompatible in plag, and partitioning is v v complex (melting, post-crystallisation metamorphic processes, etc)
- Raaaaaaaare Earth Elements. Trivalent REE can also be effected by impat shock effects and mtamorphic intermineral subsolidus equilibration, where they partition into co-existing mafic impact melt or px
-- however, metamorphic equilibration effect is thought to be far less significant in lunar anorthosites compared to terrestrial ones, and can be monitored in lunar plag by assessing chondrite-normalised LREE to HREE slopes in pristine unequilibriated plagioclase
-- Eu and Sm are useful diagnostics of the main Apollo rock suits
Lunar meteorite plagioclase:
- major trace element data reported from rock & ineral fragment
- and data also from various other bits and bobs...
Plagioclase REE in lunar rocks:
- normalised to chondrite. HUGE Europium spike - because it can exchange for calcium in plag structure, so sucks up any Eu in melt
- new feldspathic lunar meteorite REE data looks quite similar to FANS
- plag in different Apollo rock types can be distinguished by comparing EU-anomaly (chondrite-normalised Eu/Sm) to Sm values (proxy for LREE)
- Apollo FAN trend distinguished from Apollo mafic ferroan FAN sub-group: suggests ptrological variability within anorthositic plag
- lunar meteorite plag define different trend - comparatively lower Eu-anomalies for REE abundance
- suggests further petrological range in lunar meteorites compared with Apollo samples - LMO models have to account for this range (primary [FAN] vs secondary [KREEP-bearing] crust)
--> heterogeneity in source rocks...
[oh wow she has not timed this talk well]
Other evidence for lack of a homogeneous source for anorthosites:
- Gross, Treiman and Mercer - "Sinking the magma ocean: new evidence from lunar meteorites and the return of serial magmatism" - LPSC 2012
- Sm-Nd: differences in Nd-isotopics in A16 FAN and LM anorthosites; some samples from LREE enriched sources and some from HREE enriched sources, suggests 'lunar anorthosites do not derive from a single source, i.e., they are not all products of the LMO' (Nyquist et al 2010) - okay, I am so unconvinced, how does this play with e.g. the mantle zoo? what kind of heterogeneities do we see in Sept-Iles (i.e. across a much smaller volume/surface)? Flotation from multiple cycles of large-scale heterogeneities (does the silica content remain low enough for long enough)?
- ages of lunar anorthosites: variable, ish, with huge error bars
Implications:
- perhaps see any rocks from LMO, but instead see rocks taht formed from more complex geological processing, for example:
-- crust formed quickly, which was then processed by a hot, partially melted mantle; anorthositic blebs rose to the surface by serial magmatism; lots of diapirism
-- preserves initial heterogeneity, causes wide range in trace and major element chemistry and gives rise to a range of ages for anorthosite
- not arguing that LMO NEVER existed, just that we don't see ROCKS from it (so far) (see e.g. Longhi & Aswal, 1985)
Discussion:
- what would ratio of initial crust to later magmatism be? -- she reckons we see very little of the very earliest crust
- Moon asymmetry - far side mafics more magnesian?
- primary claim that anorthosite did not predominantly form within 1Ma of morn formation