TY - BOOK
T1 - The 1.9 Ga Winnipegosis Komatiite: Implications for Earth Accretion, Mantle Dynamics, and Komatiite Formation
AU - Waterton, Pedro
PY - 2018
Y1 - 2018
N2 - Komatiites are high temperature, high MgO lavas that have the potential to provide geochemical and thermal insights into the evolution of the Earth's mantle. However, komatiite studies are frequently hampered by the altered nature of all komatiites and the presence of large gaps in the komatiite record, most notably during the Proterozoic. This thesis characterises the Palaeoproterozoic Winnipegosis komatiites with a comprehensive study of their petrography, bulk rock and mineral geochemistry, isotope systematics, geochronology, and in-depth thermodynamic and geochemical modelling. These data and modelling are used to explore komatiite formation, the geochemical and thermal history of the mantle, and the relationship between komatiites and basalts.
The komatiites occur in the Winnipegosis Komatiite Belt (WKB), a greenstone belt dominated by basalt and komatiite with no surface exposure. New U-Pb SHRIMP dating of mafic zircons from the WKB yields an age of 1870.3 ± 7.1 Ma. Combined with an abundance of mafic and ultramafic magmatism and depleted geochemical signatures, this age confirms that the WKB is part of the ~1.88 Ga Circum-Superior Belt, a Proterozoic large igneous province. The komatiites are very well preserved and dominated by massive olivine porphyritic flows with a median thickness of 6 m. Differentiated flows containing layers of olivine spinifex are present, but rare.
Trace element data indicate the komatiites were derived from depleted mantle, and contaminated by interaction with continental crust. Al-in-olivine thermometry combined with olivine-melt Fe-Mg partitioning indicates that the parental melts to Winnipegosis komatiites were nominally dry, with an MgO content of 23.6 ± 1.6 wt%. The presence of olivine spinifex textures and derivation from liquids with > 18 wt% MgO demonstrates the Winnipegosis samples are komatiites sensu stricto. However, liquidus temperatures of 1501 ± 32 °C are approximately 100 °C cooler than their hottest Archaean counterparts, and are also significantly cooler than the hottest Phanerozoic picrites. The geochemical and geological evidence requires that Winnipegosis Komatiites were erupted onto continental crust, and high liquidus temperatures require anomalously hot mantle. A tectonic model is proposed in which Winnipegosis komatiites were generated in a mantle plume deflected towards the margins of the Superior craton by strong gradients in lithospheric thickness. This interpretation of komatiitic magmatism in the Circum-Superior Belt casts doubt on previous suggestions that ambient mantle potential temperatures were as high as 1600 °C during the Palaeoproterozoic.
Olivine was the liquidus phase during crystallisation of Winnipegosis komatiites, followed by chromite which saturated at ~1424 °C. Bulk rock geochemical relationships do not represent liquid lines of descent, and are instead controlled by mixing of these phenocryst phases with residual melt shortly before or during eruption. In the massive flows crystallisation continued primarily through the growth of skeletal olivine and clinopyroxene in the groundmass, but some differentiated flows also crystallised clinopyroxene phenocrysts. Plagioclase is a rare groundmass phase though, in general, glass (now devitrified) formed before plagioclase crystallisation. Thermodynamic modelling using MELTS software reproduces this crystallisation sequence and, with minor caveats, can reproduce the liquid line of descent of Winnipegosis komatiites.
The relationship between komatiites and associated tholeiitic basalts has long been a subject of debate. The link between these rock types in the WKB is investigated by extending the MELTS modelling to crystallisation at higher pressures. This modelling demonstrates that tholeiitic basalts in the WKB likely formed through ~60% crystallisation of Winnipegosis komatiite parental melt in an upper crustal magma chamber. The presence of a large dunite body, representing a komatiitic olivine-chromite adcumulate, provides clear evidence of such upper crustal fractionation in the WKB.
Re-Os isotopic data for Winnipegosis komatiites define an isochron with age 1865 ± 40 Ma, and chondritic γOs. This age agrees with the U-Pb age of the WKB, and combined with demonstrable PGE control by igneous phases, confirms that the PGE systematics of Winnipegosis komatiites were undisturbed during metamorphism. Winnipegosis komatiites parental melts contained 7.0 ± 0.8 ppb Pt and 7.2 ± 0.7 ppb Pd. These concentrations are > 60% lower than predicted by geochemical modelling under any reasonable assumptions of melting conditions, melt fraction and source composition. This discrepancy cannot be explained by melting of a source depleted in PGEs. Instead, comparison with parental melt compositions of other well-preserved komatiites and picrites indicates that both Pd and Pt may exhibit some compatibility during high degree melting, and may not be exhausted from mantle residues until extreme melt fractions are reached. It is suggested that low PGE concentrations in komatiites reflect partial retention of PGEs in their mantle sources, and differences in melting conditions, rather than evidence of low-PGE mantle generated by uneven late accretion and/or magma ocean differentiation.
AB - Komatiites are high temperature, high MgO lavas that have the potential to provide geochemical and thermal insights into the evolution of the Earth's mantle. However, komatiite studies are frequently hampered by the altered nature of all komatiites and the presence of large gaps in the komatiite record, most notably during the Proterozoic. This thesis characterises the Palaeoproterozoic Winnipegosis komatiites with a comprehensive study of their petrography, bulk rock and mineral geochemistry, isotope systematics, geochronology, and in-depth thermodynamic and geochemical modelling. These data and modelling are used to explore komatiite formation, the geochemical and thermal history of the mantle, and the relationship between komatiites and basalts.
The komatiites occur in the Winnipegosis Komatiite Belt (WKB), a greenstone belt dominated by basalt and komatiite with no surface exposure. New U-Pb SHRIMP dating of mafic zircons from the WKB yields an age of 1870.3 ± 7.1 Ma. Combined with an abundance of mafic and ultramafic magmatism and depleted geochemical signatures, this age confirms that the WKB is part of the ~1.88 Ga Circum-Superior Belt, a Proterozoic large igneous province. The komatiites are very well preserved and dominated by massive olivine porphyritic flows with a median thickness of 6 m. Differentiated flows containing layers of olivine spinifex are present, but rare.
Trace element data indicate the komatiites were derived from depleted mantle, and contaminated by interaction with continental crust. Al-in-olivine thermometry combined with olivine-melt Fe-Mg partitioning indicates that the parental melts to Winnipegosis komatiites were nominally dry, with an MgO content of 23.6 ± 1.6 wt%. The presence of olivine spinifex textures and derivation from liquids with > 18 wt% MgO demonstrates the Winnipegosis samples are komatiites sensu stricto. However, liquidus temperatures of 1501 ± 32 °C are approximately 100 °C cooler than their hottest Archaean counterparts, and are also significantly cooler than the hottest Phanerozoic picrites. The geochemical and geological evidence requires that Winnipegosis Komatiites were erupted onto continental crust, and high liquidus temperatures require anomalously hot mantle. A tectonic model is proposed in which Winnipegosis komatiites were generated in a mantle plume deflected towards the margins of the Superior craton by strong gradients in lithospheric thickness. This interpretation of komatiitic magmatism in the Circum-Superior Belt casts doubt on previous suggestions that ambient mantle potential temperatures were as high as 1600 °C during the Palaeoproterozoic.
Olivine was the liquidus phase during crystallisation of Winnipegosis komatiites, followed by chromite which saturated at ~1424 °C. Bulk rock geochemical relationships do not represent liquid lines of descent, and are instead controlled by mixing of these phenocryst phases with residual melt shortly before or during eruption. In the massive flows crystallisation continued primarily through the growth of skeletal olivine and clinopyroxene in the groundmass, but some differentiated flows also crystallised clinopyroxene phenocrysts. Plagioclase is a rare groundmass phase though, in general, glass (now devitrified) formed before plagioclase crystallisation. Thermodynamic modelling using MELTS software reproduces this crystallisation sequence and, with minor caveats, can reproduce the liquid line of descent of Winnipegosis komatiites.
The relationship between komatiites and associated tholeiitic basalts has long been a subject of debate. The link between these rock types in the WKB is investigated by extending the MELTS modelling to crystallisation at higher pressures. This modelling demonstrates that tholeiitic basalts in the WKB likely formed through ~60% crystallisation of Winnipegosis komatiite parental melt in an upper crustal magma chamber. The presence of a large dunite body, representing a komatiitic olivine-chromite adcumulate, provides clear evidence of such upper crustal fractionation in the WKB.
Re-Os isotopic data for Winnipegosis komatiites define an isochron with age 1865 ± 40 Ma, and chondritic γOs. This age agrees with the U-Pb age of the WKB, and combined with demonstrable PGE control by igneous phases, confirms that the PGE systematics of Winnipegosis komatiites were undisturbed during metamorphism. Winnipegosis komatiites parental melts contained 7.0 ± 0.8 ppb Pt and 7.2 ± 0.7 ppb Pd. These concentrations are > 60% lower than predicted by geochemical modelling under any reasonable assumptions of melting conditions, melt fraction and source composition. This discrepancy cannot be explained by melting of a source depleted in PGEs. Instead, comparison with parental melt compositions of other well-preserved komatiites and picrites indicates that both Pd and Pt may exhibit some compatibility during high degree melting, and may not be exhausted from mantle residues until extreme melt fractions are reached. It is suggested that low PGE concentrations in komatiites reflect partial retention of PGEs in their mantle sources, and differences in melting conditions, rather than evidence of low-PGE mantle generated by uneven late accretion and/or magma ocean differentiation.
KW - Faculty of Science
KW - Petrology
KW - Geochemistry
KW - Isotope Geochemistry
KW - Komatiite
KW - Proterozoic
KW - Major elements
KW - Trace elements
KW - Platinum group elements
KW - Re-Os isotopes
KW - Circum-Superior Belt
KW - Trans-Hudson Orogeny
KW - Al-in-olivine thermometer
KW - MELTS
U2 - 10.7939/R3N29PN52
DO - 10.7939/R3N29PN52
M3 - Ph.D. thesis
BT - The 1.9 Ga Winnipegosis Komatiite: Implications for Earth Accretion, Mantle Dynamics, and Komatiite Formation
PB - University of Alberta
ER -