On Archean granites, greenstones, cratons and tectonics: does the evidence demand a verdict?

Abstract

I review the geologic evidence for Archean plate tectonic processs within the framework of geophysical, geochemical and experimental observations. For the Late Archean (2.5-3.0 Ga), the evidence is primarily based on the rich data-base of the Superior craton; for the early Archean (3.0-4.0 Ga), data from the Pilbara and Kaapvaal cratons are examined. Data from other old cratons are used as supplementary evidence. The verdict is that there is a robust consensus on the validity of using plate tectonic boundary processes to decipher the late Archean rock record; and it also confirms that such processes dominated the early Archean. Modern plate tectonics probably has its roots in the Hadean-Archean transition between 4.0 and 4.2 Ga. Before that time, recycling of thin, dry oceanic lithosphere may have dominated the surface tectonics of Earth. Assuming gradual mantle dehydration during formation of oceanic lithosphere, from 4.5 Ga onward, c70% of mantle water would have accumulated in the hydrosphere by about 4.0 Ga. At this stage, Earth's seafloor-spreading plate boundaries, would, for the first time, have come to operate below sea level, in turn, initiating onset of extensive hydrothermal cooling and alteration of upper oceanic lithosphere as observed today. Only then did 'modern' production of granitoids become possible during the recycling of Earth's lithospehre. Processes at convergent plate boundaries, however, evolved more slowly into a modern subduction-dehydration style over the course of the Archean. In the earliest Archean, recycling of buoyant hydrated oceanic materials into the mantle may not have been efficient, and back-arc spreading processes may have been absent. Instead, I speculate that hydrated oceanic lithosphere 'piled-up' to form intraoceanic thrust-stacks, which evolved into Earth's first continental fragments through internal differentiation during tectonic thickening and construction of deep lithospheric keels. These fragments amalgamated into larger cratons throughout Archean times. Minimum continental growth rates of various cratons spanning the Archean period illustrate that Early Archean continental crustal-growth must have been 'explosive', and that only a small fraction of this crust has been preserved. I speculate that modern-subduction-style recycling of exosphere materials (here taken as crust, sediments, hydrosphere, atmosphere) may not have been efficient during the Early Archean, and that many Early Archean continental fragments, possibly those lacking a thick buoyant mantle keel,were episodically 'flushed' whole-scale, back into the convecting mantle. Those cratons that survived, built up and/or retained thick Archean mantle keels (between 200 and 400km). Delamination of these keels cannot, therefore, be invoked to explain subsequent Archean (and later) tectonic events recorded at the surface of these cratons. The style of modern subduction-related recycling of exosphere materials emerged during the late Archean. These conjectures have major consequences for global mass balance and chemical flux considerations. New experimental work indicates that komatiitic magmas were hydrous, and that their liquidus temperatures are comparable to those of magmas observed at modern plate boundaries. This, and other independent observations indicate that the upper mantle in the Archean was not excessively hotter than today. The presence of significantly dissolved water in komatiitic magmas raises the possibility that Earth lost heat more effectively through dehydration from a wetter, cooler, and less viscous mantle: 'wet' Archean mantle dynamics may have been significantly different than the present 'dry' one. Since hot spot volcanism is less efficient at dissipating internal heat as through spreading at mid-ocean ridges, it is unlikely that Archean plates were swamped by plumes, or replaced by plume tectonics. Rather, accretion and hydrothermal cooling processes at seafloor-spreading plate boundaries were possibly more efficient; and deep mantle plume material may have ben deflected via the asthenosphere into these boundaries more effectively than observed today. Inferred secular changes in Earth's tectonic processes are mostly based on qualitatitave statements concerning the relative increase/decrease of preserved rock types/rock associations through time. The generally accepted decline of komatiite abundance (and their average MgO contents) from the Early Archean is a good example; and so is the apparent increase in the abundance of kimberlites with time. Analysis of the rock make-up of 40 greenstone belts worldwide, for which there is reasonable quantitative data, indicates that there is no unequivocal decline in the volume of preserved komatiites, or in the maximum MgO content of ultramafic magmas from the Early Archean to the Phanerozoic. There is, however, a peak in preservation of komatiites in globally distributed greenstone belts of the Late Archean. Similarly, there..........