Archaean mantle models

In the first part of this chapter we shall examine the structure and composition of the modern mantle in order to establish how it works. In so doing we will find tantalizing clues which relate to the mantle s earlier history. It is these clues that we shall explore in the second part of this chapter and use to identify the nature and chemical evolution of the Archaean mantle. These data are then used in the third part of the chapter to constrain models for the Archaean mantle. 

Alternative views of early Archaean mantle evolution require that mantle depletion started as early as ca. 4.5 Ga (see compilation in Rollinson, 1993). These models imply significant mantle Sm-Nd fractionation in the very early Archaean and have major implications for the differentiation of the early Earth. One such study is that of Bennett et al. (1993) who measured very high eNd values (+3.5 to +4.5) in 3.81 Ga Amitsoq gneiss samples. Collerson et al. (1991) also calculated an isochron eNd value of +3.0 for 3.8 Ga-old peridotites from northern Labrador. The extreme deviation from CHUR early in Earth history (Fig. 3.27) was interpreted by Bennett et al. (1993) as evidence for an extreme and very early fractionation of the Earth s mantle relative to CHUR. Such an event implies the formation of extensive continental crust prior to 3.8 Ga, for which there is no independent geological evidence. This apparent paradox and the claim for very early extensive mantle differentiation led to a detailed reexamination of the Bennett... 

The co-ordinated Kaapvaal Project geochron-ological studies of crustal and mantle xenoliths reveal that both crust and mantle have experienced a multi-stage history, and that a simple view of cratonization as a discrete event is not a viable model for craton formation (Schmitz et al. 1998 Schmitz Bowring 2000 Moser et al. 2001). The lower crust in particular retains a comprehensive record of the tectonothermal evolution of the lithosphere. The study of lower-crustal samples has shown that much of the deep craton experienced a dynamic and protracted history of tectonothermal activity that is temporally associated with events seen in the surface record, including late Archaean magmatism (Ventersdorp) and even Proterozoic deformation (Namaqua-Natal) (Schmitz et al. 1998). Thermal events are... 

The U-Th-Pb isotope systems have half-lives ranging from 0.7 to 4.6 Ga, and can therefore be used to investigate variations in time-integrated U/Pb (pi) and Th/U (ki) ratios, and to constrain models for the evolution of the mantle and crust from early Archaean time to the present (Sinha Tilton 1973 Stacey Kramers 1977 Kramers Tolstikhin 1997). The Pb isotope compilation and model of Kramers Tolstikhin (1997) highlight the range in the initial Pb/ Pb ratios (and thus in time-integrated pi) in suites of Archaean to early Proterozoic rocks. Inferred differences between high-pi lower crust and... 

We summarize the results of several plume simulations for a small craton, using the approach of Sleep (1997). We then analyse the effect of plume material emplacement beneath cratonic keels and channels of different dimensions. This work presents new simulations of plume flow beneath the African continent that include deep keels beneath Archaean cratons, which were not considered in the simple models of Ebinger Sleep (1998). Flow velocities and strains predicted from our preferred model allow us to estimate the magnitude and direction of SKS splitting by plume flow around a cratonic keel. These patterns are then compared with SKS splitting patterns from normal mantle flow around a keel (e.g. Fouch et al. 1999). 

Fig. 13. Four important reservoirs of CO2 are shown as functions of time for the models in Figure 12. High heat flow is denoted by continuous lines, low heat flow by dashed lines. Here we have chosen models in which the crustal reservoirs are initially constant in time i.e. we have started from the equilibrium reservoirs. In particular, the equilibrium continental reservoirs are small and so these models begin with very little continental carbonate. The high heat-flow models chiun the reservoirs fast enough that if we do not start at equilibrium values, the model quickly evolves to them, but in the low heat-flow models circulation is slow enough that the arbitrary initial conditions are remembered well into Archaean time. In general, the effect of abundant Hadean impact ejecta is to remove CO2 from the continents and oceans and put it into the mantle.

Fig. 14. Fluxes of CO2 are shown as functions of time for the high heat-flow models in Figures 12 and 13. Impact ejecta are of minor importance compared with the rapid churning of the oceanic crust. In Archaean time, CO2 is mostly controlled by processes involving the creation and subduction of oceanic crust. Continents become increasingly important through Proterozoic time, with the transition from mantle to continental control occurring at c. 1.4 Ga.

---