Time after solar system formation

Fig. 27.17 Hafnium-isotope composition in ancient zircons. Hafnium is compatible in zircon and its isotope composition has been frozen in this mineral since its crystallization. The spread of the data points indicates magmatic differentiation that started ~ 100 Myr after SOS formation. The line of lowest gradient shows possible 176Hf/177Hf development in 4.4-Gyr-old continental crust (with Lu/Hf ratio from Table 26.3). The grey lines show the model development of DMM and CCR reservoirs, discussed in Chapter 28. Note that the CCR model incorporates both ancient and newly formed crust and is therefore intermediate between the lowest-gradient line and the DMM line. Data from Wilde et al. (2001), Amelin et al. (1999, 2000) and Harrison et al. (2005).

the DMM with age. A similar pattern appears for the £i76 evolution (e.g. Vervoort and Blichert-Toft, 1999; Bennett, 2003).

These trends are used to constrain chemical Earth-evolution models (Chapter 28). A simple approach already allows us to resolve a controversy on the continental crust growth since the Archaean (e.g. Taylor and McLennan, 1995). Extreme end-member hypotheses are (1) that the crust grew without any recycling into the mantle, and (2) that the mass of continental crust has been constant since the Archaean because recycling into the mantle has kept pace with the generation of new crust (see inset in Fig. 27.18). Both hypotheses allow a mean crust age of 2 Gyr and the observed

Age, Gyr

0.2807

0.2805

6 0.2801

0.2799

Zircons from gneisses:

Depleted

£

MORB-source

Qrr C) *

o Pilbara (NW Australia)

mantle

Yo • * ^

Barberton (South Africa)

fW ^^

" Itsaq (SOI) (W Greenland)

Acasta (NW Canada)

1 A

Detrital zircons:

% A

O Jack Hills (W. Australia)

A \ ^

A Bulk continental

+ ♦

- vnr'

Upper-crustal

rocks

- / *

S Bulk silicate Earth

t typical 1 uncertainty i i

i i i

i i

present-day s143 values of the crust and the DMM. However, their resulting mantle-evolution trends for £143(T are totally different from that observed; it is obvious from this simple test that neither of these extreme hypotheses is correct. In detailed modelling an intermediate crustal-growth model produces a good fit to the upper bound of the data envelope (the curve DMM in Fig 27.18). Further, the model described in Chapter 28 envisages a wholly convective mantle and an apparently isolated enriched D" reservoir (Chapters 17 and 19). The very early formation of this reservoir would have caused initial mantle depletion ~ 4.5 Gyr ago, leading to the early growth of mantle £143(T) even if very little or no continental crust was present prior to 4 Gyr.

0 0

Post a comment