Page 8 - i1052-5173-26-9

subduction pre-subduction
stage 2
stage 5 stage 4 stage 3 1 0.10

Model SiO2 (%) 58 A Time-integrated Rb/Sr

0.08

54 0.06
0.04
95% confidence 0.02
50 66% confidence

maximum probability
46

New crust formation age (Ga)

New crust thickness (km) Volume of continental crust (%)125 B late heavy bombardment

100 Dhuime et al. 2012

75 Condie & Aster 2010 Belousova et al. 2010
50

25 present day surfaceGoodwin 1996

ages distribution hot, shallow pre-subduction 1700
regime 1600
400 C 1500
T ~ 250oC 1400
35 Earth’s subduction 175oC
0.34
30 middle age Urc(r0u)st=al thickness T ~ Tp (oC)
25 subcdmoulacdntitolentemperature 80oC
T ~

20

15 1300
0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
Age (Ga)

Figure 4. (A) Variation in the estimated Rb/Sr ratios, and SiO2 contents, of new continental crust from Dhuime
et al. (2015). (B) A preliminary model for the changes in volume of the continental crust (dotted blue curve)
compared with the crustal growth curves of Belousova et al. (2010) and Dhuime et al. (2012), the model age
distribution of Condie and Aster (2010), and the present-day surface age distribution of Goodwin (1996). The
preliminary model is illustrative, and it is not unique. It assumes two types of crust generated before and after
3.0 Ga, and rates of crust generation and destruction for each. It is constrained by the volume of continental
crust at the present day and 70% at 3.0 Ga, and by the present-day curve of Condie and Aster (2010).
(C) Variation in the thickness of new continental crust through time (orange curve) as estimated from the
Rb/Sr ratio of new continental crust (Dhuime et al., 2015), and thermal models for ambient mantle for Urey
(Ur) ratio of 0.34 (Korenaga, 2013).

GSA TODAY | SEPTEMBER 2016 perspective (Fig. 4C; Dhuime et al., 2015). The high crustal then the thickness (and volume) of the crust may have increased
growth rates before 3.0 Ga were marked by relatively thin conti- with little or no increase in area.
nental crust, and at that time there is no link between estimated
crustal thickness and crustal growth rates. However, by 3.0 Ga the One issue is the extent to which the crustal thickness at the sites
estimated volume of crust was at least ~70% of the present-day of generation of new crust can be linked to crustal volume. Our
volume, and the crustal thickness was ~50% of the present-day preliminary models suggest that the predominantly mafic crust
crustal thickness. It therefore appears that before 3.0 Ga the area generated before 3.0 Ga was largely destroyed by 2.0 Ga, and that
of continental crust increased with crustal volume, but that since since that time the crust predominantly consisted of post–3.0 Ga
crust generated in subduction-related settings. Because the relation

8