2007 GSA Denver Annual Meeting (28–31 October 2007)

Paper No. 13
Presentation Time: 11:00 AM

TOWARD A THERMAL MODEL FOR THE SKAERGAARD LIQUIDUS


MORSE, Stearns A., Dept. of Geosciences, Univ of Massachusetts, 611 North Pleasant Street, Amherst, MA 01003-9297, tm@geo.umass.edu

The advent of cumulus augite at LZa/b in the Skaergaard intrusion defines the equilibrium L = OL + PL + AUG at an estimated pressure of 1290 bars based on stratigraphic depth and a reference point of 600 b at the Sandwich Horizon (Lindsley et al. 1969 MSA Sp Pap 2, 193). The same equilibrium occurs in the Kiglapait intrusion at a well-defined experimental P = 5 kb, T = 1203 +/- 2 C in graphite (Morse et al. 2004 JPet 45:2225). By coincidence, the plagioclase component of the liquid in both intrusions has the composition An 39, as estimated from linear partitioning using a pressure-sensitive KD = 0.4 for Skd and 0.525 for KI (see Morse et al 2004 again). Assuming our observed liquidus gradient of 8 C/kb, this 4-phase eqm scales to 1173 C at the pressure of Skd LZa/b. We now use cumulus cores = liquidus plag An from Toplis et al. 2007 and an experimental variation of 4.8 C/unit An to calculate all other temperatures from the base of LZa to the base of UZc. Using the same pressure scale, the melting experiments of McBirney and Naslund (1990 CMP 140:235) are adjusted for pressure and also plotted against zone volumes scaled as stratigraphic height. The result is a straight-line variation of T = -0.0382H (H in meters) + 1167, the fictive T at base of LZ, with a Y axis error estimate of 3 C and Rsqd = 0.993. ALL of the PL-based points from LZb to UZa lie within error of the M&N results, yielding a powerful agreement between difficult melting experiments and the plag-based scale. The M&N LZa value falls low and has the greatest uncertainty because of the experimental difficulty of estimating the end of intercumulus melting. The PL-based scale gives 1191 C at the base of LZa and 1207 C at the base of HZ where An = 68 in the Cambridge drill core (M. Holness PC 2007). It gives 1040 C at the base of UZc, implying a sharp drop to less than 1000 C at SH. These temperatures are far more reliable than those reported at 1-atm FMQ, for liquidus compositions in OL and PL that are far (5-20 mole%) more refractory than in any Skd rock, by Thy et al. (2006 Lithos 92:154), which range from +20 C to -34 C relative to the combined M&N and PL-based scale. By contrast, the M&N results were found at variable oxygen fugacities consistent with the mineral equilibria.