Northeastern Section - 53rd Annual Meeting - 2018

Paper No. 32-6
Presentation Time: 8:00 AM-12:00 PM

USING THE MELTS ALGORITHM TO MODEL MAGMA CRYSTALLIZATION WITHIN AN UPPER-CRUST PLUMBING SYSTEM IN THE WESTERN NEWARK BASIN, CENTRAL ATLANTIC MAGMATIC PROVINCE, USA


WEAVER, Alexandria1, SROGI, LeeAnn1, LUTZ, Tim1 and POLLOCK, Meagen2, (1)Department of Earth & Space Sciences, West Chester University, 720 S Church St, West Chester, PA 19383, (2)Department of Geology, The College of Wooster, 944 College Mall, Scovel Hall, Wooster, OH 44691

We are investigating mafic igneous rocks of the Central Atlantic Magmatic Province that formed at about 201.5 Ma. In the western Newark Basin, post-magmatic tilting, folding and erosion have exposed cross-sections through the upper crust from the basalt lava flow down to 6 km. Samples of basalt and chilled margins of underlying diabase plutons approximate the parental liquid. These are used as starting compositions in the rhyolite-MELTS (v. 1.0.x) algorithm of Ghiorso and Sack (1995) and Gualda et al., (2012) to model the crystallization of the magma under upper-crustal conditions. Results will be compared with bulk rock and mineral compositions from samples throughout the plumbing system that correspond to paleodepths from about 0.3 to >5 km (~ 0.1 – 1.5 kb). All MELTS models were started at oxygen fugacity = QFM and then run without oxygen buffering, with fractional crystallization and a 10°C Temperature step. Average basalt and diabase chill compositions are very similar and MELTS runs at P = 1 kb and 2 wt.% H2O produced very similar mineral sequences and mineral and liquid compositions; however, opx crystallized in the basalt but not the diabase chill. We attempted 16 initial MELTS runs to bracket conditions at P = 0.1, 0.5, 1, and 2 kb; and 0.5, 1, 2, and 3 wt.% H2O, but some runs were unsuccessful due to initial water contents above saturation. In successful runs, olivine composition and the behavior of augite, spinel (magnetite), and rhombohedral-oxide (ilmenite), are very similar. By contrast, P and water content cause variations in liquidus T (higher at higher P and lower H2O) and liquidus phase: opx (En83) at P = 1-2 kb and 0.5-1% H2O and oliv (Fo82-83) at lower P and higher H2O. Compositions of pigeonite and plagioclase also vary with water content and P: pig has higher Mg# and plag has lower An at higher P and lower H2O. MELTS runs predict late Fe-rich oliv (Fo31-19) which has not yet been identified in thin section. The water saturation T in the liquid is lower at higher P or lower initial H2O, and saturation was not reached before the runs ended with 0.5% initial H2O at 1 and 2 kb. Additional MELTS runs will more narrowly bracket P-T-H2O conditions and will be used to test if different mineral compositions and sequences observed in the samples are due to the effects of P and H2O or require different magma compositions.