THERMODYNAMIC MODELING AND POTENTIAL CONSTRAINTS ON THE THERMAL HISTORY OF THE WICHITA GRANITE GROUP, OKLAHOMA

The Wichita Granite Group (WGG) is a series of compositionally similar alkali-feldspar granites. These were emplaced as shallow sheets during Cambrian rifting, and they now underlie much of southern Oklahoma. Multi-phase thermodynamic modeling using Rhyolite-MELTS 1.1.0 (Ghiorso and Gualda, 2015, CMP 169.6) provides thermal and phase-reaction constraints for WGG crystallization. Initial modeling evaluated the two WGG P-T paths for finer vs. coarser microstructure as outlined in Hogan et al. (2000, ESTrans. Roy. Soc. Ed., 91.1-2) using averaged compositions of individual lithodemic units and 1 wt.% F. Path 1 follows the ponding and initial crystallization (~30%) at 7-8 km, followed by ascent, and soldification at 1 km. Path 2 has the magma crystallize entirely at 3.5-4 km. WGG PH₂O is presumed to be low, so models compared “dry” (0.5 wt% initial H₂O_i) to “damp” (2 wt.%) magma.

Naturally, H₂O_i has the greatest impact on liquidi T’s (960-1000 °C, dry; 870-910 °C, damp). All models place ternary Ksp and Mag on the liquidi; Qz becomes stable with cooling. H₂Oi fails to impact the low-T assemblage of subequal % of Qz + Ab + Or, with lesser Mt + Bt ± Ttn. This outcome is consistent with some of the WGG, although the subsolvus feldspar ratios differ, undoubtedly affected by metastability. H₂O saturates only in damp models, but may resorb with Bt growth. In path 1 models, Qz appears before ascent, and is destabilized with Ksp during decompression, consistent with some WGG microstructure. Path 2 can produce a Pxn at intermediate T’s, but disappears at Bt in.

The models poorly predict the nature of Ttn and Hbl common to many WGG. Titanite, when stable in modeling, is late and at the expense of Ilm, and Hbl is entirely absent. In reality, many WGG yield Hbl T’s of 800 and Ttn T’s of 750 °C, and both are stable in experimental melts at 850 °C. Raising model fO₂ fails to extend Ttn stability and produces unrealistic liquidi. Fluorine is unaccounted for in modeled Ttn and Hbl, and it likely stabilizes these phases.

Except for vapor saturation, modeling produces similar outcomes for a given path. Damp path 2 and dry path 1 are more consistent with the thermal profile given by geothermometry and rock microstructure. These preliminary results suggest that felsic melt production may capture different source volatile concentrations.