Empirical garnet–muscovite geothermometry in low-grade metapelites, Selwyn Range (Canadian Rockies)

Abstract Partitioning of Fe and Mg between garnet and phengitic muscovite was calibrated as a geothermometer by Green & Hellman (1982) using experimental data at 25–30 kbar. When the thermometer is applied to pelites regionally metamorphosed at pressures of between 3 and 7 kbar it yields temperatures much higher than those from the garnet–biotite thermometer. A new empirical calibration is proposed for use with such rocks, with particular application where garnet occurs at lower grades than biotite. The new calibration is where K is given by: In K = In K _d and X _ii are mole fractions in the garnets.
The calibration was derived from comparison with the garnet–biotite thermometer of Ferry & Spear (1978), assuming no pressure-dependence for the partitioning between garnet and muscovite, no ferric iron partitioning, ideal mixing in muscovite, and the garnet mixing model of Ganguly & Saxena (1984) modified for a non-linear Ca effect. This latter garnet mixing model was selected because it gave the geologically most reasonable results. It has not proved possible to distinguish a pressure effect from a ferric-iron effect.
Despite the simplifying assumptions used to derive the calibration, it yields temperatures generally within 15°C of those given by the garnet–biotite thermometer, and has been used to supply thermometric data in a low-grade region of the Canadian Rockies.