Mullen, E. and McCallum, I.S., 2009, Petrogenesis of Mt. Baker Basalts and Andesites: Constraints From Mineral Chemistry and Phase Equilibria. American Geophysical Union, Fall Meeting 2009, abstract #V51C-1703. http://adsabs.harvard.edu/abs/2009AGUFM.V51C1703M

Petrogenesis of Mt. Baker Basalts and Andesites: Constraints From Mineral Chemistry and Phase Equilibria. American Geophysical Union, Fall Meeting 2009, abstract #V51C-1703. http

Basalts in continental arcs are volumetrically subordinate to andesites and this is the case for Mt. Baker in the northern Cascade magmatic arc. However, basalts provide indirect evidence on mantle compositions and processes that produce magmas parental to the abundant andesites and dacites of the stratocones. Basalts at Mt. Baker erupted from monogenetic vents peripheral to the andesitic stratocone. Flows are variable in composition; some samples would more appropriately be classified as basaltic andesites. The “basalts” have relatively low Mg/(Mg+Fe) indicating that they have evolved from their original compositions. Samples studied are Park Butte, Tarn Plateau, Lk. Shannon, Sulphur Cr. basalts, and Cathedral Crag, Hogback, and Rankin Ridge basaltic andesites. Mt. Baker lavas belong to the calc-alkaline basalt suite (CAB) defined by Bacon et al. (1997) and preserve arc geochemical features. High alumina olivine tholeiite (HAOT) are absent. Equilibrium mineral pairs and whole rock compositions were used to calculate pre-eruptive temperatures, water contents, and redox states of the “basalts.” All samples have zoned olivine phenocrysts with Fo68 to Fo87 cores and chromite inclusions. Cpx and zoned plagioclase occur in all flows, but opx occurs only in Cathedral Crag, Rankin Ridge, and Tarn Plateau. Ti-magnetite and ilmenite coexist in all flows except for Sulphur Cr., Lk. Shannon and Hogback, which contain a single Fe-Ti oxide. Liquidus temperatures range from 1080 to 1232°C and are negatively correlated with water contents. Water contents estimated using liquidus depression due to H2O (0.8 to 5.4 wt.%) agree well with plag core-whole rock equilibria estimates (1.2 to 3.9 wt.%). Park Butte, Sulphur Cr. and Lk. Shannon had <1.5 wt.% H2O, and Cathedral Crag is most hydrous. Redox states from ol-chr pairs (QFM +0.1 to +2.8) and Fe-Ti oxide pairs (QFM -0.6 to +1.8) indicate that Park Butte and Sulphur Cr. are most oxidized and Cathedral Crag most reduced; however, the two methods do not give consistent results. The water content and redox state of the basalts are inversely correlated, inconsistent with data from andesites that show the reverse correlation. Using published experimental data and the BATCH algorithm (Longhi, 2002) we constructed an array of phase diagrams in the multi-component basalt system relevant to arc basalts and andesites ranging from 0 to 3 GPa and variable water contents. Projections of Mt. Baker lava compositions (corrected for loss or gain of olivine and plag where appropriate) on these diagrams reveal: (1) with the exception of Sulphur Cr., primary basaltic compositions equilibrated with depleted hydrous mantle harzburgite/lherzolite at pressures from 1 to 1.5 GPa, coincident with the crust-mantle boundary in the Mt. Baker region, (2) except for Sulphur Cr., melt fractions were >10%; Sulphur Cr. basalt is alkalic and formed by smaller degrees of partial melting comparable to basalts from the northern Garibaldi belt, (3) evidence for shallow fractionation of basalts (5-10 km), (4) Mt. Baker andesites delineate a low pressure fractionation trend coincident with the 0.2 GPa, water-saturated, oliv+cpx+plag and cpx+amph+plag cotectics (Sisson and Grove, 1993, Grove et al., 2003).