Petrogenetic Constraints at Mount Rainier Volcano, Washington

S. C. Kuehn and P. R. Hooper, Department of Geology,
Washington State University, Pullman, WA

A. E. Eggers and C. Kerrick, Department of Geology,
University of Puget Sound, Tacoma, WA

Poster presented at the American Geophysical Union Fall Meeting, December 11-15, 1995 in San Francisco, CA


Abstract

The Mount Rainier lavas are a typical calc-alkaline suite with LIL elements enriched relative to HFS elements and ranging from basaltic andesite to dacite. Chemical variations may be observed in stratigraphic section where there are significant chemical breaks with relative stratigraphic height. At least one break corresponds to an unconformity in the edifice. The lavas form two principal groups, main cone and satellite vent flows, distinguished on the basis of geographic location, petrography, and chemical composition. The satellite vent flows, despite being temporally, spatially, and volumetrically limited, have a greater range in LIL and LREE compositions than the main cone flows. Major and trace element data for over 300 well-located samples of the Rainier lavas cluster into six sub-groups which cannot be related to each other by simple fractional crystallization; multiple source components are required. Simple two end-member mixing and simple mantle melting are also unable to individually explain the chemical variations between and within the chemical groups. No simple mixing lines are discerned; some combination of processes is required. Within chemical groups, positive correlations between relatively incompatible elements including P2O5 and TiO2 and Ba, Rb, Th, and Zr and negative correlations between Sr and Zr, for example, suggest that the Rainier flows have undergone some fractional crystallization. The presence of xenoliths (both exotic and of Rainier origin) and xenocrysts indicate that some assimilation has occurred. The Tatoosh plutonic complex beneath Rainier is a potential assimilant. The presence of quenched magmatic inclusions in the lavas, the nearly continuous 59-75% range in glass SiO2 content in the Rainier C-tephra, and visible heterogeneity in the C-tephra offer evidence for mixing and mingling of magmas at Rainier.


Panel 1

The Rainier lavas are a typical calc-alkaline suite ranging in composition
from basaltic andesite to dacite with more mobile LIL elements (K, Rb, Ba, Th)
enriched relative to less mobile HFS elements (Nb, P, Zr, Ti, Y).


Panel 2

Chemical discontinuities are present in stratigraphic section and
could prove useful as stratigraphic markers. At Liberty Ridge,
a 2% discontinuity in SiO2 corresponds to a mapped
unconformity (pers. comm. Tom Sisson, USGS).

Liberty Ridge Section

Success Cleaver Section

Additional plots of elements versus stratigraphic height


Panel 3

The Rainier lavas fall into two principal groups, the andesites and dacites of the main cone (Qra of Fiske, et. al., 1963) and basaltic andesites and andesites erupted from two satellite vents on the volcano's northwest flank (Qroa). The Qroa flows contain olivine phenocrysts even in flows with SiO2 as high as 62%. Major and trace element data for over 300 well-located samples show that these two groups differ in SiO2, P2O5, and TiO2, excepting the higher silica Qroa flows. The Qroa flows have greater trace element variability between flows than the surprisingly tight plotting Qra flows. This is in spite of the greater number of samples and greater time span represented by the Qra analyses.


Panel 4

The Qroa flows may be subdivided into 4 groups. They may be first distinguished into higher P2O5 (HP) and lower P2O5 flows. The lower P2O5 Qroa flows fall into high SiO2 Qra-like flows (HSi), higher TiO2 flows (HT), and lower TiO2 flows (LT). Geographically, the HP (and HSi) flows appear to be associated with the Echo Rock satellite vent.


Panel 5

The Qra fall into higher P2O5 (Qra-HP) and lower P2O5 flows (main bulk of samples). The Qra-HP flows occur principally at Liberty Ridge. The lower P2O5 Qra flows include a small group of possible plagioclase cumulates distinguished by higher Al2O3 and lower Mg#.


Panel 6

Compositional variations between and within groups provide constraints on the processes involved in the genesis of the Rainier lavas. Strong positive correlations are apparent between SiO2 and the alkalies (see panel 1). There are equally impressive correlations between these elements and MgO. These correlations are typical of calc-alkaline suites and might be interpreted as evidence for crystal fractionation, varying degrees of partial melting of a common source material, or a combination of these two processes. Positive correlations between P2O5 and TiO2 (panel 4) and between relatively incompatible elements such as Ba, Rb, and Th (and between these and SiO2) are also compatible with fractional crystallization.


Panel 7

Relatively incompatible minor and trace elements are often used to gauge fractional crystallization as these elements tend to remain in the melt and thus increase in proportion during fractional crystallization. This produces positive correlations between incompatible elements. The reliability of these elements can be tested by checking for negative correlations with Mg # which decreases during fractional crystallization. Rb, Ba, and Th do correlate negatively with Mg #. P2O5, TiO2, and Zr do not for most groups.


Panel 8

Plots of Rb versus Ba (see panel 5) and Ba and Rb versus Th show strong, positive, one-to-one correlations. These correlations could be explained by fractional crystallization. However, they could as easily be explained by mixing with an evolved component.


Panel 9

Elements such as Eu (and Sr) which are compatible during fractional crystallization of plagioclase are also used. The lack of a negative Eu anomaly for most flows indicates that fractional crystallization could not have been dominated by plagioclase.

Plagioclase, two pyroxenes, and Fe-Ti oxides are present as phenocrysts in most main cone flows. Amphibole (typically oxidized) and are olivine found only in a limited number of flows. Apatite is not uncommon, and at least one zircon has been observed.


Panel 10

A plot of Ba against Sr suggests that varying degrees of partial melting may have played a role in producing some of the differences between groups. Since Ba and Sr are equally incompatible during mantle melting, varying degrees of partial melting of a mante source produces a positive correlation. Some of the chemical groups are separated along a positive trend. Negative trends on the same plot can be produced by fractional crystallization. Small variations in degree of partial melting and fractional crystallization produces a scatter.

Simple mixing of only two independently derived magmas or of a magma and a single assimilant cannot be the process controlling the overall chemical variation. No simple mixing line is observed. Mixing of multiple source components would be needed, including a mantle source enriched in LIL over HFS elements. Two end-member mixing may have been a controlling factor some individual eruptions, however.

The presence of xenoliths (both exotic and of Rainier origin) and probable xenocrysts (i.e. of quartz) indicate that some assimilation has occurred. The Tatoosh and related plutons on which the Rainier cone is built and their probable more deeply emplaced cousins are potential assimilants.


Panel 11

Two-end member mixing may have produced the 2,200 year old C-tephra, the largest and most extensive tephra of Rainier origin. This unit has a consistent fine-coarse-fine grain size pattern at all sites indicating a single eruptive event. Mixing of two end-member magmas is compatible with the observed linear pattern of glass compositions and visible sample heterogeneity.


Further Study

To further constrain the petrogenesis of the Rainier lavas, additional detailed study is required. Sampling combined with detailed mapping and age dating are needed to provide improved control on the temporal and geographic relationships between compositional groups. That a series of samples is in elevation sequence does not guarantee that they are in temporal sequence. Repeated episodes of erosion between eruptions have produced inverted sequences in many locations.

Detailed study is needed to distinguish between the results of the different petrogenetic processes. This should include detailed petrography coupled with microprobe analysis of both phenocrysts and glass. Geochemical modelling including work on fractional crystallization, partial melting, and mixing/assimilation is needed. For work on mixing/assimilation, additional chemical data on xenoliths (including quenched magmatic inclusions) and local plutons is needed. Stable and radiogenic isotope is needed to help distinguish source components. Oxygen isotope analyses of crystal separates might also be used to investigate magma-H2O interaction.


Home

Copyright © 1995-1998 Stephen C. Kuehn