Resolving atmospheric contaminants in mantle noble gas analyses

[1] As part of an ongoing initiative to constrain mantle noble gas budgets, we evaluate both recently published high-quality noble gas data and report new data in order to rationalize the use of threedimensional (3-D) modeling techniques. Modeling of these data shows that traditional 2-D mixing plots are not adequate tools to fully constrain mixing systems and mantle end-member compositions. We show that these mantle noble gas analyses are compromised to varying degrees by the addition of a fractionated atmospheric contaminant, irrespective of eruption setting. This component can also be present in analyses with mantle-like neon or argon isotope ratios. Previous estimates for the mantle end-member 40Ar/ 36Ar and 129Xe/ 130Xe compositions of most of these samples are likely to be minimum values. In fact it may well be that there are no analyses in the present literature that have completely unadulterated mantle volatiles present. © 2003 by the American Geophysical Union.