TESTING COSMOGENIC-NUCLIDE PRODUCTION-RATE SCALING MODELS: A COMPARISON OF 40AR/39AR AND COSMOGENIC 3HE AGES OF LAVA FLOWS ON EREBUS VOLCANO, ANTARCTICA

A goal of Earthtime is to precisely calibrate multiple geochronometers towards an accurate timescale for at least the Phanerozoic. Cosmogenic nuclides are valuable for this goal and widely used for the Quaternary. Although modern mass spectrometers are able to measure nuclide concentrations with 2σ uncertainties of ~5%, the accuracy of calculated exposure ages is limited by large uncertainties in the models used to scale time-integrated cosmogenic-nuclide production rates to the latitude, elevation, and exposure duration of a sample. Because most independently dated production-rate calibration sites are located in regions in which the difference between ages calculated by the two leading scaling models is minimal, it is not currently known which model most accurately reproduces production variation over the globe.

Mt. Erebus, a 3794-m-high active volcano located at 77°S in Antarctica, is a superb location for evaluating the accuracy of exposure ages because of its high cosmogenic production rates, low erosion rates, and large difference between exposure ages calculated using different scaling models. If accurate and precise independent ages can be determined, Erebus has potential to be the first production rate calibration site situated at high latitude and high elevation. A challenge at Erebus is determining accurate ⁴⁰Ar/³⁹Ar ages on anorthoclase because of melt-inclusion-hosted excess argon. Aggressive anorthoclase crushing and HF etching to remove melt inclusions, coupled with high-precision multicollector mass spectrometry, results in accurate ages for samples between 0 and 70 ka. These ages are compared to high-precision cosmogenic ³He ages of five Erebus lava flows calculated using the two leading production rate scaling models—the Lal (1991)-based “St” model of Stone (2000) and the “LSD” model of Lifton et al. (2014)—which differ by 16-21% on Erebus. Because the ⁴⁰Ar/³⁹Ar ages were determined with 2σ precision of a few hundred years to 2 ka, they are sufficient to discriminate between scaling models. The ⁴⁰Ar/³⁹Ar ages best agree with the LSD-calculated ³He ages and are significantly younger than the St-calculated ³He age. These results strongly suggest that the LSD scaling model more accurately predicts the strength of the cosmic ray flux at high latitude and high elevation.