VOLCANIC GLASS AS A METEORIC WATER PROXY: DETERMINING HYDROGEN ISOTOPE FRACTIONATION FROM THE MAZAMA ASH IN WESTERN NORTH AMERICA

Plate tectonics creates and alters topography at convergent margins through orogenesis. High mountains in turn control climate and precipitation patterns. Topography controls the isotopic composition of meteoric water by driving air masses to rise, adiabatically cool, and condense, where heavier mass isotopes preferentially move into the condensate. To reconstruct past topography, many researchers use proxies for precipitation that have incorporated ancient meteoric water into their structures. The hydrogen isotope composition (δD) of meteoric water directly reflects the amount of isotopic distillation and thus past topography. Hydrated volcanic glass reflects the δD of meteoric water near the time of ash deposition. The glass and water phases may have different δD values, however, if there is a preference for one isotope over the other as the glass hydrates. We aim to establish a temperature-dependent measure of the magnitude of the isotopic fractionation when water diffuses into glass (fractionation factor). A new water-glass fractionation factor is needed as currently used estimates are based on small datasets and do not include temperature. Our temperature-dependent fractionation factor will allow for accurate comparisons of the glass proxy to any other isotope proxy and account for past climate change.

The Mazama ash, erupted approximately 7,600 years ago, blanketed much of the western North America and provides an ideal proxy for meteoric waters similar to modern soil waters. By measuring the δD of soil water and the δD of volcanic glass from a single ash layer, and the mean annual temperature at the depth of sampling, we will determine the fractionation factor at a range of locations with variable elevations and mean annual temperatures. In 2015, ash and soil water samples were collected and temperature loggers installed across the northwest. 34 samples are currently being separated, and soil waters are being extracted via vacuum line. Both glass and liquid samples will be analyzed on a TC/EA – MAT 253 IRMS. Preliminary results show that higher elevations tend to correlate with higher δD calibrated values. For instance, a sample from British Columbia has a mean δD value of -74‰ at an average elevation of 601.5 m, while a sample in south central Alberta had a mean δD value of -139‰ at an elevation of 1337 m.