WATER CONCENTRAIONS IN TUYA BASALTIC GLASS: IMPLICATIONS FOR OVERLYING ICE THICKNESS

Thickness of overlying ice sheets can be calculated from the volatile contents of basaltic glass preserved in subglacial basaltic lava flows (Tuffen et al., 2010). Tuyas are the edifices that remain from a subglacial eruption once the ice sheet has retreated. We have attempted to calculate the minimum ice thickness under which Hlöðufell tuya (south-central Iceland) erupted, and have investigated constraints on the quality of data that can be expected from this technique.

Hlöðufell is a Holocene tuya located in the Southwest Rift Zone of Iceland, 9 km south of Langjokull ice cap. Dissolved H₂O and CO₂ concentrations in glass rinds from pillows from Hlöðufell have been determined using Fourier transform infrared (FTIR) spectroscopy. Ice thickness was calculated on the basis of the pressure dependence of solubility of water and carbon dioxide in basaltic melt. Multiple measurements within individual samples from various locations, but at similar elevations were taken to test the assumption of homogeneity of volatile concentration in glass.

Average H₂O concentrations range from 0.08 to 0.19 wt. %. These small concentrations of H₂O could be dissolved in basaltic magma at very low pressures, which would indicate a very thin ice-cap. Hlöðufell erupted ~ 8.2 ka (Liccardi, 2007). By 8.7 ka, glaciers in Iceland had retreated to nearly their present locations. Hence, Hlöðufell would have been situated at or near the edge of the ice-sheet. So, it is possible that the ice cover under which Hlöðufell erupted was relatively thin (<50 m thick). Alternatively, measured H₂O concentrations might be low because either because a) original H₂O was lost through vesiculation or b) the Hlöðufell basaltic melt hosted very little water initially, so that the melt was not water-saturated. The presence of vesicles indicates oversaturation of volatiles (Tuffen et al., 2010). Vesicles are present in the Hlöðufell samples, indicating that the melt was most likely volatile-saturated. So, measured water concentrations may be low due to water loss through vesiculation, or the data may indicate that ice cover was indeed quite thin. We will next measure H₂O in melt inclusions, which ideally retain original water concentrations of the melt from which the host mineral crystallized.