New Estimates Of Ice And Oxygen Fluxes Across The Entire Lid Of Lake Vostok From Observations Of Englacial Radio Wave Attenuation

Over deep Antarctic subglacial lakes, spatially varying ice thickness and the pressure-dependent melting point of ice result in distinct areas of melting and freeze-on at the ice-water interface, that is, at the lake lid. These ice mass fluxes drive lake circulation and, because basal Antarctic ice contains air clathrate, affect the input of oxygen to the lake, with implications for subglacial life. Inferences of rates of melting and freeze-on, that is, accretion rates, from radar layer tracking and geodesy are limited in spatial coverage and resolution. Here we develop a new method to estimate accretion rate, and the resulting oxygen input at a lake lid, using airborne radar data over Lake Vostok together with ice-temperature and chemistry data from the Vostok ice core. Because the lake lid is a coherent reflector of known reflectivity (at our radar frequency), we can infer depth-averaged radio wave attenuation in the ice, with a spatial resolution of similar to 1km along flight lines. Spatial variation in attenuation depends mostly on variation in ice temperature near the lid, which in turn varies strongly with ice mass flux at the lid. We model ice temperature versus depth with ice mass flux as a parameter, thus linking that flux to observed depth-averaged attenuation. The resulting map of melt and accretion rates independently reproduces features known from earlier studies but now covers the entire lid. We find that freeze-on is dominant when integrated over the lid, with an ice imbalance of 0.05 to 0.07km(3)/year, which is robust against uncertainties.Plain Language Summary Where ice meets water at the upper surface (i.e., the lid) of a subglacial lake, freezing to the lid or melting into the lake can occur. Distinct areas of freezing and melting affect lake water circulation. Furthermore, melting can introduce to the lake oxygen from ancient air trapped in deep ice. That oxygen input is important for understanding what organisms can live in the lake. Knowledge of lake lid melting and freezing has been sparse, but here we develop a new way to map melting and freezing using data from radar sounding and an ice core. We apply our method at Lake Vostok, Antarctica. We find, surprisingly, that the total amount of lid freezing exceeds that of melting. Because Antarctic ice flow carries away ice frozen to the lid, the lake would lose water, and its level would go down, unless there is some compensating inflow. The level of Lake Vostok is observed to be neither rising nor falling, so we infer that meltwater flows into the lake from surrounding ice-covered terrain. That flow carries minerals and other chemicals important for any life in the lake, so our result may change views on what can live there.