Physical Controls on the Hydrology of Perennially Ice‐Covered Lakes, Taylor Valley, Antarctica (1996–2013)

Abstract

The McMurdo Dry Valleys, Antarctica, are a polar desert populated with numerous closed‐watershed, perennially ice‐covered lakes primarily fed by glacial melt. Lake levels have varied by as much as 8 m since 1972 and are currently rising after a decade of decreasing. Precipitation falls as snow, so lake hydrology is dominated by energy available to melt glacier ice and to sublimate lake ice. To understand the energy and hydrologic controls on lake level changes and to explain the variability between neighboring lakes, only a few kilometers apart, we model the hydrology for the three largest lakes in Taylor Valley. We apply a physically based hydrological model that includes a surface energy balance model to estimate glacial melt and lake sublimation to constrain mass fluxes to and from the lakes. Results show that lake levels are very sensitive to small changes in glacier albedo, air temperature, and wind speed. We were able to balance the hydrologic budget in two watersheds using meltwater inflow and sublimation loss from the ice‐covered lake alone. A third watershed, closest to the coast, required additional inflow beyond model uncertainties. We hypothesize a shallow groundwater system within the active layer, fed by dispersed snow patches, contributes 23% of the inflow to this watershed. The lakes are out of equilibrium with the current climate. If the climate of our study period (1996–2013) persists into the future, the lakes will reach equilibrium starting in 2300, with levels 2–17 m higher, depending on the lake, relative to the 2020 level.

The hydrologic and geochemical contributions from snow to streamflow in the McMurdo Dry Valleys of Antarctica

“…Our data from the 2021–2022 field season have a total potential input of 84 000 m 3 from within channel snow patches in 8 streams across Fryxell Basin (excluding Commonwealth Stream as it does not flow into Lake Fryxell). Our analysis does not include ungauged streams that flow into Lake Fryxell, or the water inputs from snow patches outside of stream channels that may melt directly into the lake, without flowing through the streams, and contribute to lake level rise, so the extra 84 000 m 3 year −1 from our analysis may explain only a fraction of the water deficit found by Cross et al (2022). However, these data show that snow has the potential to contribute significantly to lake level rise and could be an explanation for at least some of this deficit in the hydrologic model of Lake Fryxell.…”

“…the streams, and contribute to lake level rise, so the extra water deficit found by Cross et al (2022). However, these data show that snow has the potential to contribute significantly to lake level rise and could be an explanation for at least some of this deficit in the hydrologic model of Lake Fryxell.…”

“…A recent study that modelled the hydrologic inputs to Lake Fryxell from long-term MCM LTER datasets indicated a water inflow deficit of about 450 000 m 3 year À1 between the modelled lake water budget and the recorded long-term streamflow or glacier records, which is hypothesized to potentially come from snow patches melting and flowing into the lake (Cross et al, 2022). Our data from the 2021-2022 field season have a total potential input of 84 000 m 3 from within channel snow patches in 8 streams across Fryxell Basin (excluding Commonwealth Stream as it does not flow into Lake Fryxell).…”

“…Summer melting of snow patches across the landscape has been shown to provide soil moisture and solutes directly to the lakes through persistent 'water tracks' (which are thin bands of high soil moisture that route subsurface water downslope), although they have no connection to the stream channels (Gooseff et al, 2013;Levy et al, 2011). Along with this, hydrologic modelling of the Lake Fryxell Basin has shown a data gap of about 450 000 m 3 year À1 of water inputs to Lake Fryxell, which has been hypothesized to potentially (at least partly) come from snow patches across Fryxell Basin (Cross et al, 2022).…”