Proglacial Lakes Control Glacier Geometry and Behavior During Recession.

Ice‐contact proglacial lakes are generally absent from numerical model simulations of glacier evolution, and their effects on ice dynamics and on rates of deglaciation remain poorly quantified. Using the BISICLES ice flow model, we analyzed the effects of an ice‐contact lake on the Pukaki Glacier, New Zealand, during recession from the Last Glacial Maximum. The ice‐contact lake produced a maximum effect on grounding line recession >4 times further and on ice velocities up to 8 times faster, compared to simulations of a land‐terminating glacier forced by the same climate. The lake contributed up to 82% of cumulative grounding line recession and 87% of ice velocity during the first 300 years of the simulations, but those values decreased to just 6% and 37%, respectively, after 5,000 years. Numerical models that ignore lake interactions will, therefore, misrepresent the rate of recession especially during the transition of a land‐terminating to a lake‐terminating environment. Plain Language Summary: Lakes form at the margins of glaciers as meltwater accumulates against hillsides and behind ridges of glacier debris. Lakes at a glacier terminus are known to affect its behavior. However, glaciers terminating into such lakes are usually absent from computer simulations, and the effects of these lakes on the rates of deglaciation and on glacier behavior are poorly quantified. In this study, we tested the effect of a lake on glacier recession under two different scenarios; a land‐terminating versus a lake‐terminating glacier. We used an ice flow model called BISICLES and applied it to what was once the Pukaki Glacier in New Zealand during the end of the last ice age. We found that the presence of a lake caused the glacier to recede more than 4 times further and it accelerated ice flow by up to 8 times when compared to the same glacier that terminated on land under the same climate. Our simulated lake processes predominantly influenced the glacier over decades to centuries rather than over millennia. We suggest, therefore, that simulations of glacier evolution ignoring glacial lakes will likely misrepresent the timing and rate of recession, especially during the transition from a land‐terminating to a lake‐terminating environment. Key Points: A reverse bedslope, combined with deep water and frontal ablation, produces a terminus geometry with a wider and steeper ice surface slopeA lake‐terminating glacier receded >4 times further and flowed up to 8 times faster than a land‐terminating glacier forced by the same climateOur simulated lake processes predominantly influenced the glacier over decades to centuries rather than over millennia [ABSTRACT FROM AUTHOR]