Effect of Subshelf Melt Variability on Sea Level Rise Contribution From Thwaites Glacier, Antarctica

• Published in: Journal of Geophysical Research: Earth Surface
• Main Authors: Hoffman, Matthew James, Asay‐Davis, Xylar, Price, Stephen F., Fyke, Jeremy, Perego, Mauro
• Language: unknown
• Published: 2021
• Subjects: 54 ENVIRONMENTAL SCIENCES; West Antarctica; Thwaites Glacier; Antarc*; Antarctica; Ice Sheet; Ice Shelf; Thwaites Ice Shelf
• Online Access: http://www.osti.gov/servlets/purl/1593115; https://www.osti.gov/biblio/1593115; https://doi.org/10.1029/2019JF005155

Modeling and observations suggest that Thwaites Glacier, West Antarctica, has begun unstable retreat. Concurrently, oceanographic observations have revealed substantial multiyear variability in the temperature of the ocean water driving retreat through melting of the ice shelf that restrains inland glacier flow. Using an ensemble of 72 ice–sheet model simulations that include an idealized representation of ocean temperature variability, we find that variable ice–shelf melting causes delays in grounding line retreat, mass loss, and sea level contribution relative to steady forcing. Modeled delays are up to 43 years after 500 years of simulation, corresponding to a 10% reduction in glacier mass loss. Delays are primarily caused by asymmetric melt forcing in the presence of variability. For the “warm cavity” conditions beneath Thwaites Ice Shelf, increases in access of warm, deeper water are unable to raise water temperatures in the cavity by much, whereas increases in access of significantly colder, shallow water reduce cavity water temperatures substantially. This leads to lowered mean melt rates under variable ocean temperature forcing. Additionally, about one quarter of the mass loss delay is caused by a nonlinear ice dynamic response to varying ice–shelf thinning rate, which is amplified during the initial phases of unstable, bed–topography–driven retreat. Mass loss rates under variability differ by up to 50% from ensemble mean values at any given time. Our results underscore the need for taking climate variability into account when modeling ice sheet evolution and for continued efforts toward the coupling of ice sheet models to ocean and climate models.