Exploring ice-ocean boundary layer dynamics in climate models, idealized simulations, and outdoor lab experiments

Abstract

Everything that happens in the Arctic Ocean, be it of physical, biological, or chemical nature, is constrained by the vertical distribution of heat and salt. In this talk, I will share recent results and on-going work aimed at examining questions directly related to vertical mixing below sea ice: (1) How accurately are the physical properties of the Canada Basin simulated in climate models? (2) How do observed changes to the size and speed of a sea ice floe and ocean stratification impact ocean mixing in 2D numerical simulations? (3) Can we, for the first time, examine seasonal ice-ocean boundary layer dynamics in a 20 m × 10 m × 3 m outdoor saltwater pool?Our results indicate that the majority of climate models do not accurately simulate the surface freshening trend observed in the Canada Basin between 1975 and 2006-2012, nor do they simulate heat from Pacific Water in the same region. We suggest that both of these biases can be partly attributed to unrealistically deep vertical mixing in the models. We next explore one possible source of this model bias related to decadal changes to the underside of ice floes, called ice keels. Results from idealized numerical simulations highlight the importance of ice keel depth, which controls the range over which ocean mixing occurs, as well as ice keel speed and ocean stratification. Further, we estimate that observational uncertainties related to ice keel depth may translate into uncertainties in the sign of current and future changes to below-ice momentum transfer into the ocean. Lastly, we present the instrument setup for our 2022-2023 pilot experiment and on-going outreach work at the Sea-ice Environmental Research Facility (SERF) in Canada. This is a unique facility centres around an outdoor saltwater pool where sea ice evolves under natural atmospheric conditions in a semi-idealized and well-instrumented setting.