Atmosphere - Sea ice Exchanges and Teleconections

Over the last decades, the Arctic sea ice has experienced a drastic decline which is expected to continue. The inability of state-of-the-art climate models to capture both the rate of Arctic sea ice changes and their impact on lower latitudes is hypothesized to originate mainly from inaccuracies in the representation of atmosphere-sea ice heat exchanges. Under ASET WorkPackage 1, an extensive observational database was built from which, as part of WorkPackage 2, a novel neural-network based parameterization of turbulent surface momentum, sensible heat and latent heat fluxes was developed. This algorithm performs respectively marginally better than (for momentum), substantially better than (for sensible heat) and comparatively bad with (for latent heat) the more traditional bulk-based parameterizations on the pre-MOSAiC data (Cummins et al, 2024). Its added-value over the bulk approach emerges clearly when evaluating on the more recent MOSAiC data (Cummins et al, 2023). A fortran-based code applying this algorithm has been released publicly. Still under ASET WP2, a detailed documentation of the sources of uncertainties and their respective contributions to the uncertainties in turbulent exchange coefficients estimated from observational data, which form the basis of traditional development of bulk turbulent flux parameterizations has also been carried out. An article gathers these results as well as recommendations towards an optimal selection of data from observational campaigns for calibration (Blein et al, 2024). As part of ASET WorkPackage 3, the most recent turbulent flux parameterizations from literature have been included into the CNRM-CM6 coupled climate model used for the Coupled Model Intercomparison Project Phase 6 (CMIP6), as well as our neural-network based parameterization. Sensitivity experiments were carried out which are currently being analysed in order to guide strategic choices for the future versions of the CNRM-CM coupled model. As a complementary activity, an advanced snow scheme including a variable snow density, grain size and a multi-band albedo and accounting for processes such as blowing-snow sublimation, snow melt and metamorphism has been implemented into the ocean-sea ice component of our climate model. An article evaluating its performance in ocean-sea ice and thermodynamic-only mode is under review (Brivoal et al, 2024).