Announcing the SHEBA bulk turbulent flux algorithm

Mesoscale models and global climate models that are used over sea ice must couple the lower atmosphere to the sea ice through flux boundary conditions. Modelers often refer to the code that makes this connection as a flux coupler. In our jargon, we make the connection with a bulk flux algorithm. In most sea ice and atmospheric models, this flux coupler is a potpourri of equations, parameterizations, and algorithms selected from many sources. In contrast to this standard piecemeal approach, we describe here the first bulk turbulent flux algorithm for coupling the atmosphere with sea ice regions that has been developed from a large data set and comprehensively tested as a unified package against flux data. We base our algorithm on data from SHEBA, the experiment to study the Surface Heat Budget of the Arctic Ocean. SHEBA yielded roughly 18,000 hours of turbulent flux data from multiple sites over Arctic sea ice.

The SHEBA year presented two aerodynamic seasons: winter and summer. We therefore developed two complementary versions of the flux algorithm. One version treats winter, when the sea ice is compact and snow-covered and the snow is dry enough to blow and drift. The second version treats summer, when the snow is wet and sticky and thus does not drift; later in summer, the snow melts entirely to expose bare ice, and extensive open water covers the surface in the form of melt ponds and leads. We also recognized that summer sea ice behaves aerodynamically like the marginal ice zone; our summer algorithm thus also works in marginal ice zones in any season.

Components of our algorithm include new expressions for the roughness length for wind speed (z₀) over winter and summer sea ice, validated expressions for the roughness lengths for temperature (z_T) and humidity (z_Q), new expressions for the profile stratification corrections in stable stratification, and tested expressions for the effective wind speed in very light winds in both stable and unstable stratification. To demonstrate our algorithm, we will compare its predictions with measurements of the friction velocity and the sensible and latent heat fluxes made during both winter and summer.