Turbulent transfer coefficients and roughness lengths over sea ice: The SHEBA results

The turbulent transfer coefficients for momentum (the drag coefficient) and for sensible and latent heat provide the means for coupling the sea ice surface to the lower levels of the atmospheric boundary layer for data analyses and in numerical models. The roughness lengths for momentum (z₀), heat (z_T), and humidity (z_Q) are related directly to these transfer coefficients but are usually easier to treat theoretically. Thus, in practice, we need the transfer coefficients; but to understand the physical processes that support the turbulent transfer, we tend to focus on the roughness lengths.

Our participation in SHEBA, the experiment to study the Surface Heat Budget of the Arctic Ocean, resulted in approximately 13,000 hours of data suitable for evaluating the roughness lengths z₀ and z_T. These data came from our main 5-level, 20-meter Atmospheric Surface Flux Group tower and from four remote portable automated mesonet (PAM) stations. Briefly, each site measured the turbulent fluxes of momentum and sensible heat directly with eddy-correlation instruments. We determined the roughness lengths by analyzing hourly averages of these fluxes in the context of Monin-Obukhov similarity theory.

On the basis of almost a year-long time series of drag coefficients from these multiple sites, we divided the SHEBA year into two aerodynamic seasons: winter, when snow was available to blow and drift, and summer, when it wasn't. Winter ran from the start of the experiment in October 1997 through 14 May 1998 and resumed on 15 September 1998. Summer was the period from 15 May through 14 September 1998. Our parameterization of z₀ for winter includes the influence of drifting snow and a contribution from the roughness of the underlying surface. In summer, we found a unique relation between the drag coefficient and the fractional area of the surface covered by leads and melt ponds. In the winter, the SHEBA z_T values corroborated a theoretical model that Andreas (1987, Bound.-Layer Meteorol., 38, 159-184) developed. In summer, sensible heat transfer is more complex because, at any time, the surface temperature ranges from 2°C in some leads and melt ponds, down to 0°C on melting ice, to even a few degrees lower when a cold snap occurs. We believe that some form of mosaic technique will, thus, be necessary for parameterizing the sensible heat flux in summer.