Annual variability of heat in the Arctic mixed layer beneath pack ice from drifting buoy observations

The oceanic heat flux (Fw) from the ocean to the ice in the Arctic, the product of heat and turbulence at the ice-ocean interface, is a sensitive but poorly understood component of the heat budget in climate studies. Although models indicate that the equilibrium mean thickness of the ice pack is sensitive to small changes in annual average Fw, the sparseness and variability of direct observations has made it difficult to produce credible regional estimates at annual and longer timescales. In this investigation, we compare observations of heat in the Arctic mixed layer and dynamics at the ice-ocean interface with parameterizations and climatologies in order to obtain a better understanding of the large-scale structure and temporal variability of Fw. Long term drifting platform observations of temperature and salinity between 8 and 12 m (primarily from SALARGOS and IOEB buoys) are used to describe the annual cycle of temperature above freezing (Taf) in the mixed layer beneath Arctic pack ice between 1975 and 1998, and Fw is estimated by modulating the observed Tafs with ice-ocean friction velocities (u*) determined from the platform drifts, according to the McPhee (1992) relationship. In the Transpolar Drift, Taf is not negligible in winter, which implies a positive Fw to the ice pack by means other than solar heating. In the Beaufort Gyre, variability of Taf (and Fw) between different years is apparent and sometimes not negligible in winter. A parameterization based solely on the solar zenith angle (with a 1 month lag) is found to largely describe the observed Tafs (with root-mean-square error less than 0.05 °C), despite the lack of an open water term. Correlations between the observed annual Tafs and the parameterization are high (median R²=0.75), compared to Tafs determined from a hydrographic dataset based on the US-Russian EWG Atlas (median R²=0.16). Deviations of observed Tafs from the parameterization cannot consistently be explained by local open water fraction anomalies (determined from satellite ice concentration data), but are likely due to heat advected horizontally, or entrained from below the halocline (such as from synoptic storms). A monthly Fw "climatology" from 1979 to 1998 is produced by modulating parameterized Tafs with u* based on monthly ice drift averages from the IABP, also using the McPhee (1992) equations. Correlations are moderate between the derived climatology and Fw estimates from the drifting observations (median R²=0.45), with deviations predominantly due to errors associated with u* which are enhanced by the nonlinearity of the heat flux relationship. In the derived climatology, the interannual variations in Taf are fixed by the parameterization, but the dynamics cause an overall positive trend in average Arctic Fw since 1981.

Reference: McPhee, M.G., "Turbulent heat flux in the upper ocean under sea ice," J. Geophys. Res., 97, 5365-5379, 1992.