Convective heat transfer over wintertime leads and polynyas

Leads, and polynyas are areas of open water or thin ice surrounded by thicker ice. In winter they are ideal natural laboratories for studying convective heat transfer from the ocean. First, the relevant length scales are much larger than those possible in the laboratory. Second, the large water-air temperature difference provides a wide range of unstably stratified conditions. Third, the surrounding sea ice is a very stable platform on which to place turbulence instruments. Here we analyze three data sets obtained over Arctic leads and polynyas in winter. First, we compute the bulk aerodynamic transfer coefficient for sensible heat at neutral stability appropriate at a reference height of 10 m, C-HN10. For fetches over the warm lead or polynya larger than similar to 100 m, C-HN10 has the value typically reported over the open ocean at lower latitudes, 1.00 x 10(-3). At fetches much <100 m, C-HN10 Can be as large as 1.8 X 10(-3). That is, smaller leads and polynyas transfer sensible heat more efficiently than large ones, presumably because the transfer is by mixed free and forced convection. In the second part of the study, we look at this transition from forced to free convection as parameterized by C-* in the free convection relation Nu = C*Ra1/3, where Nu is the Nusselt number and Ra is the Rayleigh number. We demonstrate that delta/L, where L is the Obukhov length and delta is the fetch-dependent height of the thermal internal boundary layer, is a useful free convection parameter. At large - delta/L, our data show that C-* approaches the often quoted limit of 0.15, For - delta/L roughly <6, a flow is exchanging heat by mixed free and forced convection, and C-* can be substantially >0.15. Because we have data over a wide enough stability range to evaluate how C-* depends on delta/L, in effect, we develop a new algorithm for computing sensible and latent heat transfer in fetch-limited, convective conditions. This algorithm could be especially useful for sensing the turbulent heat fluxes over leads and polynyas remotely because it depends only weakly on surface level wind speed, which is difficult to determine remotely over an ice-covered ocean.