The temperature dependence of frost flower growth on laboratory sea ice and the effect of the flowers on infrared observations of the surface


  • Seelye Martin,

  • Yanling Yu,

  • Robert Drucker


This paper describes a laboratory study of frost flower growth on young sea ice at different temperatures and the effect of these flowers on the surface temperature observed with an infrared radiometer. The flowers grew on sea ice which formed in a salt water tank at room temperatures of −20, −24, and −30°C, with an additional experiment at −16°C, where no flowers appeared. The growth habit and height of the observed crystals depended on the existence of a region of supersaturated vapor adjacent to the surface and on the range of temperatures in the surface boundary layer. The source of the surface brine from which the flowers grew was probably brine transport within the ice toward the cold upper surface driven by the thermomolecular pressure gradient. The evaporation of vapor from this liquid into the atmospheric boundary layer provided the supersaturated region adjacent to the ice surface. Two kinds of flowers were observed; at −20 and −24°C, dendritic crystals grew approximately between the −12 and −16°C isotherms, and, at −30°C, rod-like flowers appeared between −16 and −25°C. These limits correspond to earlier work on crystal growth from the vapor. In each case, the maximum flower height approximately equaled the height of the isotherm corresponding to the colder temperature limit for each crystal type, −16°C for the dendrites and −25°C for the rods. The effect of the flowers on the radiometer surface temperature was as follows: because the flowers protrude 10–20 mm above the surface into the boundary layer, the infrared temperature of the flower-covered ice was about 4–6°C colder than that of the same ice cleared of flowers. We also found that the insulating effect of the flowers caused the ice surface temperature beneath the flowers to be 1–2°C warmer than the surrounding bare ice. The importance of the flower growth is that infrared satellite observations of thin ice in winter will be colder than the actual surface temperature, which may account for the absence of warm young ice in infrared satellite images.