Evaluation of snow/ice albedo parameterizations and their impacts on sea ice simulations

Abstract

Abstract:Climate models use a variety of snow/ice albedo parameterizations for the ice covered ocean. In this study, we applied in situ measurements (surface temperature, snow depth and ice thickness) obtained from the Surface Heat Budget of the Arctic Ocean (SHEBA) as input values to four different snow/ice albedo parameterizations (representing the spectrum of parameterizations used in stand-alone sea ice models, numerical weather prediction and regional climate models of the Arctic Basin, and coupled global climate models), and evaluated the parameterized albedos against the SHEBA observed albedo. Results show that these parameterizations give very different representations of surface albedo. The impacts of systematic biases in the input values on the parameterized albedos were also assessed. To further understand how sea ice processes are influenced by differences in the albedo parameterizations, we examined baseline sea ice characteristics and responses of sea ice to an external perturbation for the simulations of the albedo parameterizations using a stand-alone basin-scale dynamic/thermodynamic sea ice model. Results show that an albedo treatment of sufficient complexity can produce more realistic basin-scale ice distributions, and likely more realistic ice responses as climate warms.

Albedo of coastal landfast sea ice in Prydz Bay, Antarctica: Observations and parameterization

“…This suggests that albedo and ice thickness are not correlated significantly when the ice is thicker than the optical thickness of ice. The result is consistent with former studies (Barry, 1996;Curry et al, 2001;Liu et al, 2007). The correlations between albedo and surface temperature in 2010 and 2011 were −0.27 and −0.30, respectively. However, this negative correlation between albedo and surface temperature was very weak before 1 November, when surface temperature was far below the melting point; the correlations in 2010 and in 2011 were 0.42 and −0.10, respectively.…”

“…Following Liu et al (2007), four existing snow/ice albedo parameterizations ranging from simple to complex were evaluated in this study. The parameterization of Parkinson and Washington (1979) only considers broadband albedo for snow and bare ice (PW79).…”

“…Besides weighting snow and sea ice albedos with snow thickness, the albedo scheme used in CCSM3 further distinguishes the visible and near-infrared albedos (Briegleb et al, 2004). For a detailed description of these parameterizations, see Parkinson and Washington (1979), Lynch et al (1995), Dethloff et al (1996), Briegleb et al (2004) and Liu et al (2007). Figure 4 shows the observed and simulated albedos during the measurement period.…”

“…Currently, global climate models use a number of snow/ sea-ice albedo parameterizations with different complexities (e.g., Barry, 1996;Curry et al, 2001;Liu et al, 2007). However, sea-ice albedo parameterizations have been largely determined and validated with observations that have been collected in the Arctic (e.g., Curry et al, 2001;Liu et al, 2007), and little is known about how well these parameterizations perform for Antarctic sea ice (Weiss et al, 2012), especially the complex parameterizations that consider snow thickness and ice thickness.…”

“…However, sea-ice albedo parameterizations have been largely determined and validated with observations that have been collected in the Arctic (e.g., Curry et al, 2001;Liu et al, 2007), and little is known about how well these parameterizations perform for Antarctic sea ice (Weiss et al, 2012), especially the complex parameterizations that consider snow thickness and ice thickness. Thus, continuous time series of the Antarctic sea ice albedo are required to quantify the influence of different factors on variations in albedo, and provide a dataset for evaluating and developing Antarctic sea-ice albedo parameterizations.…”