Albedo parameterization is of fundamental importance for accurate representation of high-latitude climate variability by modeling studies. Field observations show that near-infrared snow albedo decreases dramatically when surface air temperature exceeds −2 °C. This can influence reproduction of sea ice simulations taking into consideration the importance of the drastic change in albedo in early melt season for the seasonal change of sea ice extent. Therefore, we conducted global ocean data-assimilative simulation experiments using a modified snow albedo parameterization. The modified parameterization reduced the albedo directly and achieved a comparable indirect reduction via changes in the modeled snow and sea ice distributions (ice–albedo feedback). As a result, sea ice thickness was reduced by more than 0.4–1 cm over most of the central Arctic Ocean. Sea ice velocities were also reduced by enhanced ocean drag with weakened surface ocean circulation in the Beaufort Gyre. In the Southern Ocean, the modified parameterization caused snow thicknesses to be decreased by up to 2 cm in the Weddell Sea. These impacts, which were generally larger than the spread of ensemble experiment results and therefore robust, at least in our model, provide useful information for quantifying the results of albedo modification in climate modeling studies.

反照率参数化对于通过建模研究准确表示高纬度气候变化至关重要。现场观察表明，当地面气温超过-2°C时，近红外雪反照率会急剧下降。考虑到早融化季节反照率的急剧变化对于海冰范围的季节性变化的重要性，这可能会影响海冰模拟的再现。因此，我们使用修改后的雪反照率参数化进行了全球海洋数据同化模拟实验。修改后的参数化直接减少了反照率，并通过模拟的雪和海冰分布的变化（冰–反照率反馈）实现了类似的间接减少。结果，海冰厚度减少了超过0。北冰洋中部大部分地区4-1厘米。Beaufort Gyre的海洋阻力增加，地表海洋环流减弱，海冰速度也降低了。在南大洋，修改后的参数设置导致韦德尔海的积雪厚度减少了2 cm。这些影响通常大于整体实验结果的传播范围，因此至少在我们的模型中是可靠的，它们为量化气候模拟研究中反照率修改的结果提供了有用的信息。