This paper presents a sea ice prognostic model involving a one‐dimensional thermodynamic diffusion model, nudging satellite‐derived snow/ice temperatures, and two‐dimensional Lagrangian ice tracking. The aim of the model is to produce the evolvement of the physical properties of the snow and ice over the Arctic Ocean during the winter season. While the one‐dimensional column process solves the solution at a specific time and location, the evolvement of physical properties of the same ice target can be continuously simulated along the trajectory of ice movement determined by the Lagrangian tracking method. The main inputs were reanalysis‐based atmospheric forcings, thermal conditions constrained through nudging of snow skin temperature and snow‐ice interface temperature, and satellite‐derived ice motion vectors. The simulation results showed that the model can successfully reproduce well‐known regional features and geographical distributions of snow depth and ice thickness. The model‐simulated variables (i.e., snow depth, total freeboard, ice freeboard, ice thickness, and temperature) showed high correlations with the in situ or satellite measurements. In particular, the simulated temperatures were in excellent agreement with drifting buoy measurements. Since the nudging of the satellite‐derived temperature data into the model improved the thermal structure considerably, these data appear to be a key element for the successful simulation of other variables as well.

中文翻译：

---

一维热力学模型的实现，用于模拟北冰洋冰雪物理性质的冬季演变

本文提出了一种海冰预测模型，该模型包括一维热力学扩散模型，对卫星衍生的雪/冰温度进行微调以及二维拉格朗日冰跟踪。该模型的目的是在冬季产生北冰洋上冰雪物理特性的演变。虽然一维圆柱过程在特定的时间和位置解决了问题，但可以沿着拉格朗日跟踪方法确定的冰运动轨迹连续模拟同一冰目标的物理性质的演变。主要输入是基于重新分析的大气强迫，受雪皮温度和雪冰界面温度微调约束的热条件，以及卫星衍生的冰运动矢量。仿真结果表明，该模型可以成功地再现著名的区域特征以及雪深和冰厚的地理分布。模型模拟的变量（例如，积雪深度，总干舷，干冰干舷，冰厚和温度）与原位或卫星测量值具有高度相关性。特别是，模拟温度与浮标测量值非常吻合。由于将源自卫星的温度数据掺入到模型中大大改善了热结构，因此这些数据似乎也是成功模拟其他变量的关键要素。干冰，冰厚和温度）与原位或卫星测量值显示出高度相关性。特别是，模拟温度与浮标测量值非常吻合。由于将源自卫星的温度数据掺入到模型中大大改善了热结构，因此这些数据似乎也是成功模拟其他变量的关键要素。干冰，冰厚和温度）与原位或卫星测量值显示出高度相关性。特别是，模拟温度与浮标测量值非常吻合。由于将源自卫星的温度数据掺入到模型中大大改善了热结构，因此这些数据似乎也是成功模拟其他变量的关键要素。