Arctic Sea Ice Variability in the 20th and Early 21st Centuries: Synthesis of Observations and Models

卫星前海冰记录的不确定性很大程度上源于难以根据船舶日志、侦察飞行等历史观测建立覆盖整个北极的海冰可变性的一致记录。这些观测在空间和时间上分散，根据这些记录绘制海冰可变性代表性图景的主要挑战是对北极海冰的强季节性周期和高度空间变异性进行充分采样。因此，海冰的重建必须依赖于时间和空间的内插，而这一点由于海冰的动态性质而变得困难。对于一些至关重要的气候变量，如海冰厚度和北极海冰总量，没有历史数据。PIS将进行世纪尺度的北极海冰再分析(CASIR)，它综合了从1872年到现在的观测和模型，以产生北极海冰范围、密度、厚度和体积的长期记录。大气再分析、海冰模拟和历史数据恢复方面的进展使CASIR成为可能。为了建立CASIR，他们将恢复和收集早在19世纪末的新的和现有的历史海冰观测数据，并将它们整合到基于泛北极冰洋模拟和同化系统(PIOMAS)的模式验证和同化框架中。这将包括从税务局/海岸警卫队的历史船舶日志中提取海冰变量，这些日志目前正在转录，并有望填补北极太平洋地区的重要空白。通过数据同化和模式校准和验证，观测记录将被整合到PIOMAS中，以生成CASIR。CASIR的大气强迫数据集合将从已经完成或目前正在进行的20世纪大气再分析项目中提取。除了与观测值进行比较外，这一集合还将用于确定CASIR海冰变量中的不确定性。PIS将分析CASIR的冰层范围、浓度、厚度和体积，以表征20世纪的时空变异性，提高对这种变异性驱动机制的理解，并将最近的海冰下降放在更长期的背景下。

Much of the uncertainty in the pre-satellite sea ice record stems from the difficulty establishing a consistent record of sea ice variability that covers the entire Arctic from historical observations such as ship logs, reconnaissance flights, etc. These observations are scattered in space and time and the primary challenge in constructing a representative picture of sea ice variability from these records is to adequately sample the strong seasonal cycle and high spatial variability of sea ice in the Arctic. Sea ice reconstructions must therefore rely on interpolation in time and space, which is made difficult by the dynamic nature of sea ice. For some critically important climate variables, such as sea ice thickness and total arctic sea ice volume, no historical data exist. The PIs will conduct a Century-scale Arctic Sea Ice Reanalysis (CASIR) that synthesizes observations and models from 1872 to the present to produce a long-term record of arctic sea ice extent, concentration, thickness, and volume. Advances in atmospheric reanalysis, sea ice modeling, and recovery of historical data make CASIR possible. To construct CASIR, they will recover and collect new and existing historical sea ice observations as far back as the late 1800s and integrate them into a model validation and assimilation framework based on the Pan-arctic Ice-Ocean Modeling and Assimilation System (PIOMAS). This will include the extraction of sea ice variables from historical Revenue Service/Coast Guard ship logs that are now being transcribed and promise to fill important gaps in the Pacific sector of the Arctic. Through data assimilation and model calibration and validation, the observational record will be integrated into PIOMAS to generate CASIR. An ensemble of atmospheric forcing data for CASIR will be extracted from atmospheric reanalysis projects for the 20th century that are completed or currently underway. This ensemble, in addition to comparisons with observations, will be used to determine uncertainties in CASIR sea ice variables. The PIs will analyze CASIR ice extent, concentration, thickness, and volume, to characterize 20th century temporal and spatial variability, improve understanding of the mechanisms driving this variability, and place the recent sea ice decline in a longer term context.