冰间湖是极地冰区中尺度物理海洋现象，是海-冰-气间物质与能量交换的重要场所，具有独特的热力学效应，表现为高热通量、冰产量和盐产量等，对极地气候和生态系统有重要影响。冰间湖对极地冰冻圈系统的微弱变化较为敏感，其位置、面积变化是气候变化和极地海洋环境变化的指示器。对冰间湖开展长期监测并探究其与极地环境变化的相互作用机制，对深入了解极地气候变化和极地海-冰-气系统相互作用等具有重要意义。在全球变暖和北极放大作用影响下，北极海冰快速减少，冰间湖的热力学效应在北极放大效应中的作用尚不明确，冰间湖的长时序时空变化特征尚不清楚，冰间湖对全球变暖的响应研究也有待完善。

针对这一前沿科学问题，本文以获取长时序北极冰间湖时空动态变化特征为出发点，以冰温数据和海冰密集度数据为基础，评估了热力学遥感反演冰间湖薄冰厚度的不确定性，发展了基于海冰厚度和海冰密集度的冰间湖范围自动提取算法，创新性地研制了2002/03 ~ 2021/22年11 ~ 3月全北极逐日1 km分辨率的冰间湖数据集，该数据集包括冰间湖范围、薄冰厚度、热损失量、冰产量和盐产量等参数。在此基础上，分析了近二十年北极40个冰间湖范围和热力学效应的时空变化特征，揭示了北极冰间湖范围变化与极地环境变化的关联。主要内容和结论如下：

（1）系统地评估了冰间湖薄冰厚度反演模型的不确定性，优化了模型参数，提高了模型反演精度。薄冰厚度是冰间湖的关键参数之一。本文基于经典一维热力学海冰模型反演海冰厚度，检验了模型冰厚对不同参数化方案的敏感性，估计了模型冰厚对输入变量不确定性的响应。模型参数化变量包括大气发射率、空气密度、空气比热、海冰蒸发潜热、海冰热传导率、雪厚和雪传导率。模型输入变量包括空气温度、海冰表面温度和风速。研究发现，在参数化变量中，雪传导率和雪厚方案引起的冰厚偏差最大，其次是海冰热传导率，这些方案对冰厚的敏感性也较高。使用现场测量冰厚对各参数化方案所得冰厚进行验证，得到一个相对较优的参数化组合方案，提高了模型精度。使用实测气象数据验证发现，模型输入的空气温度和表面温度不确定度较高。根据输入数据误差的概率分布，使用其扰动来评估由此产生的模型冰厚的不确定度。结果表明，气温、冰温和风速对模型反演冰厚的不确定性的贡献分别约为0.09 m、0.049 m和-0.005 m。

（2）优化了基于冰厚的冰间湖范围提取算法，创建了基于海冰密集度的冰间湖范围提取算法，研制了2002/03 ~ 2021/22年11 ~ 3月全北极逐日冰间湖数据集。本文采用MODIS冰温和ERA5气象数据作为一维热力学海冰模型输入，反演逐景的薄冰厚度，结合冰间湖特征重构和云处理等方法，优化了基于冰厚数据的逐日冰间湖范围自动提取算法，与前人研究结果对比表明本方法的提取结果有效可靠。采用基于像素结构的分区迭代膨胀法，建立了基于海冰密集度的全北极冰间湖范围自动提取算法，分层随机抽样验证精度达98%。基于已有的模型公式和冰厚数据，反演了冰间湖热力学参数，建立了长时序、高时空分辨率的北极冰间湖数据集，包括冰间湖范围（6.25 km 和 1 km）、冰厚、冰温、热损失量、冰产量、盐产量和高盐陆架水产量（1 km）等。

（3）开展了北极冰间湖范围和热力学参数的时空变化特征分析研究。利用本文提取的2002/03 ~ 2021/22年11 ~ 3月全北极逐日冰间湖数据集，分析了北极40个主要冰间湖的范围、冰厚和热力学参数在年际、季节和逐日尺度的时空变化特征及趋势。主要发现包括：一是北极冰间湖的稳定性在降低，主要表现为近五年冰间湖范围在加速增长、范围的波动幅度显著增加、发生在多年冰区的冰间湖事件增多以及冰间湖正异常频次在近十年翻了5倍等；二是北极冰间湖在极地环境系统中的热力学作用增强，热力学产值占整个北极薄冰区的50%以上，平均每年增加约0.8%，近三年来该比例显著增加，这与11月和1月东北极冰间湖冰厚增加高度相关；三是北极40个冰间湖时空变化特征具有明显的区域性，可以分为三个区，包括边缘冰区型冰间湖（Marginal Sea Polynya, MP）、东北极冰间湖（East Arctic Polynya, EP）和西北极冰间湖（West Arctic Polynya, WP）。对北极冰间湖范围和热力学产值增加贡献最大的来自EP，其次是WP，而MP却呈减少趋势。

（4）分析了北极冰间湖变化与极地气象环境的关系，揭示了北极冰间湖对气候变化的响应。本文详细分析了北极冰间湖区域气温、冰温、温差、风速、风向和海平面气压等气象因子的时空变化特征，探究了气象要素与北极冰间湖范围变化的相关性。结果显示，近十年冰间湖范围与气温的决定系数由0.17增加至0.45，气温变化与北极冰间湖范围变化的关联在加强。冬季MP范围与气温高度负相关，其范围先增加后下降呈∩ 型，与气温的变化趋势恰好相反；EP和WP的范围在初冬（11 ~ 12月）与温度正相关，在深冬（1 ~ 3月）与风场的相关程度加强，这使得其范围在整个冬季呈下降趋势。对冰间湖异常事件的统计分析发现，风向在冰间湖异常事件中具有重要作用，正异常事件发生时的北极气压场分布，使得北极大部分海域沿岸的盛行风向利于当地冰间湖的发展，负异常时相反。此外，分析了地形等固有因子对冰间湖分布的影响，结果表明地形的走向和坡度影响冰间湖边界的走向和位置，有93.26%的北极冰间湖分布在水深500 m以浅的海域。

Polynya is a mesoscale physical oceanography phenomenon in the polar ice region, which is also a main material and energy exchange place between sea water, sea ice, and the atmosphere. It has an important effect on the polar climate and ecosystem with special thermodynamic effects, such as high heat flux，ice production and salt production. Polynya’s locations and extent changes are sensitive indicators of climate change and polar marine environment change. Long-term monitoring of Arctic polynya and their interaction with polar environment factors is of great significance to the understanding of polar climate change and the interaction of the polar sea-ice-air system. Under the influence of global warming and Arctic amplification, Arctic sea ice is rapidly decreasing, and the role of the thermodynamic effects of polynya in Arctic amplification is still not clear. The long-term spatial-temporal variation characteristics of polynya are waiting to be revealed, and the study of the response of Arctic polynya to global warming also needs to be improved.

Focusing on such a frontier topic, the present thesis sets the goal of obtaining the long-term spatial-temporal variations of Arctic polynyas. Based on the thermal sea ice temperature data and sea ice concentration (SIC) data, this thesis evaluated the uncertainty of thin ice thickness (TIT) retrieval and developed automatic polynya extent retrieval algorithms for TIT and SIC data, respectively. Based on those algorithms, a daily 1 km Arctic polynya dataset was innovatively developed for the entire Arctic from November to March between 2002/03 and 2021/22. This dataset includes polynya extent, TIT, heat loss, ice production, and salt production. Based on this dataset, the spatial-temporal variations of 40 Arctic polynya extents and thermodynamic effects in the last 20 years are analyzed, and the correlation between the Arctic polynya extents and polar environmental factors is studied. The main contents and innovative conclusions are as follows:

(1) Evaluation and optimization of thin ice thickness retrieval model. TIT is one of the key parameters of polynya. Retrieval of thin-ice thickness using thermodynamic modeling is sensitive to the parameterization of the independent variables and the uncertainty of the measured input variables. This thesis examines the deviation of the classical model’s TIT output when using different parameterization schemes and the sensitivity of the output to the ice thickness. Moreover, it estimates the uncertainty of the output in response to the uncertainties of the input variables. The parameterized independent variables include atmospheric longwave emissivity, air density, specific heat of air, latent heat of ice, conductivity of ice, snow depth, and snow conductivity. Measured input parameters include air temperature, ice surface temperature, and wind speed. Among the independent variables, the results show that the highest deviation is caused by adjusting the parameterization of snow conductivity and depth, followed ice conductivity. The sensitivity of the output TIT to ice thickness is highest when using parameterization of ice conductivity, atmospheric emissivity, and snow conductivity and depth. The retrieved TIT obtained using each parameterization scheme is validated using in situ measurements and satellite-retrieved data. From in situ measurements, the uncertainties of the measured air temperature and surface temperature are found to be high. The resulting uncertainties of TIT are evaluated using perturbations of the input data selected based on the probability distribution of the measurement error. The results show that the overall uncertainty of TIT to air temperature, surface temperature, and wind speed uncertainty is around 0.09 m, 0.049 m, and -0.005 m, respectively.

(2) A dataset of daily Arctic polynya extent and its thermodynamic effects was generated from November to March between 2002/03 and 2020/21. Based on the thermodynamic sea ice model, this thesis used MODIS ice temperature and ERA5 meteorological data as model inputs to retrieval TIT. Then combined the polynya feature reconstruction method and cloud processing method to realize the automatic retrieval of daily polynya extent. By comparing with previous research results, the validity of our retrievals is verified. In addition, based on the SIC data, a fast automatic retrieval algorithm for the entire Arctic polynya extent was established by using the pixel-based partition iterative expansion method. The accuracy was validated based on stratified random sampling method with an accuracy up to 98%. Based on the existing formulas and ice thickness data, the values of thermodynamic parameters of Arctic polynya were evaluated. Furthermore, the long-term Arctic polynya data set with high spatial-temporal resolution was established, including polynya extent (6.25 km and 1 km), ice thickness, ice temperature, heat loss, ice production, salt production and high salt shelf water production (1 km).

(3) The spatial-temporal variations of the Arctic polynya extent and the its thermodynamic effects are systematically analyzed. Based on the above daily data set, the variation characteristics and trends of the extent and thermodynamic parameters for 40 Arctic polynyas at interannual, seasonal and daily scales were analyzed. The main findings are as follows: first, the stability of the Arctic polynya is decreasing, which is mainly manifested as 1) the accelerated growth of the polynya extent in the past five years, 2) the significant increase of the extent fluctuation, 3) the increase of the polynya events in the multi-year ice area, and 4) the frequency of the positive anomaly of the polynya in the past ten years. Secondly, the thermodynamic effect of Arctic polynyas in the polar environmental system is enhanced. The polynya thermodynamic productions account for more than 50% of the whole Arctic thin ice area, with an average annual increase of about 0.8%. In the past three years, this proportion has increased significantly, which is highly correlated with the increase of ice thickness in the east Arctic polynya in November and January. Thirdly, the spatial-temporal variations of the 40 Arctic polynyas have obvious regional characteristics, which can be divided into three regions, i.e., marginal sea polynya (MP), east Arctic polynya (EP) and west Arctic polynya (WP). The EP contributed the most to the increase of the Arctic polynya extent and thermodynamic production, followed by WP, while MP extent showed a decreasing trend.

(4) The relationships between the Arctic polynya and the meteorological and topographical factors are analyzed. The spatial-temporal variations of meteorological factors such as air temperature, ice temperature, temperature difference between them, wind speed, wind direction and sea surface level pressure in the Arctic polynya are analyzed in detail, and the correlation between meteorological factors and the variation of the Arctic polynya extent is explored. The results show that the correlation between air temperature change and Arctic polynya extent is increasing from 0.17 to 0.45 in the last decade. The MP extent in winter is negatively correlated with the air temperature, and its range increases first and then decreases in a ∩-shaped pattern, which is opposite to the trend of air temperature. The ranges of EP and WP were positively correlated with temperature in early winter, and were more correlated with wind field in late winter, which made the ranges of PE and PW decrease in winter. The statistical analysis of the polynya anomaly events shows that the wind direction plays an important role in the polynya extent anomaly events. The distribution of the Arctic air pressure field during the occurrence of the positive anomaly events makes the prevailing wind direction favorable to the development of the local polynya in most of the Arctic sea coasts. In addition, the inherent influence of topography on the extent of polynya is analyzed. The results show that the strike and slope of topography affect the strike and location of the polynya boundary and 93.26% of Arctic polynyas are distributed in the shallow area with a depth less than 500 m.