2015-2019年青藏高原、西伯利亚、阿拉斯加河湖区30m分辨率湖冰类型数据集

湖冰是冰冻圈的重要参数，其变化与气温、降水等气候参数密切相关，而且可以直接反映气候的变化，因此是区域气候参数变化的一个重要的指标，但由于其研究区往往位于自然环境恶劣，人口稀少的区域，大规模的实地观测难以进行，因此利用哨兵1号卫星数据，以10m的空间分辨率和优于30天的时间分辨率对不同类型的湖冰变化进行了监测，填补了观测空白。利用HMRF算法对不同类型的湖冰进行分类，通过时间序列分析三个极区中部分面积大于25km2的湖泊的不同类型湖冰的分布，形成湖冰类型数据集，可以获得这些湖泊不同类型湖冰的分布，数据包括了被处理湖泊的序号，所处年份及其在时间序列中的序号等信息，矢量数据集包括采用的算法，所使用的哨兵1号卫星数据，成像时间，所处极区，湖冰类型等信息，用户可以根据矢量文件确定时间序列上不同类型湖冰的变化。

Lake ice is a critical parameter of the cryosphere. Its changes are closely correlated with climatic parameters such as air temperature and precipitation, and can directly reflect climatic variations, making it an important indicator of regional climatic parameter changes. However, since most study areas are located in regions with harsh natural environments and sparse populations, large-scale field observations are difficult to conduct. Therefore, Sentinel-1 satellite data were utilized to monitor changes in different types of lake ice at a spatial resolution of 10 m and a temporal resolution better than 30 days, filling the observational gap. The Hidden Markov Random Field (HMRF) algorithm was applied to classify different types of lake ice. Through time-series analysis of the distribution of different lake ice types in some lakes with an area greater than 25 km² across three polar regions, a lake ice type dataset was developed. This dataset enables acquisition of the distribution of various lake ice types for these lakes: the attribute data includes the serial number of each processed lake, the corresponding year, and its serial number within the time series, while the vector dataset contains information such as the adopted algorithm, the utilized Sentinel-1 satellite data, imaging time, the corresponding polar region, and lake ice type. Users can determine the changes of different lake ice types over the time series based on the vector files.