基于Sentinel-1卫星的北冰洋海冰信息提取

Information Extraction of Arctic Sea Ice Based on Sentinel-1 Satellite

为研究北极海冰分布、形态特征的变化规律,基于Sentinel-1卫星的SAR数据,对弗雷姆海峡区域的海冰进行长时间序列的提取及变化分析。对试验影像预处理,利用K-均值聚类算法提取海冰。基于浮冰面积大小,将浮冰进行分类,研究各参数与海冰破碎物理过程的关系,海冰分布以密集度体现,海冰的形态特征以面积、周长、圆度等参数描述。结果表明:利用高分辨率SAR数据,能获取精细的海冰分布信息,将其与传统的海冰密集度产品结合,可构造信息更丰富的多尺度冰情图,为极地海冰监测提供支持,对维护北极航道安全、分析北冰洋海冰气候效应等有重要意义。

In order to study the distribution and morphological characteristics of Arctic sea ice over time,a long time serial images from Sentinel-1 satellite are used to study the extraction and change of sea ice in the Fram channel region in this paper.After the image preprocessing,the sea ice is extracted by K-Means clustering algorithm.The distribution of sea ice is described in sea ice concentration.The morphological features of sea ice in floating ice area are characterized by area,perimeter and roundness.According to the classification of sea ice area,the variation of each parameter with time is analyzed.In addition,the relationship between the parameters and the physical process of ice breaking is studied.The results show that,based on high resolution SAR data,fine sea ice distribution can be obtained.Combining it with traditional sea ice concentration products,the multi-scale ice map with more abundant information can be constructed.It can provide support for polar sea ice monitoring.It is of great significance for maintaining the safety of Arctic waterway and analyzing the climate effect of Arctic sea ice.