极地研究

极地研究 ›› 2025, Vol. 37 ›› Issue (4): 681-701.DOI: 10.13679/j.jdyj.20240058

• 研究论文 • 上一篇    下一篇

基于MODIS热红外数据的北极冰间水道产品不确定性评估

翁志龙1, 刘岩1, 贾鹏1, 王彩虹1, 程晓2,3,4, 陈卓奇2,3,4   

  1. 1北京师范大学地理科学学部遥感科学国家重点实验室, 北京 100875;
    2中山大学测绘科学与技术学院, 广东 珠海 519082;
    3南方海洋科学与工程广东省实验室(珠海), 广东 珠海 519082;
    4中国高校极地联合研究中心, 北京 100875
  • 收稿日期:2024-05-30 修回日期:2024-09-02 出版日期:2025-12-30 发布日期:2026-01-12
  • 通讯作者: 刘岩
  • 基金资助:
    国家自然科学基金资助

Uncertainty assessment of Arctic sea ice lead products based on MODIS thermal infrared data

WENG Zhilong1, LIU Yan1, JIA Peng1, WANG Caihong1, CHENG Xiao2,3,4, CHEN Zhuoqi2,3,4   

  1. 1 State Key Laboratory of Remote Sensing Science, Faculty of Geographical Sciences, Beijing Normal University, Beijing 100875, China;
    2 School of Geomatics Science and Technology, Sun Yat-sen University, Zhuhai 519082, China;
    3 Guangdong Laboratory of Southern Marine Science and Engineering (Zhuhai), Zhuhai 519082, China;
    4 Polar Joint Research Center of Chinese Universities, Beijing 100875, China
  • Received:2024-05-30 Revised:2024-09-02 Online:2025-12-30 Published:2026-01-12

PDF (PC)

156

摘要： 北极冰间水道时空分布观测对探究极地地区水热通量、海-气物质交换及北极航道的开辟具有重要意义。星载热红外数据因其较高的时空分辨率和空间覆盖能力, 被广泛应用于冰间水道产品制作, 但目前不同水道产品仍存在较大不确定性。本文通过对国际上Willmes、Reiser和Hoffman共3种主要的热红外冰间水道产品进行比较分析和精度评估, 探索了冰间水道提取的不确定性及其来源。结果表明, 3种产品提取的日冰间水道面积存在显著差异, 就水道平均提取面积而言, Willmes产品占53.1%, Hoffman产品占106.0%, Reiser产品占140.8%。冰间水道产品提取的不确定性主要来源于3个方面: (1)狭窄冰间水道产品提取能力不足, 导致3种产品漏分率均高于50%; (2)3种产品是基于一天内多期热红外数据提取冰间水道的日合成产品, 海冰漂移导致同一冰间水道在不同位置被重复提取, 从而造成对快速漂移区冰间水道宽度和面积的高估; (3)Willmes产品与Reiser产品设计了云伪影过滤器(fuzzy cloud artefact filter, FCAF), 以减弱云等高温地物对冰间水道提取结果的影响, 但不同FCAF方案导致冰间水道提取面积差异显著。本研究不仅可为不同情境下的冰间水道产品选择提供参考依据, 同时也可为提升水道产品精度和未来发展指明优化方向。

关键词: 冰间水道, 中分辨率成像光谱仪(MODIS), 热红外, 不确定性, 北极

Abstract: Monitoring the spatial and temporal distribution of Arctic sea ice is important for improving our understanding of oceanic heat flux and sea-air mass exchange in the polar regions and for supporting the potential opening of Arctic shipping routes. Moderate resolution imaging spectroradiometer (MODIS) thermal infrared data, characterized by high spatial and temporal resolution and large coverage, are widely used to identify leads (ice-free waterways) in sea ice. However, there are significant discrepancies and uncertainties in the different products. This paper compares and evaluates the accuracy of three major thermal infrared lead products (the Willmes, Reiser, and Hoffman products, respectively) and analyzes lead extraction uncertainties and their sources. The results show that there are significant discrepancies between the daily lead areas in the three products, with the Willmes, Hoffman, and Reiser products extracting 53.1%, 106.0%, and 140.8% of the daily average lead area, respectively. There are three main sources of uncertainty in lead extraction: (1) narrow leads are inadequately extracted, resulting in misclassification rates exceeding 50% in all three products. (2) All three products are daily synthesis products that are based on the multiperiod thermal infrared data of a single day. Because of sea ice drift, the same lead is extracted repeatedly at different locations, resulting in overestimation of the width and area of rapidly drifting leads. (3) The Willmes and Reiser products are designed with a fuzzy cloud artifact filter (FCAF) to mitigate the effects of clouds and other high-temperature features on the extraction of ice-covered waterways. However, the different FCAF algorithms result in significant discrepancies in the extraction of the extent of ice-covered waterways. The results of this study can serve as a reference for the use of lead products in a variety of applications.

Key words: sea ice lead, moderate resolution imaging spectroradiometer (MODIS), thermal infrared, uncertainty, Arctic