极地研究

极地研究 ›› 2023, Vol. 35 ›› Issue (4): 508-516.DOI: 10.13679/j.jdyj.20220419

• 研究论文 • 上一篇    下一篇

基于PlanetScope影像的格陵兰冰面融水监测

朱雨欣1  张闻松1  杨康1,2,3
  

  1. 1南京大学地理与海洋科学学院, 江苏 南京 210023; 
    2江苏省地理信息技术重点实验室, 江苏 南京 210023; 
    3南方海洋科学与工程广东省实验室(珠海), 广东 珠海 519080
  • 出版日期:2023-12-30 发布日期:2023-12-30
  • 通讯作者: 杨康, E-mail: kangyang@nju.edu.cn
  • 作者简介:朱雨欣, 女, 2001年生。本科生, 主要从事冰冻圈水文遥感研究。E-mail: yuxin_zhu_nju@163.com
  • 基金资助:
    国家自然科学基金(41871327)和中国科学院战略性先导科技专项(XDA19070201)资助

Surface meltwater on the Greenland ice sheet mapped from PlanetScope small satellite imagery

Zhu Yuxin1, Zhang Wensong1, Yang Kang1,2,3   

  1. 1School of Geography and Ocean Science, Nanjing University, Nanjing 210023, China; 
    2Jiangsu Provincial Key Laboratory of Geographic Information Science and Technology, Nanjing 210023, China; 
    3Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai 519080, China
  • Online:2023-12-30 Published:2023-12-30

PDF (PC)

894

摘要： 格陵兰冰盖的消融及其对海平面上升的贡献成为国际上研究的热点。每年消融期, 格陵兰冰盖表面消融, 融水会导致冰面形成冰面湖、冰面河、注水冰裂隙等形态。格陵兰冰面融水规模庞大、结构复杂、变化迅速, 区域气候模型难以准确模拟冰面融水分布, 中等分辨率卫星影像难以反映冰面融水的时空变化。以 PlanetScope 为代表的 CubeSat 小卫星空间分辨率高达3 m, 理想情况下重访周期可达1 d, 这为精细化动态监测格陵兰冰面融水提供了可能。本研究利用 PlanetScope高空间分辨率小卫星遥感影像提取格陵兰冰盖西南部典型流域冰面融水遥感信息, 构建了针对 PlanetScope遥感影像的冰面融水深度反演公式, 对比了 MAR v3.11区域气候模型模拟的融水径流量与遥感反演的融水体积。结果表明: 在2019年7—8月, 流域内冰面融水开放水体比率先上升后下降, 在7月12日达到峰值8.7%; 流域内冰面融水深度介于0.2~1.5 m 之间, 冰面湖最深(0.9 m±0.2 m), 冰面河干流次之(0.6 m±0.1 m), 冰面河支流最浅(0.5 m±0.1 m); 遥感观测的开放水体比率、冰面融水体积与区域气候模型 MAR 模拟的融水日径流量具有正相关关系, 故融水径流对于冰面湖与冰面河具有直接供给作用; 流域冰面融水存储比例(遥感观测的冰面融水体积与模型模拟的融水累积径流量之比)先升高后降低, 冰面融水储存比例小于1%, 储存能力十分有限, 冰面流域能够高效输送融水进入冰盖内部。

关键词: 冰面融水, PlanetScope, 遥感观测, 区域气候模型, 格陵兰冰盖

Abstract: Surface melt of the Greenland ice sheet and its impact on rising sea levels have become hot topics in international research. Each summer, surface meltwater forms on the Greenland ice sheet, forming supraglacial lakes, supraglacial rivers, and water-filled crevasses. There is a large area of surface meltwater, which exhibits complex structure, and changes rapidly over time. Regional climate models do not accurately simulate the distribution of surface meltwater, and medium resolution satellite images may not reflect spatial and temporal changes in surface meltwater. CubeSats, such as PlanetScope, allow dynamic monitoring of surface meltwater by providing near-daily satellite observation at 3 m spatial resolution. In this study, we extracted surface meltwater in a typical internally drained catchment in the southwest Greenland ice sheet using PlanetScope images. We then constructed a formula to estimate meltwater depth from the PlanetScope images and compared meltwater volume estimated by satellite images and meltwater runoff simulated by the MAR v3.11 model. The results showed that the open water fraction (OWF) of surface meltwater in the catchment first rose then decreased during July and August 2019, peaking on 12 July (8.7%). The surface meltwater depth in the catchment ranged from 0.2 to 1.5 m, and the supraglacial lakes were the deepest (0.9 m±0.2 m), followed by the mainstem of supraglacial rivers (0.6 m±0.1 m), and the tributaries of supraglacial rivers (0.5 m±0.1 m). The OWF, meltwater volume and daily runoff simulated by MAR were positively correlated, indicating that meltwater runoff directly supplies supraglacial lakes and rivers. The surface meltwater storage capacity is limited (<1%), and the catchment is efficiently transporting meltwater into the ice sheet.

Key words: surface meltwater, PlanetScope, remote sensing observation, regional climate model, Greenland ice sheet