摘要
星载合成孔径雷达(synthetic aperture radar,SAR)具备全天时、全天候工作能力,在海洋监测方面已展现出巨大的应用价值。本文从海洋动力环境要素和海洋目标两个方面,系统总结了星载SAR技术在海洋监测领域的研究现状。对于前者,重点梳理了SAR用于海浪、内波、涡旋、风场、海流及海底地形等环境要素监测的主流技术与算法,并深入讨论了多波段、多极化与多模式SAR数据在提升反演精度方面的作用;对于后者,则系统总结了海冰、溢油、船舶及海上人工设施等海洋目标的识别方法,阐明了多极化信息在刻画目标散射特性中的关键贡献。此外,本文进一步评述了人工智能技术在SAR海洋监测中的进展,并对未来SAR海洋遥感技术的发展方向进行了探讨。
Spaceborne synthetic aperture radar(SAR),with its all-weather and day-and-night imaging capability,has demonstrated tremendous value in ocean monitoring.This paper provides a systematic review of the current research status of spaceborne SAR technology in the field of marine monitoring,from the perspectives of ocean dynamic environmental parameters and maritime targets.For the former,we summarize mainstream SAR-based techniques and algorithms for monitoring ocean environmental parameters such as waves,internal waves,eddies,winds,currents,and seafloor topography,and further discuss the roles of multi-frequency,multi-polarization,and multi-mode SAR data in improving inversion accuracy.For the latter,we review SAR-based methods for the detection of maritime targets including sea ice,oil spills,vessels,and offshore infrastructures,highlight the importance of multi-polarization information in characterizing target scattering properties.In addition,this paper reviews and evaluates recent advances in applying artificial intelligence to SAR-based ocean monitoring and discusses future development directions for SAR ocean remote sensing technologies.
作者
朱泠
陈鹏
郑罡
杨劲松
朱海天
任林
ZHU Ling;CHEN Peng;ZHENG Gang;YANG Jinsong;ZHU Haitian;REN Lin(Second Institute of Oceanography,MNR,Hangzhou 310012,China;State Key Laboratory of Satellite Ocean Environment Dynamics,Hangzhou 310012,China;National Satellite Ocean Application Service,Beijing 100081,China)
出处
《海洋学研究》
2026年第1期109-123,共15页
Journal of Marine Sciences
基金
国家重点研发计划(2022YFB3902400)
十三五预研项目“海洋卫星应用示范”。
关键词
SAR
极化特征
海洋遥感
海洋动力环境
海上目标
海洋信息提取
人工智能
反演算法
SAR
polarimetric features
ocean remote sensing
ocean dynamic environment
maritime targets
ocean information extraction
artificial intelligence
inversion algorithms
