摘要
中尺度涡信息的提取包括涡旋的识别和轨迹追踪,其自动识别与追踪对于基于海量数据的中尺度涡分析十分重要。传统涡旋轨迹自动追踪方法一般需要预先设定搜索半径的阈值,存在一定的主观性。针对传统中尺度涡轨迹追踪方法存在的问题,论文从聚类的角度出发,提出基于密度峰值聚类算法实现对涡旋轨迹的自动追踪,并以南海中尺度涡追踪为例,将基于聚类的追踪算法与传统的相似度追踪算法进行比较分析。结果表明:(1)基于密度峰值聚类算法,可实现对海洋中尺度涡的自动追踪,该算法涡旋追踪准确率优于传统相似度算法;(2)该涡旋追踪算法对资料的完整性依赖度较低,特别是对于存在部分缺损数据的情况仍能较准确追踪;(3)该追踪算法克服了传统涡旋追踪算法需要预先设定搜索半径阈值的问题,自适应性更强。
Information extraction of ocean mesoscale eddies includes the eddy identification and its trajectory tracking,both of which are very important for the research on mesoscale eddies based on the massive data.The traditional methods of trajectory tracking generally need to be set the threshold of the search radius beforehand,which could introduce a certain degree of subjectivity.To improve the existing problem of the traditional methods,an automatic tracking method of the mesoscale eddies is proposed in this study based on the Clustering by Fast Search and Find of Density Peaks（CFSFDP）from clustering point of view.Then it was compared with the traditional similarity algorithm by taking the South China Sea as a testbed.Our results show that：（1）Based on the CFSFDP algorithm,the automatic tracking of mesoscale eddies is realized,and the accuracy is better than the traditional similarity method;（2）The proposed tracking algorithm is less dependent on the data integrity especially for the presence of partial missing data;（3）Our proposed tracking method has stronger adaptability,which overcomes the problem that the search radius need to be set beforehand in the traditional method.
作者
王辉赞
郭芃
倪钦彪
李佳讯
Wang Huizan;Guo Peng;Ni Qinbiao;Li Jiaxun(Institute of Meteorology and Oceanography,National University of Defense Technology,Changsha d 10073,China;State Key Laboratory of Satellite Ocean Environment Dynamics,Second Institute of Oceanography,State Oceanic Administration,Hang zhou 310012,China;Unit 94587,Lianyungang 222345,China;State Key Laboratory of Marine Environmeural Science,Xia men University,Xiarnen 361005,China;Naval Institute of Hydrographic Surveying and Charting,Tianjin 300061,China)
出处
《海洋学报》
CAS
CSCD
北大核心
2018年第8期1-9,共9页
基金
中国科学院战略性先导科技专项(A类)资助(XDA11010103)
国家自然科学基金(41706021
41775053
41206002)
国家海洋局第二海洋研究所专项资助(JG1416)
中国博士后科学基金(2014M551711)
江苏省自然科学基金(BK20151447)
关键词
中尺度涡
轨迹追踪
密度峰值聚类算法
南海
mesoscale eddies
trajectory tracking
Clustering by Fast search and Find of Density peaks（CFSFDP）
South China Sea
