Research On Data Management And Sharing Method Of Global Argo Data

More than 12,000 floats have been deployed since 1998,the year that Argo program was born.Argo data has very important scientific significance of basic researches of ocean and atmosphere sciences.People at home and abroad,have developed a variety of Argo data sharing platform,but all of them are less than real-time and flexibility and have high price of updating in terms of sharing of float trajectory and profile chart.These shortcomings affect the obtainment and utilization of Argo data and promotion of Argo program.To solve above problems,this paper start with the structure of Argo data itself and analyze Argo data structure and its characteristics:structured metadata store location and other attribute information,semi-structured store observation data as multiple columns digital matrix in the text.Then,this paper proposes a method of collaborative management of structured and semi-structured Argo data,establishes Argo database,designs varieties of data access interface based on Web Service and realizes real-time visualization of float trajectory.This research focuses on solving following key problems:(1)Design a multi-level tree JSON data model used for organizing profile observation data extracted,which can solve the tough problem of original data is not easy to parse and cannot manage.Moreover this model is able to adapt to the requirements of the changing variables,the observation depth and sampling density of"Global Argo Array".Furthermore,this paper establish a relationship between profile metadata and observation data by means of extended support of relational database to a semi-structured property,and finally accomplish the construction of heterogeneous database.(2)Propose a multi-table storage method to reduce storage pressure of single data table as data volume growing rapidly,and global Argo data in period of 1997-2016 are parsed in database,the results indicate that the method has good scalability of both storage and query,as well as high efficiency of float trajectory query.(3)Realize the real-time dynamic visualization of multi-float trajectory.First unify the frequency of different trajectories' drawing.Second,realize the time synchronization during asynchronous rendering by setting the starting delay.In addition,design a method of adding auxiliary line,in order to correct the mistakes in project of Web Mercator during drawing line cross 180 degree meridian.In the last part,this paper designs "Global Argo Data Sharing and Service Platform"using GIS services,map slicing and double caching technology and FSH(Flex + Spring+ Hibernate)layered-architecture.This platform integrates varieties of functions,including multi-condition query,sea area query,multiple trajectory real-time rendering,interactive profile chart and so on.And this platform has been deployed in China Argo Real-time Data Center to pilot run,which can provide Argo information and data services to customers at home and abroad quickly and conveniently,and can promote Argo program.Our result will offer a reference for other similar observation profile data management.