Research, Development And Integration Of Web-MAGIS-Oriented Spatial-temporal Statistical Analysis For The Ocean And Atmosphere

Spatial-temporal statistical analysis is an effective method to deal with large amounts of data and to reveal the inherent law.The ocean atmospheric environment is multi-dimensional and dynamic,and due to the introduction of the time dimension and the height(or depth)dimension,the amount of data has a sharp increase.Large amounts of multivariate data could certainly provide more valuable information for researchers and decision-makers,but it is more difficult to control and use a larger amount of data.Marine and meteorological researchers often use the spatial-temporal statistical analysis to deal with the mass ocean data,so as to avoid the duplication of information,grasp the principal components of the data,and to find out the rule or the abnormal phenomenon hidden in the ordinary phenomena or the complex data, thereby provide reference and convenience for the marine work and human life.This dissertation takes the Web-MAGIS-oriented spatial-temporal statistical analyses for the ocean and atmosphere as the main research topic and aims to build an effective spatial-temporal statistical software platform for the ocean and atmosphere. The main research content of this dissertation focuses the development and integration of the spatial-temporal statistical analyses,ranging from the research,design and development of the spatial-temporal statistical methods to the integration of Web-MAGIS and the statistical analysis module.It takes a general research on the applications of the statistical analyses and focuses on the algorithms of the statistical analyses.Based on the research results,an order-oriented Web-MAGIS platform will be developed and integrated.The main contents and results of the research in this dissertation are summarized as follows:(1)The application research of 46 kinds of the spatial-temporal statistical analyses The dissertation discusses the definitions,functions and applications of 46 kinds of spatial-temporal statistical analyses of the Web-MAGIS:the basic statistics,Buffer Analysis,Empirical Orthogonal Function,Empirical Mode Decomposition,Singular Value Decomposition,Powerful Spectrum,Fast Fourier Transformation(FFT), Cross-spectral Analysis,Harmonic Analysis,Spectral Entropy Analysis,Wavelet Analysis,Singular Spectrum Analysis,Trend Analysis,Cluster Analysis,Factor Analysis,Discriminant Analysis,Two-dimensional Fourier Analysis,Multivariate Regress Analysis,Canonical Correlation Analysis,Anti-distance Weighting Interpolation,Kriging Interpolation and so on.(2)The research and development of the algorithms of 19 kinds of spatial-temporal statistical analysesThe dissertation discusses in detail the theories of 19 kinds of spatial-temporal statistical analyses of the Web-MAGIS:Buffer Analysis,Empirical Orthogonal Function,Empirical Mode Decomposition,Singular Spectrum Analysis,Trend Analysis,Factor Analysis,Multivariate Regress Analysis,Anti-distance Weighting Interpolation,Kriging Interpolation,Error Analysis and so on,and focuses on the algorithm,design and implementation of the Buffer Analysis Based On Line, Multivariate Regress Analysis,Empirical Orthogonal Function,Empirical Mode Decomposition,and Kriging Interpolation.(3)The integration of the order-oriented Web-MAGIS platformThe order-oriented Web-MAGIS platform is the product of the current computer network technology and the Web-MAGIS platform.According to the independency of the orders and the sharing of the resources in the order system,this dissertation uses Web Service and multi-threading technology to distribute the different services to multiple Web Service servers,and distribute the different orders to different threads. The Web Service servers can not only share the burden of the Website server,but also carry on the multiple orders concurrently.As a result,the total execution time of orders is reduced and the efficiency is improved.(4)The integration of spatial-temporal statistical analyses in Web-MAGIS Owing to the fact that the spatial-temporal statistical analysis is numerous in category and wide in the difference of operands and the complexity of the algorithms, it is the most complex part in the integration of Web-MAGIS.According to the difference of the operands and the complexity of the algorithms,the analyses are divided into multiple parts in this dissertation,and every part uses a method to execute. Through Batch technology,the analyses with a high complexity of time and space are carried out effectively with the help of disk space,which is the key of the dissertation. Moreover,to reduce the maintenance cost of Web services,the dissertation adjusts the constructive method of Web service,focusing on the functions of the analyses and the structures of the operands.In short,the dissertation provides an effective integrating method of spatial-temporal statistical analyses,ranging from the construction and management of the Web services to its implementation.