Open AccessConception of Geomagnetic Data Integrated Space
Author(s) -
A. V. Vorobev,
Gulnara Vorobeva,
Nafisa Yusupova
Publication year - 2019
Publication title -
trudy spiiran
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.242
H-Index - 9
eISSN - 2078-9599
pISSN - 2078-9181
DOI - 10.15622/sp.18.2.390-415
Subject(s) - earth's magnetic field , data processing , computer science , data integration , outlier , data mining , discretization , missing data , algorithm , mathematics , database , magnetic field , artificial intelligence , machine learning , physics , quantum mechanics , mathematical analysis
. As is known, today the problem of geomagnetic field and its variations parameters monitoring is solved mainly by a network of magnetic observatories and variational stations, but a significant obstacle in the processing and analysis of the data thus obtained, along with their spatial anisotropy, are omissions or reliable inconsistency with the established format. Heterogeneity and anomalousness of the data excludes (significantly complicates) the possibility of their automatic integration and the application of frequency analysis tools to them. Known solutions for the integration of heterogeneous geomagnetic data are mainly based on the consolidation model and only partially solve the problem. The resulting data sets, as a rule, do not meet the requirements for real-time information systems, may include outliers, and omissions in the time series of geomagnetic data are eliminated by excluding missing or anomalous values from the final sample, which can obviously lead to both to the loss of relevant information, violation of the discretization step, and to heterogeneity of the time series. The paper proposes an approach to creating an integrated space of geomagnetic data based on a combination of consolidation and federalization models, including preliminary processing of the original time series with an optionally available procedure for their recovery and verification, focused on the use of cloud computing technologies and hierarchical format and processing speed of large amounts of data and, as a result, providing users with better and more homogeneous data.