A Big Aurora Data Management Framework Toward Aurora Classification

摘要

Aurora is the only geophysical phenomenon observed directly with the naked eye in the high latitude regions. The systematic observation of aurora plays an important role in the magnetosphere and solar-terrestrial electromagnetic activity. With the observed aurora data growing quickly, aurora morphology research becomes data intensive. This paper presents a big aurora data management framework toward aurora classification analysis considering the scalability and performance. The framework integrates aurora data acquisition and management, aurora data preprocessing, data query and explore, and aurora classification experiment management. The NoSQL database is employed to manage the aurora data and the classification experiment data. The in-memory index is constructed to support the in-situ analysis.