Research on Domain Variable Identification among Different Data Cubes

Large spatio-temporal Earth observation (EO) data are organized as multidimensional data cubes, which is a new paradigm for users to interact with EO data. For the multidimensional data cubes, geospatial dimensions, denoting by latitude and longitude, are indispensable. However, time series, climate and ocean data, etc. also require time axes and non-spatio-temporal axes in the n-dimensional structure. Therefore, at this level, a major challenge in data cubes interoperability concerns the harmonization of domain variables among different data cube implementations. Harmonization of data cubes from data with significantly different spatio-temporal reference systems always requires domain variable knowledge base. We propose to identify domain variables via a standardized resolver approach. In this way, cubes interoperability of different domain variables can be referred to and interoperated across multidisciplinary applications.