Concept of Processing and Interpretation of Spatiotemporal Vector Data

Abstract

The purpose of this work is to develop an approach to complex processing and interpretation of spatio-temporal vector data. In this regard, it is necessary to analyze existing methods for visualizing vector and tensor fields and organizing a database for their storage and develop a concept which contains the main advantages of these methods and allows one to interpret spatial data represented as a tensor. The developed concept proposes an approach to field visualization based on drawing superellipse glyphs at each point in space. Axes of glyphs correspond to the rank of the visualized tensor, and attribute values are expressed by varying the color gradient in its monochrome representation. The architecture of a web-based application, which implements the storage of spatial data in a hybrid DBMS, and interpretation of the necessary data upon request is also proposed. For the physical storage of tensor information, a hybrid relational-hierarchical model is proposed, in which the representation of direct spatial data and the corresponding metadata, including those characterizing the source of information, are separated. It is advisable to present the final result and intermediate data in GeoJSON format. This format is actually currently the standard for presenting geospatial data and is supported by almost all software platforms and libraries aimed at solving problems of this kind.