Democratizing the Access to Geospatial Data: The Performance Bottleneck of Unified Data Interfaces

Abstract

The exponential growth of geodata collected through satellites, weather stations, and other measuring systems presents significant challenges for efficient data management and analysis. High-resolution datasets from Earth observation missions like ESA Sentinel are vital for climate research, weather forecasting, and environmental monitoring, yet their multidimensional nature and temporal depth increasingly strain traditional spatial data models. This research evaluates the performance characteristics of various geospatial data access and processing technologies through systematic benchmarking. The study compares file formats (NetCDF, Zarr) and interface standards (OpenEO, OGC WCS) on a Rasdaman database instance to determine optimal configurations for interactive data exploration and analysis workflows. Performance tests reveal a clear hierarchy in processing efficiency. Direct RasQL queries consistently outperform both OGC WCS and OpenEO API interfaces across all test scenarios. Array file formats demonstrate superior query processing speeds, likely attributable to reduced database technology overhead. The findings provide a robust foundation for selecting appropriate geodata processing technologies based on specific use case requirements, data volumes, and performance needs. This research contributes to enhancing the efficiency of multidimensional geospatial data handling, particularly for time-critical applications that require interactive visualization and analysis capabilities