INTEGRATION OF THE OPEN DATA CUBE ON COMMON CLOUD FRAMEWORKS

The Open Data Cube (ODC) [1, 2, 3] is an open-source geospatial data management and analysis software package growing in popularity. With increased popularity there is increased demand for the computational power and storage capabilities required to analyze large spatial and temporal datasets. Cloud resource providers have bolstered their computational and satellite data offerings, simplifying access and management so that average users are able to tailor systems to their specific needs. Consequently, there exists substantial community interest in the comparative capabilities and technological performance of different cloud providers. Making informed choices in cloud providers offerings is crucial for optimizing geospatial data management and processing to meet user-specific needs. In this paper we begin to evaluate the deployment and performance of the ODC, standard notebooks, and datasets, to better determine functional differences. Our work finds that datasets saved in the local environment execute operations significantly faster than from the cloud, Google Earth Engine (GEE) has longer execution times when the metadata for a given asset is indexed locally, and that more thorough benchmarking is required to better understand the end-to-end performance of EO processes on the cloud.