Using Earth Observation Data Cubes for dynamic mapping at scale: A Pacific Perspective
To ensure the continually increasing quantity of earth observation (EO) imagery is utilised to its full potential, data access needs to be efficient. Important advancements in distributed computing, cloud-based infrastructures and the creation of open-source technologies are lowering the barrier to accessing big data, including satellite imagery.
The Open Data Cube is a core technology that has enabled the creation of multiple EO data cubes around the world. This talk will discuss the ongoing implementation of the IPP Common Sensing data cube, located over the South Pacific region (covering Fiji, Vanuatu and the Solomon Islands) which has been funded by the UK Space Agency’s International Partnership Programme. The key technologies used, as well as the lessons learnt and realities of developing a data cube from the perspective of an EO data scientist will be presented.
For Pacific Island nations the ability to respond to, prepare for and anticipate hazardous events and disturbances related to the climate, known as climate resilience, is increasingly vital. Based on a user-centred design approach, the IPP Common Sensing project has developed a platform to assist government level ministries in making decisions around climate resilience of which the data cube is part of the solution.
From a data perspective, several open imagery collections including Landsat, Sentinel-2 and Sentinel-1 have been processed to approved Analysis Ready Data (ARD) standards, which are directly accessible to the users. Derived products have also been generated such as advanced water masks, as well as products from set workflows which enable users to query the data cube in a flexible and customisable way via a portal. The implementation of this system has taken advantage of several open-source tools, including Dask, Kubernetes, Docker, STAC and COGS, to create a scalable way of gaining valuable information from the imagery.