Quality Assurance

Supporting the ambition to “make the UK a global centre for trusted EO data”, a unique aspect of the EODH will be its attention to data quality and quality assurance (QA). It will provide easily understood data quality information built into user interactions via simple query options, aiming to raise awareness of data quality and foster data literacy for non-experts. This will lead to the provision of trustable insights from EO data that are fit-for-purpose. Led by NPL, the EODH QA Service aims to give users confidence in the quality of both:

• Data Products in the EODH Catalogue, through independent review of the processes by which Data Products are created, underpinned by quantitative validation.
• Applications that build on the EODH platform, by review through a similar process of underpinning evidencing relevant quality assurance information.

The EODH Data Product QA Process is composed of two core components. Results from both components are published in the EODH Catalogue for user interaction.

Quality Processes Review. Verifies that data products are developed according to best practices by reviewing provider documentation against established standards.

Validation. Quantitatively validates data product performance through automated peer-reviewed workflows, comparing observed performance against provider claims for defined metrics.

During the pathfinder phase of the programme the EODH QA Service piloted these QA processes for a set of core Data Products. The Data Products reviewed include: Sentinel-2 MSI L1C, Planet SuperDove constellation L3B, Airbus Pléiades MS-ORTHO.

Applications powered by the EODH must meet the following standards to be 'quality assured'. This framework ensures developers demonstrate the quality of their products to users, guaranteeing trusted data and outputs, and building confidence that results meet the necessary standards for their intended applications.

Self-evaluation. Review and collate evidencing documentation of processing methodologies and generated data for user transparency.

Evaluate Application input data. Using existing data product quality checks, defined by the QA Service, on the datasets being accessed from the EODH.

QA Service Review. Finalise QA reporting and publish application on the EODH website.

During the pathfinder phase of the programme the EODH QA Service piloted these QA processes for a set of core Applications built on top of the Hub. Find out more about how EO Pro and the Climate Asset Risk Analysis Tool (CLARAT) are quality-assured.

Building on the capability developed to validate optical product absolute radiometric calibration uncertainty, a beta version of a Calibration Dashboard Application for CEOS was developed to allow users to explore in more detail the validation results. Once released, the link to the dashboard will be accessible here.