Data Quality Guidelines for AGDC Data Collections (Draft) - GeoscienceAustralia/agdc GitHub Wiki

Data Quality Guidelines for AGDC Data Collections (Draft)

The following document is based on a draft by Simon Oliver, Technical Lead, Business Systems Development, National Earth and Marine Observations Group, Environmental Geoscience Division at Geoscience Australia. Simon is currently overseas on secondment but discussions on the data quality criteria required for true interoperability should continue.

Fundamentals:

  • The AGDC presents a common analytical environment – the AGDC is not just a data warehouse. All data should be transformed to conform to the nested grid specification of the AGDC to facilitate true interoperability
  • The AGDC vision is to be able to perform data fusion analyses between heterogeneous gridded data collections with minimal overhead, e.g. products of different resolution and sensor
  • To achieve interoperability we need a few common elements for a proposed data collection to ensure compatibility:
  1. Data has a geometric base which is compatible with the existing data cube contents: i.e. Landsat uses the USGS GLS as a base with the SRTM 1 second DEM for ortho-correction
  2. Spectral data products should represent an inter-comparable metric: i.e. reflectance, which can be directly compared between different acquisitions and sensors
  3. Inclusion of Pixel Quality information is highly desirable
  4. Advocates should be aware of the National Nested Grid specification as it relates to geographic grids and projections
  5. Reflectance products should have common settings in terms of the nadir view / solar elevation etc.
  6. Appropriate metadata including both data and process provenance information should be provided for all data
⚠️ **GitHub.com Fallback** ⚠️