Evaluating and Improving Metadata for Data Use and Understanding

The last several decades have seen an extraordinary increase in the number and breadth of environmental data available to the scientific community and the general public. These increases have focused the environmental data community on creating metadata for discovering data and on the creation and population of catalogs and portals for facilitating discovery. This focus is reflected in the fields required by commonly used metadata standards and has resulted in collections populated with metadata that meet, but don’t go far beyond, minimal discovery requirements. Discovery is the first step towards addressing scientific questions using data. As more data are discovered and accessed, users need metadata that 1) automates use and integration of these data in tools and 2) facilitates understanding the data when it is compared to similar datasets or as internal variations are observed. When data discovery is the primary goal, it is important to create records for as many datasets as possible. The content of these records is controlled by minimum requirements, and evaluation is generally limited to testing for required fields and counting records. As the use and understanding needs become more important, more comprehensive evaluation tools are needed. An approach is described for evaluating existing metadata in the light of these new requirements and for improving the metadata to meet them.