Archived: 30 January 2006
University of Adelaide Masters by coursework thesis, November 2005.
Supervisors: Paul Coddington and Andrew Wendelborn
Most scientific communities include many different research groups who have collected large amounts of data over many years. Typically this data is only easily accessible and understandable by the researcher who collected it, since almost every research group will have different ways of specifying data and metadata, and even different ways of storing the data (e.g. paper records, files, spreadsheets, databases). In order to make this data readily available to all researchers, the community must define standard data formats, metadata schemas, and interfaces for querying, accessing and processing data from distributed data servers. Ecological Metadata Language (EML) is an XML schema that has recently been developed to describe ecological data.
The School of Earth & Environmental Sciences at the University of Adelaide has a huge amount of data collected from many different freshwater environments (lakes and rivers) around the world over the past few decades. In order to make this data more useful to researchers, a new way to organise the data is required.
This project involved designing and implementing a prototype online data warehouse for environmental data, by using the EML metadata standard and extending it to fit the specific requirements of water data, and developing a standard database schema for the data warehouse that can be used instead of the different schema for each different data set. The project has also implemented a prototype web-based system for querying and accessing the data.