Spatial data infrastructure for ecological research

The Biodiversity and Ecosystem Informatics Research Agenda (http://bio.gsfc.nasa.gov) notes current inability to compare data across spatial scales as a critical problem: Biological data from different sources are frequently collected and presented in different scales and resolutions resulting in a loss of detail when multiple data sets are required for data synthesis and analysis. The proposed infrastructure would allow a scientist to define a data set by putting together spatial building blocks represented in conceptually familiar, domain-specific terms. Data sets thus constructed (or recast) would be amenable to the automatic integration of spatial locations and measurements with automatic spatial data transformations. This would allow spatial analysis of individual field data sets and the linking together at same and different spatial scales data sets defined within the spatial infrastructure. We aim to include productivity enhancing research tools such as field data gathering devices and data validation and analysis, and to automatically tag field data with metadata early in the research cycle. We also contribute to the synthesis of metadata and data. This paper describes a recently funded NSF project that aims to build such an infrastructure. We describe the need for the infrastructure, our design approach, work to date and future work.