Semantic search

The Geoweb Server searches for and indexes geographically-referenced scientific data made public on the servers of organizations outside of the EPA, or on selected data servers within the EPA network. The outside sources are currently concentrated at NASA, NOAA, and the Center for Ocean-Land-Atmosphere. The inside sources include Community Mesoscale Air Quality data. The Geoweb Service has the potential to index any geographically-referenced data published on servers inside or outside of the EPA. User connection to search the index is only available from workstations on the EPA internal network.

The Geoweb Server normally performs its dataset indexing at a weekly interval. The Server is instructed to web-crawl an adjustable list of scientific data servers. Metadata about the files found on the indexed servers is brought back to EPA’s Geoweb Server and assembled into an xml database. This database contains all computer-obtainable metadata about each indexed file, and the location of the file on the Internet. The data files themselves remain on the servers where originally published. Currently, about 700,000 scientific files are indexed.

Users of the Geoweb Service connect to it at http://gds.rtpnc.epa.gov. They can follow links to specific data sources found on the home page, or perform searches for data by providing search terms. The searches execute quickly because the metadata being searched is all local to the Geoweb Server. The user is presented with a search results list, similar to Google or Yahoo search results. The user can then view the metadata associated with each dataset, or initiate a visualization of the data contents. The visualizations are performed using the World Wind Geo tool, based on NASA’s World Wind, and developed here at EPA; or by using Google Earth, NASA’s World Wind, or ArcGIS Explorer. The reading of metadata and the visualization of the data correctly referenced to the World Wind, Google Earth, or ArcGIS globe is supported for all data set types. Data served from OPeNDAP servers can also be subsetted by geographic bounding boxes, date, or time and the data subset can then be downloaded across the Internet to the user’s workstation.

The Atmospheric Science Data Center is responsible for processing, archival and distribution of NASA Earth science data in the areas of radiation budget, clouds, aerosols, and tropospheric chemistry. It is unique among NASA data centers in the size of its archive, cutting edge computing technology and full range of data services.