6.12.2007

Panel: Successes in Optimizing Geospatial Access, or What ArcIMS has Meant for Geospatial Access

6/07/07 Moderated by Jeremiah Trinidad-Christensen

Keith Jenkins: Cornell University, CUGIR
Spoke about CUGIR http://cugir.mannlib.cornell.edu
  • All datasets have metadata and are downloadable
  • All data are free
  • 6700 datasets (geospatial) from ~11 different providers
  • Data providers are responsible for data, CUGIR is responsible for metadata
  • Data are .shp, .e00, .asc, .tif
  • CUGIR does not change data
  • Have “active” and “passive” providers
    • “Active” providers contact CUGIR to ask to have their data in CUGIR
    • “Passive” providers are contacted by CUGIR to have their data in CUGIR
  • Decide whether to keep older datasets that have been updated
  • CUGIR allows three different ways to find data
    • Browse
    • Select Area
    • Search

  • Not a Web mapping application, but static HTML image maps (HTML Image Mapper-?)
  • Created from Vector GIS data (batch-processed)
  • Simple, functional, reliable; don’t need to worry about an IMS going down
  • Don’t have aerial photos
  • Can save multiple datasets into a basket and download the zip file (automatically downloads metadata for user)
  • People weren’t interested in knowing what was newly entered into the catalog, so they decided not to have a link to that on the home page
  • Users may not know if the data they want is by quad or county, so it doesn’t make sense to have them select their area by choosing one of those

John Callahan: University of Delaware

  • Spoke about geospatial data at the University of Delaware
  • University of Delaware uses ArcGIS, which is used by 45+ departments and research centers
  • http://www.udel.edu/gis
  • Is pushing Open Source Software
  • Have statewide orthoimagery, landsat imagery, other imagery as needed
  • All raw data on the shared file server in Research & Data Management Service (RDMS) GIS research lab
  • Can be taken out on portable hard drives
  • Have an orthophotos index viewer
  • Have ArcIMS & WMS image services for each year for orthophotos (finds they run better than a catalog)
  • Delaware DataMIL (http://datamil.delaware.gov/) offers state framework data and topographic maps
  • Have U.S. Census data and University of Delaware campus data
  • Most people use mapping sites for viewing data
  • Many use Web for downloading data
  • Many use ArcGIS Desktop
  • Some request data on DVDs and/or hard drives
  • Few people request data in special formats (Geography Markup Language (GML), MapInfo)
  • Almost none (on campus) use Open Geospatial Consortium (OGC)

Dave Siegel: Harvard University

  • Harvard Geospatial Library (HGL) http://peters.hul.harvard.edu:8080/HGL/jsp/HGL.jsp
  • Presented, “Optimizing Access via ArcIMS, Portals, and the Desktop”
  • Infrastructure for cataloging, storing, and distributing geospatial data
  • ArcIMS is a metadata catalog based on FGDC
  • Raster and vector data
  • Everything in HGL is georeferenced
  • Log4J (an open source tool) allows them to track users and gather statistics
  • Not all files are in the same projection; they rely on ArcGIS to take those files and bring them into a common projection
  • Have a full-time FGDC metadata cataloger
  • Have a library portal, Web interface
  • HOLLIS, Harvard’s Library Catalog, provides Internet links http://hollis.harvard.edu/
  • If a map isn’t georeferenced, it is stored as JPEG2000
  • Allow a custom map service
  • Usability issues are a constant struggle (it isn’t like Google Maps)
  • They are looking to create an OGC-compliant service


Rick Lathrop: Center for Remote Sensing and Spatial Analysis at Rutgers University

  • Presented, “Web-Based Geospatial Data Visualization”
  • Not trying to be a one-stop shop
  • Work with ArcIMS to get the data out there
  • No need for GIS software for all users
  • Users wanted to be able to go in and have querying capabilities
  • Had problems with zonal summary statistics
  • They work with vector and raster data
  • With JavaScript, the delineated polygons become hotspots (when hovered, population data for the cities appear)
  • Using Google Earth and Google Maps because there is an end-user familiarity with the interface
  • Want to get customized information out to the users, not just data

No comments: