By Jaime Martindale, University of Wisconsin, Madison

New York City census blocks from Cornell University Geospatial Information Repository (CUGIR) highlighted with land cover and shaded relief from the USGS National Map.

The spectrum of student and faculty research projects at Cornell University in Ithaca, New York, is immense, and the potential for GIS to enhance inquiry, analysis, and display spans traditional disciplinary boundaries. Whether plotting invasive plant species in China or mapping social factors in Spain, GIS provides a critical tool for academic exploration.

In keeping with its mission to provide comprehensive physical and digital collections, the university's Albert R. Mann Library serves as a GIS resource for students, staff, and faculty seeking spatial data, adding maps to a thesis or dissertation, or implementing GIS in teaching or research projects. In addition to in-person consultations with GIS librarians, the library also offers instructional workshops for using GIS in areas as varied as water quality monitoring, endangered species tracking, and international development. In 1998 Mann Library extended its GIS activities beyond Cornell, launching the Cornell University Geospatial Information Repository (CUGIR) as a publicly available online resource (cugir.mannlib.cornell.edu).

Developed with funding from the Federal Geographic Data Committee (FGDC) National Spatial Data Infrastructure (NSDI) Cooperative Agreements Program, CUGIR is one of two NSDI Clearinghouse nodes in New York State and one of the few nodes developed and maintained by a university library. The CUGIR repository offers free and open online access to GIS data from 10 federal, state, and local data partners with special emphasis on natural features relevant to agriculture, ecology, natural resources, and human–environment interactions. CUGIR work group members also provide answers to technical inquiries and general GIS questions initiated via the Web site. Because the vast majority of users experience CUGIR remotely, CUGIR has become an important outreach service from Mann Library to the citizens of New York State.

A repository such as CUGIR can never remain static, and over the past two years the CUGIR work group has restructured CUGIR using relational database capabilities, cross-referencing 7,000 data sets by thematic category, data partner, and geographic location. As with many embryonic GIS projects, original assumptions about data uniformity proved insufficient, and the new database driven structure now accommodates data by diverse geographic units, including watersheds and individual regions within the state. In a step toward long-term preservation, CUGIR also supports archiving and the distribution of multiple data sets.

With additional seed funding from an FGDC Cooperative Agreements Program grant, CUGIR recently gained an important new feature: Web-based mapping. Any CUGIR user can now interactively preview popular CUGIR data sets online. The CUGIR work group chose ArcIMS 4.1 on Windows 2000 for generating the maps from data stored in ArcSDE 8.3 with Oracle9 on Linux. Data sets originally stored and distributed in the form of individual maps have been merged into seamless statewide ArcSDE layers for the map service, providing users a new perspective on statewide spatial data patterns and the freedom to browse across map boundaries.

The CUGIR Web mapping interface also features enhanced map service interoperability patterned on the U.S. federal Geospatial One-Stop E-Gov Initiative site (www.geodata.gov). Visitors can visualize and overlay data from the CUGIR collection and a variety of preselected map services from the U.S. Geological Survey, neighboring states, other universities, and Canadian data providers. Users more familiar with GIS may also connect to any map service published using the native ArcIMS mode or the Open Geospatial Consortium Web Map Service format by entering the appropriate parameters. This level of interoperability across multiple map servers and multiple formats assures access to a broader range of data than any individual server provides, allowing users to compare the quality and appropriateness of data from multiple sources without having to go through the effort of downloading and integrating the data on their own computers.

Support for interoperability across multiple map services also greatly facilitates GIS projects that span political boundaries. The CUGIR work group is currently collaborating with GIS data providers in Ontario, Canada; Pennsylvania; and Vermont to offer seamless data viewing across borders, further extending the value of GIS data beyond any single jurisdiction.

In future developments, CUGIR will leverage more ArcIMS capabilities for data extraction.

For more information, contact Jeff Piestrak, computer support specialist, Cornell University (e-mail: jmp36@cornell.edu, tel.: 607-255-9569), or Jon Corson–Rikert, programmer/analyst, Cornell University (e-mail: jc55@cornell.edu, tel.: 607-255-4608). Author Jaime Martindale was a contributor to CUGIR and to enhancements made to Mann Library GIS services during her tenure as the library's GIS librarian.