MapServer: Geospatial Access to Data

There are several ways that websites provide geospatial access to geospatially enabled data. We sought a solution that would be data driven (that is, display data directly from our metadata database and data servers and not have to rely on pre-made maps and images), and use open source software. We decided on the MapServer software, initially developed by the University of Minnesota. The MapServer software has several advantages, including its dependence on standards, such as OGC's Web Mapping Service. We were also lucky enough to have our contract programmer, Charlton Galvarino, be very proficient in using this package, so he has been able to tailor the interface extensively to suit our users' needs.

MapServer, because of its OGC underpinnings, enabled us to provide Web Feature Service (WFS) access to our data with little effort.

Our metadata database provides the basis for all of the maps generated on the fly and displayed on the user's browser. However, we also wanted people to be able to view the data on the browser as well. We are interested in enabling the user to decide whether the datasets are of sufficient interest to warrant their download to their own platform for further analysis and study. However, we do not intend to provide the analysis software via our web site. We do not want to decide how to contour the raw data into a regular grid, for example. These decisions are up to the investigator to pursue once they have the data with which to work.

To facilitate the viewing of data, we provide several ways of displaying data. These include:

  1.  Basic X-Y plot (such as salinity versus depth)
  2.  Abundance plots (circle size log proportional to count or weight)
  3.  So called time-series plots
  4.  X-Y plots of data from multiple stations
  5. Ability to display the data when the data are themselves images or movies

Since we wanted to facilitate the downloading of the data (not just the metadata) we added the Web Feature Service OGC protocol to our list of download options, which already included downloads as ASCII (tab, comma and space separated), Matlab binary file, netCDF, ODV and KML files. KML files can be used by such applications as Google Earth and Google Maps.