Baltimore Ecosystem Study

Baltimore Ecosystem Study Data and Metadata - Extensible Markup Language

Josh Cogan prepares a wireless remote sensor, 2007. Photo: Kathy Szlavecz

Overview

The Baltimore Ecosystem Study has adopted Extensible Markup Language (XML) as a means to transport and display metadata. An early adopter of the LTER Networks Ecological Metadata Language (EML), BES has been an integral contributor to its development.

Extensible Markup Language supports more applications than Hypertext Markup Language (HTML). HTML is the dominant form for transport of text and images in websites. XML differs in that it's "dialect" can be modified so that it can transport binary streams of information as well, lending it the ability to convey more than standard text content.

Additionally, XML is readily understood by database systems and web "robots" and thus lends itself nicely to the field of information discovery. It also allows the information to be passed across differing computer languages and operating systems.

At the same time, XML is designed to be reasonably human-legible. (Wikipedia.org has more on the syntax and form of XML online here - http://en.wikipedia.org/wiki/XML.)

Methods

BES metadata is kept in a relational database system along with online pointers to the data. The metadata describe the data as fully as possible.

A script, or program on the BES server reads the metadata and writes it to files in the EML format. It also writes a file listing the names and locations of those files into a file called the "harvestlist". The harvestlist file will be used to tell other computers the names and locations of the metadata.

A central "harvester" knows the location of the "harvestlist" file via a pre-assigned address. It periodically reads this file and learns the location of the XML record files for each BES metadata record. One by one, it reads these files and makes them available to a central search engine operated by the LTER Network Office in Albuquerque, New Mexico.

The result is a server with the ability to handle searches of every participating site. It supports search queries and also allows one to browse the records. You may reach the server online [HERE].

Note that the use of XML "translators" known as style sheets allow the page on the preceeding link to take on the appearance of the BES website.

GIS spatial data can be transported via XML as well. A second script ingests the output from BES geodatabase collections at the University of Vermont (and soon University of Maryland, Baltimore County and University of North Carolina) and performs the same operation, making those metadata records available for harvest. In this case, there are not yet direct pointers to the data, but a link to the website.

An additional script converts all the metadata records to "flat", human-readable files. This allows internet search engines, such as Google, to also discover the data.

Finally, a script has been developed to write out all this information to a webpage on the BES website so that the entire list of metadata can be viewed online. This page is online HERE.

Jonathan M. Walsh, 2007.

This material is based upon work supported by the National Science Foundation under Grant No. 0423476. Any opinions, findings, conclusions, or recommendations expressed in the material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.