Skip to Content

LTER Information Management– Past, Present and Future – a Rough Timeline

Printer-friendly versionPrinter-friendly version
Issue: 
Spring 2014

John Porter (VCR)

In considering where LTER Information Management is going, it is valuable to think a bit about where it has been.

When LTER started:

  • Only a few investigators had personal computers, most of them had black-and-white, text only, monitors
  • No one had electronic mail
  • Network connections were all via modem, and ran at 0.0003 to 0.0012 megabits per second
  • Floppy disks would not fit in your pocket (5-1/4” or 8”  across)pictures of floppy disks and cassette tape
  • Visicalc was the dominant spreadsheet program
  • A very large hard drive was 20 MB
  • No LTER site shared data in any systematic way

When the Internet became available to scientific users
(circa 1989):

  • Only a few investigators had connections to the Internet
  • Tools to use the Internet were primarily Telnet and FTP
  • There were no graphical user interfaces for working on the Internet
  • The first LTER-wide personnel and bibliographic databases started to be assembled
  • Many email systems could not intercommunicate with one another (e.g., Forest service could not talk to universities) so LNO IM Rudolf Nottrott set up an LTER-wide mail forwarding system that crossed these boundaries, along with the first email lists
  • The LTER Network Office was founded
  • A “minimum-standard installation” (MSI) was established focusing on getting GIS capabilities at every LTER site. It was supported by a series of large NSF technology supplements.
  • The LTER Network was working to create its very first Data Catalog, which was limited to 10 datasets per site
  • The LTER Network prepared the first guidelines for site data access policies
  • Only one LTER site (HBR) had a “Bulletin Board” system for sharing data via dial-up connections
  • LTER Databits was started
  • Work on establishing a minimum standard content for LTER metadata started

The “modern,”  Internet-centric  version of LTER came into being in 1994 with the advent of the first web browser.  At that time:

  • Many sites were operating online systems using Gopher and later HTTPD servers
  • The LTER Coordinating Committee mandated that each site share at least one dataset online
  • Many LTER sites had developed data access policies governing how data from the site could be shared
  • LTER Information Managers had defined a list of common metadata elements
  • ClimDB started to be planned (1996)
  • The idea of a network-wide information system was adopted by LTER
  • LTER Databits went on a 5-year haitus, returning in 1999
  • Hard drives are typically less than 100 MB in size

The age of network data systems arrived in 2002:

  • The Ecological Metadata Language standard version 2.0 was released
  • The first versions of the Metacat data catalog and Morpho EML editor were produced
  • The previous year, the LTER Lead Principal Investigators had added “sharing data” to the list of LTER Network goals
  • The line between web page and database was beginning to be blurred, content management systems began to be seen
  • Hard drives are approaching 1 GB in size

More recently (2010-2014):

  • The LTER Data Portal moved to use the PASTA system, assuring better quality metadata and data and more reliable downloads
  • LTER increasingly uses web services to share information among programs running at many different sites
  • The Drupal content management system is used by many sites, and specialized tools are developed for use at ecological research sites
  • LTER Information Managers have established controlled vocabularies for data keywords and units
  • LTER data is included in DataONE
  • Multi-terabyte hard drives are common
  • Many LTER systems run on virtual computers

The next decade:

  • Data collection is increasingly digital from end-to-end, with automated sensors and electronic field notebooks coming to dominate paper forms
  • Control of field tablets and computers will be via audio or gestures
  • More LTER sites will start to use a variety of drone aircraft for real-time remote sensing and for in situ measurements using drone-deployed sensors
  • Institutions start to evaluate research productivity using citations of published datasets, as well as traditional manuscripts
  • Now that every researcher needs to have a data management plan, more commercial products and services aimed at meeting the needs of scientific researchers will be created
  • Researchers will start to use standard ecological datasets, such as those produced by NEON

The next century:

  • Few LTER researchers will have anything recognizable as a computer of today, but will instead have a variety of interface devices that link them to cloud-based resources
  • Computer-mediated semantic interpretation will help to automate assembly of data for analysis
  • The data point (with associated metadata), rather than the dataset, will become the focus of information management efforts.  A query, rather than selecting datasets that qualify under search conditions, will return a collection of individual data points assembled from a large number of individual studies
  • The analysis and interpretation of large amounts of data will be the primary challenge, not the collection of data
Don Henshaw made several helpful comments on this timeline, but any errors remain my own.