History of Geographic Information Systems (GIS) in the LTER Network
Theresa Valentine (AND)
GIS use started growing within the LTER Network in 1989 after a report to NSF (Shugart Committee Report) and the development of the Minimum Standard Installation (MSI) resulted in supplemental funding to LTER sites, where the majority of the proposals sought to develop site competency in the areas of GIS and connectivity. There was a series of workshops that were funded by NSF, one a GIS training and the other a Remote Sensing Workshop. Sites acquired GIS software and hardware (mostly UNIX boxes with color screens, digitizer, plotter, and ARC/INFO software). The GIS and computer industry were in transition from mini-computers to UNIX systems.
In 1990 a survey was conducted to provide a snapshot of the technologies used for information management at each site. This survey was used to develop a report (Foster and Boose) to NSF on the effectiveness of the supplemental funding program.
GIS was a hot topic with information managers, based on a review of the early DataBits articles and IMC meeting notes. There was a GIS Corner column in the newsletter, and additional articles on issues around GIS and UNIX administration. During the early 1990’s, GIS and computer technology continued to evolve quickly, with increased computing power on personal computers, and the development of PC ARC/INFO and user interface programs such as ArcView. Plotting capability went from pen plotters, where shading was difficult, to the development of HP InkJet plotters. Costs for hardware and software were decreasing and universities started offering classes in GIS and automated cartography. Access to GIS technology moved from working in a laboratory setting, to desktop accessibility.
Many of the LTER sites cooperated with Universities in setting up GIS laboratories, with digitizers, plotters, computer terminals, and software for GIS and Remote Sensing capabilities. The Network Office, located in Seattle, had a GIS/Remote Sensing computer lab that site researchers could access remotely or locally.
GPS technology was becoming available, although the cost of GPS units was high, and satellite access was somewhat restricted. The Network purchased several GPS units, and distributed them around the sites, and the technology and expertise was shared. Sites would combine training with field campaigns to gather GPS locations for study sites and site infrastructure.
The All Scientists Meeting (AMS) has been a forum for discussion, sharing on GIS technology, spatial data management, and research or science topics related to spatial analysis. The 1990 ASM had workshops on GIS, remote sensing, and GPS.
The 1994 Information Managers meeting in Seattle stands out as a critical point, where open sessions were held on the management of spatial data and inter-site data access. There were invited speakers with breakout sessions. Groups worked on spatial metadata standards, spatial data exchange, user access and image catalogs, and proprietary issues. Recommendations included defining the capabilities of a spatial, visual, statistical, and modeling lab. “This should include GIS/RS capabilities, a wide range of statistical and visualization software and a combination of hardware, software, and technical personnel necessary to promote ecological modeling.” Nancy Tosta presented on the Federal Geographic Data Committee (FGDC) and the National Spatial Data Infrastructure. This was the beginning of the FGCD geospatial metadata standard, and metadata as something LTER should be thinking about.
Something interesting happen after this meeting, as GIS fell off the radar, at least at the network level. Focus for the Information Managers switched to developing the Network Information System (NIS), the Network Office moved to University of New Mexico, Jerry Franklin retired as chair of LTER, and the DataBits newsletter went on a hiatus when John Porter left for a detail to NSF. The next time GIS is mentioned is in the 1999 report of the LTER Technology Committee. The Technology Committee was working to acquire Landsat data for LTER sites, and was working with the San Diego Super Computing Center, focused on high performance computing, data storage, modeling, and visualization of ecological data.
LTER Information Management Committee GIS Working Group Formed
LTER GIS picked up again at the 2000 ASM in Snow Bird, Utah. Ned Gardner organized a workshop called “GIS on the Internet and LTER: A frontier for research, applications, and solutions.” The IM GIS Working Group was formed from this workshop to develop cross-site and intra-site research proposals that would help push internet mapping technologies in directions appropriate to ecological research. Early members of this group included Ned Gardner, Theresa Valentine, Barbara Nolen, Todd Ackerman, Peter McCartney, John Vande Castle, and Ken Ramsey.
Several workshops on Internet mapping and web services were held jointly with the San Diego Super Computing Center, with representatives from LTER sites, universities, and the private sector. The big topics were Internet mapping, spatial data and EML, the LTER Spatial Data Workbench, web services, spatial and non-spatial data integration, and delivery of data to researchers. The group conducted a survey to find out what types of web mapping sites were doing, and in 2003, a GIS/Internet Mapping edition of DataBits was produced (Fall 2003). It is interesting to note that an “editorial” section in DataBits was developed after this somewhat controversial issue.
GIS software was now established on the PC desktop, and emerging technologies included managing spatial data within a relational database structure through ArcSDE or Oracle Spatial, ArcGIS with extensive desktop GIS tools in a GUI (command line and macro tools were gone), Internet mapping programs (ArcIMS and Minnesota Map Server). The 2002 Winter IMExec meeting notes in the Spring 2002 DataBits edition recommend that Spatial data, traditionally managed by GIS and remote sensing experts, should be incorporated into IM systems at the site and network level.
The GIS Working Group focused their interactions on workshops at the All Scientists Meetings, with little interaction between, although there was some when GIS folks attended the annual IMC meetings. Workshops at the ASM in 2003 and 2006 were successful, but it was difficult to sustain involvement in between meetings.
GIS technology has moved to the desktop and the internet, with mapping applications such as Google Maps and Google Earth, the availability of world wide data sets through map services, and the increased proliferation of inexpensive GPS units. GPS satellites became more abundant and available to non-military users. Phones were geo-referenced and citizens became more spatially aware with MapQuest, and navigation programs for vehicles. This is the start of the downsizing of computer equipment, with smart phones and tablets.
A GIS survey was conducted in 2007 and identified a contact for each site, and what base layers were available, information on status of digital ortho-photography, Landsat scenes, core site locations, EML, spatial data on-line, and internet mapping capability. At that time only 8 sites did not have Internet mapping capability, most sites had digital elevation models, and all sites were using ESRI products. This survey was conducted by Barbara Nolen (JRN). The 2007 IM Committee meeting in San Jose, California had a town hall discussion on integrating diverse data types and specifically spatial data into the LTER cyber infrastructure. Mike Goodchild from UC Santa Barbara addressed the challenges of managing and utilizing spatially enabled data. The GIS working group updated recommendations for GIS at sites. Recommendations included basic spatial information and directions for geo-spatial referencing for study locations. The idea about developing a centralized portal for users to search, display, and access all site spatial data and a strategy for providing basic map service capacity for every site in the network was formulated. The group met for two planning sessions in 2008, and made formal recommendations to the IMC. The idea of a centralized portal for access to LTER GIS data was brought up again. A core GIS team worked together over the next year, meeting via teleconferencing, and produced a prototype to be hosted on the Georgia Coastal Website. Thus, LTERMapS was born.
The first LTERMapS internet mapping workshop was held in March of 2010, as a post-ASM product-oriented workshop. The workshop allowed several of the LTER Information Managers to work on porting the project to the Network Office and connect directly to the SiteDB database. The group completed a guide for using the Google Maps API, and scoped out specifications for the next phase of the project. Key players included Jamie Hollingsworth, Adam Skibbe, Theresa Valentine, Jonathan Walsh, and John Carpenter.
The working group continued to work on developing standards for GIS data, and tools for creation of EML from FGCD metadata. Jonathan Walsh and Theresa Valentine participated in a Best Practices for EML workshop, to help with examples of how to document site locations, geographic coverage, and GIS data formats within the EML structure. A workshop on documenting GIS data in EML was held in 2010 at the Andrews Forest. Eleven sites were represented, and new versions of esri2eml style sheets were tested and developed. A document on the best practices for documenting GIS data to LTER standards was developed, along with the specifications and requirements for a centralized service for LTER GIS data called LTERMapS Phase 2. The LTER Cartographic Atlas was developed by Jamie Hollingsworth through an IM time buyout. The Atlas provides cartographic overlays for use in making presentation graphics, across the entire LTER network. Another survey was conducted, to try and get an idea of what types of GIS data sites had and what it would take populate a centralized server.
LTERMapS Phase 2 morphed into GeoNIS, which became a reality through supplemental funding by the LNO for face to face meetings of the team, and for programming support. The group edited the Spring 2012 edition of DataBits, highlighting geospatial activities at both individual and cross-site scales. This edition provided an opportunity to update the LTER community on the cross-site activities.
The GeoNIS project moved forward in 2013 and was developed to interface with PASTA, and provide a way to quality check spatial data in PASTA, and to provide web mapping services for LTER sites. The GeoNIS is currently in beta stage, with automated processes to ingest EML files from PASTA , download data, and create web mapping services for each site. There is an end user application to check for errors, and view the data ( http://geonis.lternet.edu/). Key personnel included Adam Skibbe, Aaron Stephenson, Jamie Hollingsworth, Theresa Valentine, Inigo San Gil, Mark Servilla, Ron Beloin, and Jack Pederson.
Key products and links for the LTER GIS Working Group:
- LTERMapS (http://www.lternet.edu/sites/map)
- Documenting Spatial Data to LTER Network Standards (http://im.lternet.edu/project/GIS_document),
- GeoNIS (http://im.lternet.edu/project/GEONIS)
- Cartographic atlas (http://www.lternet.edu/sites/lter-interactive-cartographic-atlas).
- GIS Working Group Terms of Reference (http://im.lternet.edu/sites/im.lternet.edu/files/GIS_working_group_TOR_v2.doc)
- GIS Working Group project page (http://im.lternet.edu/projects/gis_data)
The Future of GIS in LTER over the next 20 years GIS Technology (also known as Theresa Valentine’s crystal ball)
GIS technology is moving towards Open Source solutions, sharing code on sites such as Github (https://github.com/). The industry leaders and large organizations are pushing cloud computing, with software and data located in the cloud, and users working remotely on a variety of lightweight devices. Users should be able to access their data from any device, and applications will have to scale in size to work from large screens to small wrist phones.
Data streams continue to increase and the availability of LiDAR, satellite data, and digital orthophotography are challenges for current computer networks. Moving large data sets and models around in the cloud currently can be cost prohibitive, and it is unclear where the industry is moving in regards to high end GIS computing. ESRI is releasing a ArcGIS for Professionals, however it is in beta release, and it’s unclear how it will work with scientific needs, especially with large raster datasets. One should expect that band width and wireless accessibility should increase, although there most likely won’t be a business model to cover remote areas, where many LTER sites are located.
GPS units are becoming more plentiful, smaller in size, with increased accuracy, and are now found in phones, tablets, cameras, watches, and glasses. Technology currently exists that will track a person/vehicle as they move between boundaries of services areas (called geo-fencing), sending alerts, notifications and even advertising back to the person/vehicle (http://whatis.techtarget.com/definition/geofencing). This emerging technology might be useful in tracking animal movements.
Plotting of hard copy maps is being replaced by on-line web applications, and large and small format touch screens, with higher definition and 3D capability. 3D printing devices will also change how business is conducted and research tools are developed. Output will be videos, and screen shots.
GIS professionals will be expected to be proficient in computer science, building custom applications for personal devices, with less emphasis on cartography and geography. Data will be available over the web, and staff will need to be good at “mashing” data together from different sources and communicating with researchers to meet their analysis and application need. They will need to be able to sift through the massive amount of data to find relevant information. I suspect that there might be a few hard core GIS nerds left in 20 years, who still like the command line and long for the Workstation ARC/INFO days.
Databases will still be important, as applications will need to capture data from a variety of devices (both researcher and citizen scientist) and provide updated data for the applications. Data will also need to flow seamlessly between different programs, especially with statistical packages and modeling software. Multi dimensional datasets will become more and more important as global climate change continues to be an issue, and researchers will want better tools to visualize and analyze data across several dimensions.
Future direction of the group is to continue with the GeoNIS development moving from beta to production mode, continued coordination with the IM committee, keeping current on GIS technologies through coordination between site GIS specialists, and renewed commitment of and recruitment of members.
In twenty years, one would expect that the GeoNIS web services would be the geographic framework for LTER site researchers and cross-site projects to provide access for students, researchers, the general public, and educators to further the goals of the LTER Network and science. The GeoNIS should be transparent to the users, and integrated into the data management and science communities.