Sunday, April 09, 2006

Want to Be a GIS Librarian? 10 Things You Should Know

So, you want to be a GIS Librarian, eh? Or, perhaps you are a new GIS Librarian. Possibly you are, as there have been quite a number of available positions lately.

Well, here are 10 things that I think you should know and that I wish someone had told me three years ago. (Yes, this is a bit long-winded, but once I started to type I just could not stop...Well, eventually I did...obviously...)
  1. There are existing GIS Librarianship-related peer support structures.

  2. Involvement with local GIS agencies
    • More than most librarians, GIS librarians need to work very closely with local government agencies, such as city planning departments and local government councils. These are the creators of the datasets that our users will no doubt need, and believe me that it is never the case that everything produced by local government agencies are available for free download.
      • So, what type of involvement are we talking about here? Attending local (off-campus) GIS events, making appointments to visit planners and other GIS professionals, and often times keeping up with series of public information requests. Collaborative efforts, such as joint GIS-Day activities are a great way to start and continue great relationships.
      • Sometimes government agencies might be hesitant to turn some datasets over to the library. It is often necessary to convince them of the policies and procedures that you have set in place (see number 4, below), how authentication is handled, and how it will save them the trouble of having to deal with data requests from individual students or professors on campus. If all datasets are given to the library on a regular basis, no one on campus will need to bother the local government agencies with requests.

  3. What (if any) datasets to purchase and where to purchase them?

  4. How to store and provide access to geospatial data?
    • Learning to operate the basics of GIS software applications is challenging enough, but most libraries' first foremost geospatial endeavor is to provide access to geospatial datasets. This often requires technologies beyond base GIS applications, such as RDMS, scripting/programming, and web development.
    • How should data be collected and locally stored? How does the selection of data storage medium affect access?
      • CD-ROM -- If data is stored on CDs, where will the CDs be stored? As a reserve item at the circulation desk? At the reference desk? These locations are usually not near GIS-enabled computers. OK, then how about right in the GIS lab (if the library has one.)? Other issues arise due to staffing (what hours will staff be there to check out the CDs?), security, etc.
      • Server (Data Download)--- On the surface, this approach has numerous advantages over CDs, but new and ever more complex issues arise. Of course, server space is needed. Licensing becomes an issue as single-user CD products can not be made available to multiple simultaneous users over a network (without vendor permission). Authentication becomes an issue. Should you require authentication to pass through the library's proxy server? Should off-campus users be able to access the data?
      • Server (Web Services) -- Products such as ESRI ArcIMS and ArcSDE can be a fantastic way to provide access to geospatial data to multiple users. Always bear in mind the licensing constraints discussed above, however. You don't want a vendor to make an example out of your library. Very messy and embarrassing, I assure you. Creating web services that can be pulled directly into GIS clients is extremely handy for users, but can authentication be enforced? It would be easy to limit access to specific computers, but what if your institution has an educational site license for ESRI software? It becomes a challenge to maintain a database of GIS computers. You can limit access to the university IP domain, but the security risks become greater and greater.
        • Furthermore, there is an access distinction that must be made between ArcSDE and ArcIMS. ArcSDE allows multiple users to access data across the network with potentially the same functionality they have with locally downloaded datasets. ArcIMS allows image services to be streamed, but with severely limited functionality. They can look at it, but not download it or touch it. ArcIMS also has feature services with similar limitations, but the data is downloadable. However, this is limited strictly to vector data. And, well, if the users must download the data in order to use it effectively, then setting up an ArcIMS server might not be worth the effort.
        • Also remember that these options take a real, full-time, commitment. A dedicated server is often the best way to go and this costs money. Then there is training on installing and using these products effectively. Someone must build these activities into their job responsibilities. Should it be the GIS librarian? IT staff? A new employee?
      • At UT Arlington, we are currently using Citrix server technology to serve our CD data. Citrix is fantastic in that it allows us to serve data and/or applications without installing the application on the client computer, while limiting the number of simultaneous users. Anyone on campus can now access most of our spatial data collection, regardless of licensing and format, as long as the Citrix client has been downloaded and installed (3mb free download). Almost all of the vendors gave us permission to serve the CD products via Citrix, as long as we can limit access to the UTA domain (which is easy enough to do) and limit access to one simultaneous user.

  5. Stats & Numbers...So, you thought you were just learning how to join tables and help folks make some pretty maps, eh?
    • Rarely does someone need to make a simple map by joining features to a pre-created table to create a thematic map. This does indeed happen, but most times users need to use the GIS to answer questions that require some level of analysis and sophistication. Common examples include site selection, regression/correlation, ESDA, interpolation, and all sorts of other statistical goodies that a fear of math will prohibit a librarian from mastering.
    • It is very common to use various statistical packages in conjunction with the traditional GIS applications. Some of the more popular include GeoDa, SPSS, SAS, and The R Project.

  6. Constant training to perform day-to-day activities
    • The geospatial industry is accelerating so quickly that it really takes a full-time job just to keep track. Now, of course it is not necessary to keep track of every industry change and shift, but it is necessary to keep up with new technologies, products,and techniques as much as is humanly possible. If you do not, the struggle to get caught up will become a harder and harder struggle. When students who seek your assistance know more about GIS than you do, the value of the public services aspect of your position become horribly diminished.
    • So, how can you keep up?
      • There are quite a number of blogs, communities, and lists out there that you can pull into your feed aggregators. Two must-see websites are Slashgeo: In+ersec+tio for Spatial People and Planet Geospatial.
      • There are also quite a few GIS-related journals out there to monitor as well. Thank goodness they do not publish as frequently as blogs and Ingenta provides free RSS feeds. Here is a list of the GIS-related publications that I monitor.
      • Also try to take advantage of any training opportunities that you can. You can never, ever, know enough.
      • I have worked as a political science librarian and a government documents librarian, and most professional development endeavors helped to set big-picture goals and strategies. For the first time in my career, my professional development is helping to define how I operate on a day-to-day basis.

  7. Definitions of service must be constantly be reexamined as capabilities and needs expand.
    • As specified above, the fast development pace of the geospatial industry requires constant training and professional development. This, combined with a healthy dose of intellectual curiosity, will constantly increase the capabilities of a GIS Librarian. This will in turn require a regular inspection of the types and level of GIS services offered.
      • At first, many libraries seek to provide access to a few basic GIS datasets and provide minimal assistance to access and make use of these datasets. As competencies increase, levels of assistance might likewise increase. As knowledge of the various types of web services (WMS, ArcIMS) increase, alternative means of providing access to datasets need to be considered. As the popularity of the GIS services grow, the allocated budget for commercial datasets will grow, allowing the scope of the library's holdings to likewise grow.
      • As this simple progression implies, it is vitally important to keep goals, objectives, and scope of the GIS services offered by the library as flexible as possible.

  8. Coordinate multiple library resources.
    • The interdisciplinary nature of GIS sets this subject specialty apart from traditional subject areas, such as history, business, literature, etc. The GIS subject specialist often acts as much more of a coordinator of GIS-related library services.
      • Commercial datasets often are most often purchased by acquisition departments, who (1) might not know enough about the field to select materials on their own, and (2) usually have a severe dislike for CD products. Procedures and policies for handling geospatial data purchased on CDs might have to be drafted. All acquired datasets must be cataloged, usually in MARC and possibly FGDC metadata. The hardware and possibly server must be fully supported by IT departments, who will find that a lab or classroom of fully-functional GIS workstation will be much more work than any other configuration in the library. (The full ArcInfo installation is 3 CDs by itself). It will be necessary for continued outreach focused on various departments on campus, and so the appropriate liaison librarians will need to be consulted.
      • One of the most common ways that libraries handle the necessity to marshal various library resources to support GIS services is to form geospatial data committees, teams, or working groups. This is the best solution as the responsibilities then become established into the normal working routines of the committee members.

  9. Metadata, metadata, and more metadata.
    • So, you've decided to begin collecting geospatial data. Numerous questions must be considered concerning the catalog records.
      • Will you create FGDC metadata records and/or MARC records? MARC is not designed to handle geospatial datasets, but there is a crosswalk.
      • Who will create the records? A traditional cataloger will need some training to understand how to create FGDC metadata records. There are a number of free tools to help generate the metadata, as well as ArcCatalog.
      • How will users browse and/or search through these records? Will you create a searchable metadata catalog?
      • Always remember that records will need to be updated as data storage mediums (above) change.

  10. Have no doubts that you are embarking on an exciting career field, and one that I believe represents the types of public services librarians will be providing in the future.
    • Best of luck.


GeoMullah said...

I'm a geographer lucky enough to work in a map library. Let's talk.

Roxanne said...

This blog is great. I'm a geographer hoping to be lucky enough to work in a map library. What advice have you for the library student trying to break into this field?