But the tools for mapping haven't been particularly accessible to us and to our students until pretty recently. Suddenly we have at our fingertips a full suite of GIS tools, thanks to the site license for ArcView. And now the challenge is to figure out how to integrate GIS into teaching, learning, and research. The 'learning curve' for ArcView seems a bit steep, because the software is so amazingly powerful (there seem to be so many features and choices to make), and because the technical details of map projections, raster vs. vector, linked files of various formats, and so on are just not part of the background knowledge most of us have. I've heard it said that very few people more than 5 years out of graduate school have any GIS experience, and indeed it's really only in the last couple of years that desktop GIS has been a practical reality. It's a much newer tool than, say, word processing or spreadsheets, but no less revolutionary in the activities it enables. GIS is a transdisciplinary tool that offers the possibility of integrating perspectives and making findings accessible to broader audiences. The point I'd like to get you to today is "oh... is that all there is to it?"
So I want to show you the basics, talk a bit about kinds of data that are available, and invite you to begin thinking about how GIS capabilities might assist you in what you do. I also want to outline a bit of the directions in which I think we're headed with geographical information in the library.
For me it's a lifelong article of faith that maps are interesting, that the spatial distribution of a phenomenon is provocative, that there is pedagogical virtue inherent in maps. Maps are an essential tool for just about all aspects of Environmental Studies that I can think of, because they let us display and explore the spatial and temporal context of subject matter, and because they are a medium which facilitates communication with the various audiences a student of the Environment wants to reach. But it's really the analytical side of GIS that lets us do science.
Let's consider an example that will show us some of the basics and involve us in a problematic. Here's an outline map (a vector map, consisting of polygons with which values can be associated) that's entirely familiar to us [US counties]:
This map has a large table of data associated with it --more than 3000 county units, with data on use of 20 herbicides (acres treated, pounds used, pounds per square mile). The map and associated data came from a USGS site, where there's also data on crops, livestock, fertilizer use, and more.
We can display variables and do basic manipulations like 'normalization'. Here's a map of Roundup use in pounds per acre:
(steps: using H1099_lbs, (1) set null value at -99 (2) classify in 10 intervals (3) normalize by Area_acres (4) change colors of two highest classes for visual effect)
What do we see here? Interesting how often state boundaries turn out to be relevant, leading us to ask why? What's going on between Minnesota and Iowa, Texas and New Mexico, Indiana and Wisconsin and Illinois?
And here's one of the same manipulation for 2,4-D in pounds per acre:
Clearly a different dynamic from Roundup. North Dakota is more than striking --we need to find out what's going on there...
I've done a series of maps of the distribution of herbicide use in the counties of Virginia and West Virginia (to see the set), and I manipulated the data to make a summary map of total pounds of all herbicides per acre:
You can think of each of these maps as a layer over which other transparent layers can be constructed or laid; layers can be turned on or off, to reveal or hide details. Thus, a GeoTIFF of a topographical quad can be used as a base map upon which other layers can be plotted (viz: Miley\gis\vfic\o37079g4.tif).
Another realm, which I won't go into here and now, is raster GIS, in which the map layers are continuous variables --think of the 3-dimensional rendering of a digital contour map, or any other version of a Z-dimension representing a data surface. To keep things simple here I'll just show a couple of these as images and John Blackburn's House Mountain map.
These few maps are a tiny and not very sophisticated tip of the iceberg, long on unanswered questions and short on analysis, but we see several things nonetheless:
We have (in the data that came with ArcView) pretty good vector basemaps of the US --counties, zip codes, census tracts, geographic features like rivers, roads, etc.-- and a fair collection of kindred resources for other countries as well. Some areas (notably China) are especially well represented because of publicly-available datasets that I've downloaded. Via various US government sites we can access census data and a lot of broadly 'environmental' data --topographical, agricultural, hydrological, meteorological, etc. We have GeoTIFF (topographic map) imagery and TIGER files for most of Virginia, and could purchase more for other areas. I'd like to develop a coordinated collection of mapping resources centered on Rockbridge County and moving outward to provide mapping support for future projects in the Great Valley and surrounding mountains --an obvious ambit for Environmental Studies research. And, looking ahead more boldly, I'd like to build an archive of electronic maps, images and data to support a Global Studies program.
Teaching the use of ArcView is a bit more daunting. I've been writing web tutorials for the simple things as I've learned how to do them, and I'm in discussions with several faculty members in a pretty broad range of departments about mapping applications they'd like to develop. Clearly the Geology Department is where the existing GIS expertise resides, and I look forward to a Winter term course in which I'll certainly learn a lot. And I look forward to trying to figure out spatial analysis problems that anybody wants to bring to me. The way I figure it, this is just another important information medium, just another opportunity to develop practical analytical and communication skills.