|
|
| GeoCommunity Mailing List |
| |
| Mailing List Archives |
| Subject: | Re: [gislist] But what the hell can a GIS really DO? |
| Date: |
05/16/2005 07:35:01 AM |
| From: |
David T. Hughes |
|
|
A number of folks have talked about probability analysis within GIS as part
of this, so let me jump in and give you all a precis of what I'm looking
forward to accomplishing -- again -- using computer-based GIS this time.
My doctoral research began with a drive. I'm an archeologist. A friend and
I were playing the driving game of "find the sites" where we would debate
whether or not a given spot on the landscape did or did not have an
archeological site on it. I routinely succeeded about 80% of the time, and
finally my friend asked the deadly question: "How are you able to know so
well." The rest, as they say is history.
In the succeeding years, I analyzed a large area of the southern plains
(USA) for patterns of distribution of archeological sites in relation to
soils, vegetation, fauna, water availability and quality, drainage basin
conformation, and about 60 other factors. At the time all the data were
available on a bewildering array of paper maps at differing scales and
projections, so I spent about years at a drawing board re-scaling all the
maps to a common scale. Then began looked at the known archeological sites
dating from A.D. 1000 to A.D. 1500 and compared them to the settings I had
been mapping. When THAT was finished, I did a series of overlays of
intersections on the rest of the factors for all the sub-drainage basins
(about 35,000 of them) and compared them on the 60 environmental attributes
I'd been mapping. From that I was able to determine about 8 combined
factors which seemed to account for some 85% of the known archeological
sites. Then, to check my work, I went back to the databases and predicted
where more like them could be found and did some walking.
Time consuming, tedious, painful (anyone ever spent 2 years at 12 hours a
day on a drawing boards?) but immensely rewarding. Computers at the time
(mid-1980s) were barely up to handling the line print statistics with
cross-tabulations and ANOVA and MANCOVA I was working with, but the result
was an excellent predictive model that would account for most of the
archeological sites of the types we knew about, and the ones that were not
predicted had unusual features to them.
What the hell can a GIS really DO? Well, for the past two years I've been
studying hard on the applications of GIS to predictive modeling at all
levels in hopes of repeating that now quite dated research in a much more
usable and repeatable medium. It lay on the shelf for the past 20 years
because the work was too time consuming and tedious to be anything more than
an academic exercise. With GIS in my corner, I believe that I can finally
operationalize the model in the real world.
Why would anyone care? Well, as part of the Environmental Policy Act, the
National Historic Preservation Act is invoked and any federal agency or
corporation operating under federal permits, licenses, or with federal
funds, must take into consideration the historic and cultural resources
(including archeological sites) that may lie within their area of proposed
effect. Since the passage of EPA that has often meant either a scientific
guess about what might be in the area of effect or that some poor soul has
to go out and physically walk the entire area on parallel traverses 10 to 30
meters apart to insure that there is nothing there (or that what is there is
given protection). It seems that some kind of predictive modeling might be
most useful and could be economically productive.
Why have I not done it before now? I defended my dissertation the month
BEFORE ESRI released its first version of PC-ArcInfo. GRASS was for the
big-time computer geeks with lots of money. Since that time, I've been
building seniority and waiting for the technology to become more affordable
and powerful, obviating the need for a cost-per-minute mainframe hookup. We
are an impoverished department in a small midwestern school. But stand by
-- things are on the move at last.
That's what I HOPE to accomplish with GIS. Any other wild plans out there?
David T. Hughes, Ph.D.
Associate Professor of Anthropology
Wichita State University
Wichita, KS 67260-0052
david.hughes@wichita.edu
_______________________________________________
gislist mailing list
gislist@lists.geocomm.com
http://lists.geocomm.com/mailman/listinfo/gislist
_________________________________
This list is brought to you by
The GeoCommunity
http://www.geocomm.com/
|
|
Sponsored by:
For information
regarding
advertising rates
Click Here!
|
