----------------------------Original message----------------------------
A couple of comments on your maps....
 
First, thank you for taking on an important project for all of us.  Second,
to produce appropriate choropleth maps (area fills), you should not use
interval-level data (number of pounds/population/facilities/etc.) unless
the enummeration units are of the same size (or close to the same).
Without relatively equal-sized areal units, your data should be
standardized (as a rate, ratio or percentage) before mapping; e.g.,
ratio-level data such as pounds per acre, or people per square mile
(population density).  Or, remove area from consideration by calculating a
rate (x per 1000; or x per 100, as with a percentage).  If you don't do
this first, the resulting maps may be visually very misleading.
 
Thank you for allowing me to use your maps for examples in this mini-forum,
an occupational hazard.  ;-)
 
Dr. Dennis Fitzsimons
Department of Geography & Planning
Southwest Texas State University
San Marcos, TX  78666-4616
telephone:  512.245.3200
e-mail: [log in to unmask]
 
 
>>From: "Michael R. Meuser" <[log in to unmask]>
>>Date: Tue, 2 Dec 1997 02:02:50 +0000
>>Subject: Re: GIS and  TRI data
>
>I am a ph.d. candidate in sociology at ucsc and have worked with tri
>data quite a bit.
>
>I also have a consulting firm and we have just completed a series of
>maps (6000+) covering all TRI facilities in the US.  These will be
>part of EDF's WWW scorecard project.  Each facility is hittable.  We
>ran into some of the same problems that you describe.  The GIS we
>used was not programmable (also not arcview) so we had to output
>the image (*.bmp) and an associated data file that included the
>preferred latitude and longitude.  We then developed an in-house
>program that would allow us to physically move and resize if
>necessary little boxes over each facility.  The program then output
>a file with the pixel coordinates of the top left and bottom right of
>the hittable box.  We then wrote another program that would write
>the html from these data and image files.  Now that we have completed
>the project we are working on both of the programs hoping to release
>them to the public in a month or so.
>
>In the meantime check out our work at:
>
>http://www.mapcruzin.com/EI
>(demographic/tri maps of silicon valley)
>http://www.mapcruzin.com
>(toxic maps including a link to our own Santa Cruz TRI)
>http://www.mapcruzin.com/meuser
>
>My ucsc email is [log in to unmask]
>
>Mike Meuser
>Clary-Meuser & Associates
>[log in to unmask]