----------------------------Original message----------------------------
>Date:         Thu, 17 Aug 1995 10:27:11 EDT
>From: Charles Gilley <[log in to unmask]>
>Subject:      Computer generated maps - convention question
>
>----------------------------Original message----------------------------
>I am in the process of implementing map projection software in an
>application under development and have come across an issue that
>doesn't affect physical maps.  Many of the map projections which exist
>greatly reduce scale distortion when plotting map coordinates to
>physical media.  Everyone knows that the Mercator projection wildly
>distorts the vertical axis - but by design.  On the other hand, the
>oblique-stereographic projection has distortion as a function of
>deviation from the Standard Circle, Line of True Scale.  The Standard
>Circle, which has a center point, which also happens to be the center
>point of the O-S map projection poses a dilemma for computer maps.
>What is the center of the map?
>
>As I said above, a physical map is fixed.  But with software, all
>confusion is possible :).  I can zoom and pan.  Is the center of my map
>the visible center of my display?  Or, do I tie it to the center of
>the original map?  It's a rather minor distinction, but I'm curious as
>to how the professional map people's perspective.
>
>regards,
>
>--------------------------+------------------------------------------
>Charles Gilley            |   All comments are my own...
>Senior Software Engineer -|-
>AEL/Cross Systems         |   Cross Systems - We make your radar
>404-475-3633              |   see double, triple, or more.
>--------------------------+------------------------------------------
 
There is a problem of commenting on the above discussion because it
reflects naivete on the problem of cartographic projections.
 
First, a map represents the results of the transformation of the
surface of a sphere (or ellipsoid) to a plane surface.  This
transformation is based upon mathematical processes executed by
computer software (in precomputer days, a combination of laborious
calculations and mechanical processes were involved, but the results
are identical).  Thus, to say there is some difference between the map
and the computer software does not make much sense for a variety of
reasons.
 
The transformation process or "map projection" can be made by an
infinite number of methods, but there is no "perfect" way to flatten
the Earth onto a plane surface.  Snyder estimates that there are at
least 400 odd projections (over 100 of them are available in the PROJ.4
system on charon.er.usgs.gov).  But what has happened, is that there
are a few "pretty good" methods that are more frequently used---each
with a particular characteristic that is to be adhered to or enhanced.
There is not room here to even begin to adequately discuss this subject.
 
A problem with Mr. Gilley's query is that there is not sufficient
information to make suggestions.  For example, in a un-zoomed state is
the view the world or a smaller region?  If one starts with a world map
and then zooms in to a city size region or smaller view, the be choice
would be using at least two projections: one for the world and another
for larger scale maps.  Even then, the question of what projection
should be used will need to be based upon matters such as
equal-area/conformal and appropriatness for the geographic region.
 
But if one starts with a limited area, like a state, and zooms in
from there, one could easily get by with one projection and
projection parameters.
 
The problem is not finding a projection to use, but to get a handle
on the overall problem.
 
Gerald (Jerry) I. Evenden   Internet: [log in to unmask]
voice: (508)563-6766          Postal: P.O. Box 1027
  fax: (508)457-2310                  N.Falmouth, MA 02556-1027