No subject

[MICHAEL A. IVIE (by way of dyanega at denr1.igis.uiuc.edu Doug]

,
              Yanega)" <ueymi at TREX.OSCS.MONTANA.EDU>
Subject:      Re: latitude and longitude designation and collection data

[This was intended to go out to the list, but came to me, and Mike said to
go ahead and forward it - DY]

There are several problems and solutions that occur to me as I read the
various listings on this subject.  First, Doug Yanega's idea of putting
all the lat./long. data on the label is unnecessary if unique identifiers
keyed to a database are used.  A relational database makes it necessary
to store the data in only on place.

Another solution is quite easy.  For general data, such as "Jamacia", the
central point of the polygon is used, and the data are but in the inexact
data field.  For actual points, the GPU data are stored in the lat./long.
field.  Only these can be used for GIS analysis, and should be identified
as such.  Another solution I have seen is to use a reading only to the
minute field for inexact data, and to the smallest reading used for the
exact data.  UTM's can be used in the same way.

A problem with using digitized maps is the problem of map quality.  For
over a hundred years, ever since the Challenger miscalculated the position
of the Fort Christian Base Line, most Caribbean maps have had incorrect
Latitudes.  This is a slight error, but enough to move sites for terrestrial
species out to sea when GPU data are compared with digitized map data.

So, you see, the idea of GIS is great, but the application is messy.  For
now, I strongly agree with Doug Y. that retrocapture of the millions of
specimen labels and wasting huge amounts of time trying to introduce
exactitude where it does not exist should not be done, for now.  For
specific projects where a product requires such data, I suggest taking the
US BOard of Geographic Names, or other equivalent, and using the reading
for the lowest rank geographic place name available as inexact data.  These
gazateers are available on tape, so the searches can be automated to some
degree.

Michael A. Ivie