--- Begin Forwarded Message ---
Date: Tue, 05 Feb 2002 16:41:10 -0500
From: Tsering Wangyal Shawa <[log in to unmask]>
Subject: Re: geographical center of the continents <fwd>
Sender: Tsering Wangyal Shawa <[log in to unmask]>



Joe McCollum has sent me personal email saying that his result is different
from mine in calculating centroid of South American continent and almost
the same message was sent to Maps-L but without name.

I thought it is important to inform our reader why the value was different.
I did not get a chance to inform in time because I was down with a bad flu.

Let me explain why our result may be different: I decided to do a quick and
dirty calculation of center of the continents using ESRI ArcWorld data 1:3M
(units in decimal degrees and ellipsoid in WGS 1984) and sent that
information to Alice because she could give the information to a guy who
was bugging her. The goal of the calculation is not to get an exact or
absolute central point of the continent but to get a relative position, and
an idea of how that type of information can be acquired by using a GIS tool.

Why I decided to use ESRI ArcWorld data.....this data has a column
containing continent information. Based on continent's information, I
grouped all the countries into each continent and then calculated the
centroid. This means, calculation was not just based on one simple polygon
but group of polygons. In South America, there are 763 polygons. When I
grouped them, it did not create just one contiguous polygon but created a
sort of region that includes a big one contiguous polygon with a lots of
small polygons (islands) which are outside the main continent. The centroid
was calculated on this region.

It is understandable that a projection, a scale of a map will play a role
in off setting a position of a centroid. As we have learn in our
cartography courses that when the earth is projected on a flat surface, the
earth is distorted (directions, distances, areas and shape). So, a centroid
value calculated from a projected map will be different then one which is
not projected.

Any other comments and suggestions will be appreciated.

Thanks for bring this issue.

-Wangyal
Tsering Wangyal Shawa
Geographic Information Systems Librarian
Head, Digital Map and Geospatial Information Center
Geosciences and Map Library
Guyot Hall, Princeton University
Princeton, NJ 08544
Phone: (609) 258-6804
Fax: (609) 258-4607
www.princeton.edu/~geolib/gis



--- End Forwarded Message ---