|
Robert,
I’m not really saying anything about the ‘correct’ area of a shape, but the response from Oracle points to a difference in the way the systems perform computations. Having said that, it does seem that MapInfo is possibly taking something of a ‘short-cut’ in area calculation – spherical trig is easier that ellipsoidal trig.
I may be wrong but I think the area of similar dimensioned objects on the surface of an ellipse varies according to where on that ellipse they are situated – the ‘curvier’ bit or the ‘flatter’ bit. According to my rudimentary knowledge of ellipsoidal geometry they may even vary according to the shape and orientation of the object. All this may be at play in the discrepency you have found between total area and sum of constituent parts area.
If you would like to forward me some simple sample data off-list (that shows the discrepancy), I would be happy to do a comparison of results between the systems.
Cheers David
-----Original
Message-----
David,
Thanks for your explanation. But now I am confused. Are you saying that the correct answer for the area of a one km square piece of ground is 0.997 sq km?
This is probably related to the issue that I had a problem with, in that if you split the 1 km square polygon, calculated the proportion of the original area covered by each piece, and then summed up the proportion of those areas, you would expect to get 1, but actually got 0.997.
To a surveyor this may make sense.
R
From:
[EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Llewellyn, David
Dear List, A response you may find interesting regarding a question I had earlier about different areas being reported by MapInfo and Oracle Spatial: My original question related to the area being reported for a Mapinfo table (Lat/Long GDA94) containing a single square shaped polygon of 1km sides centred at 141 degrees East, 22 degrees South. In MapInfo the Area()
or SphericalArea() function applied to this object
returns: “Oracle's result
is smaller than the MapInfo result by a factor of .997. This is precisely the
ratio expected if MapInfo is returning spherical area using the authalic sphere
for this GRS80 ellipsoid. This authalic sphere has radius 6371007 meters. The
Oracle result is an ellipsoidal area, corrected for meridional and prime normal
radii of curvature at the average location of the polygon. At -22 degrees
latitude ( gleaned from an earlier post), this ellipsoidal correction is
precisely a factor of
Best regards,
|
_______________________________________________ MapInfo-L mailing list [email protected] http://www.directionsmag.com/mailman/listinfo/mapinfo-l
