Hello Listers

I'm sure someone is well-informed about this non-MapInfo question, and can
give me a ready answer. 

Using the Vincenty FORWARD equation (compiled FORTRAN), a traverse as
follows misses its closure (origin) by over 3km - 

Start point P1: -25/120 (25degrees South, 120 degrees East)
P1-P2: 040deg, 240,000m
P2-P4: 160deg, 100,000m
P4-P6: 220deg, 120,000m
P6-P3: 220deg, 120,000m
P3-P1': 340deg, 100,000m
The end point (P1') has a calculated lat/long of -24d59'15.85606" /
119d58'21.49343"

Using the Vincenty INVERSE equation (compiled FORTRAN) P1-P1' gives a
separation of 3078.33m approx, forward azimuth about 296deg

The compiled programs I am using are regarded as reliable (see below), and
to my knowledge (reading the Vincenty paper and NGS / NOAA comments) should
be accurate for the purpose - certainly, a 3km closure error means
something's not right here!

For Australia, I choose the options 3,C which uses the Australian National
Ellipsoid (which has a = 6378160.0000, and 1/f = 298.250000). From Appendix
G in the MapInfo user manual, this is the same as the Ellipsoid 2
(Australian).


IL Thomas
GeoSciSoft - Perth, Australia 

Background on Vincenty formulae:
The "Vincenty formulae" were published in 1975, by Thaddeus Vincenty (an
American, I think), as a publication of the Directorate of Overseas Surveys
of the Ministry of Overseas Development (don't you love the old titles -
today it would be something jazzy and stupid like "Surveys UK"), as Survey
Review XXII, 176, April 1975, pp88-93. 

This paper was the basis for 4 FORTRAN 77 programs (INVERSE, FORWARD,
INVERS3D, FORWRD3D) last compiled in 2002 by Stephen J. Frakes (and several
others, including Robert Safford, back as far as 1985) at NGS/NOAA, and
available for download (compiled, and as source-code) from
http://www.ngs.noaa.gov/PC_PROD/Inv_Fwd/



---------------------------------------------------------------------
List hosting provided by Directions Magazine | www.directionsmag.com |
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
Message number: 11905

Reply via email to