There are some very good references in the mi-l archives (http://www.directionsmag.com/mapinfo-l/IndexMapInfoL.htm) to the accuracy of spherical distances in MapInfo. Search using "great circle" or aviation might narrow it down.
One quirk for programmers on this topic is: Because of the accuracy of and rounding that happens in the MB function SphericalDistance() and Distance(), if you have geometric computations comparing distances from a point to 2 closely spaced points using geographic coords, it is necessary to use pythagoras to separate them. (You'd think distance(), cartesiandistance(), sphericaldistance() might cover the field?!) Phil. ----- Original Message ----- From: "SCISOFT" <[EMAIL PROTECTED]> To: "'MapInfo List'" <[EMAIL PROTECTED]> Sent: Tuesday, June 01, 2004 6:17 PM Subject: MI-L Accuracy of MapInfo distances Does anyone know the accuracy of MapInfo's distances? There's a standard Australian surveyors' traverse in many (Oz) books and manuals - "Smeaton - Buninyong - Flinders Peak - Bellarine - Arthurs Seat". The GDA (Geocentric Datum of Australia) technical manual deals with a series of calculations using GDA94 and MGA94 (TM, zone55). (Chapter 10, p46 of the GDA Tech Manual shows the traverse and pertinent data.) MapInfo v6.5 can handle this datum and projection. I entered the Lat/Long values for these points onto a layer constructed with Lat/Long 1,116 (Australia GDA94) - apparently (1,116) is the same coordsys as (1,33) and also (1,74) since it displays in various MI tools as NAD83 or with 1,33 in the coordsys clause. Be that as it may, I saved each table as MGA94 / Zone55 which is appropriate to this traverse. The Cartesian length of Buninyong-Flinders Peak in MI is 54992.43 metres - almost identical to the GDA Tech Manual's Plane Distance of 54,992.279m (Chapter 6, page 27). However, the spherical distance is 54925.6 metres compared with the GDA Tech Manual's value of 54,972.271m. That's a large discrepancy. If anything, I would have thought the spherical distances would agree best, but they don't. The difference is -46.7m - far too much, I would have thought. Am I misunderstanding something here? IL Thomas GeoSciSoft - Perth, Australia --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 11976 --------------------------------------------------------------------- List hosting provided by Directions Magazine | www.directionsmag.com | To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] Message number: 11999
