CompuServe Messages

Greg: See my previous post ( and the ones of last week, if they haven't scrolled yet). The "accuracy" with which you can measure a baseline is dependent on the "precision" ( spread of a number of repeat readings, expressed in Standard Deviations, or in error ellipse size/ orientations, or in error ellipsoids, or as a drms or 2drms figure. Definitely NOT expressed as a CEP {circular error probable}) and the "reliability" ( probability of detecting a measurement error of a given size). This accuracy itself relates to the measurement "frame of reference" which you are working in, Within any particular reference frame, 2 points will have "exact" mathematical separations in lat/long/ht, or in XYZ. There is an infinite number of possible reference frames, though. These are usually based on models simplifying the shape of the earth to an ellipsoid. The true "base" shape is a geoid, which is mathematically irregular, On top of that is the topography, which you rockhounds want to chip away until it actually confroms to the shape of the geoid, as far as I can see <g>. One point to note is that a pouint can have an *infinite* number of latitude/longitude coordinates… each one based on a different reference frame. Eg, the GPS WGS 84 reference datum places the Greenwhich meridian about 10 m west (? I think so… it might be east, but I haven't got my calculator here) of the meridian located by the OSGB (Ordnance Survey of Gt Britain) datum. Rob