NavList:

Herbert, you wrote:

"Paul Hirose has surely found the most sophisticated way of landing in
the reefs with the combined assistance of GPS _and_ a sextant! A GPS
receiver normally shows distance between two points along the geodesic.
For the points in the given example the answer using WGS84 is 2229 nm,
being 3 miles off the GCD. In a worst case scenario, (e.g. establishing
latitude at N45 from the sun at winter solstice) the discrepancy is 20
miles!
In order to use your GPS receiver as a makeshift sightreduction
computer, you have to cajole it into using a spherical earth model. "

 

Wow. Very interesting (btw, for those of you who were dismissive of HP's point, I would like to make clear that I, at least, think it's worth considering).

 

Near as I can tell, you're saying that new (or new-ish) GPS receivers calculate geodesics on the Earth as shortest distance arcs on the ellipsoid. If that's so, then yes, Paul Hirose's suggestion will land you on the reefs. But is this so? I would think that it would depend very much on the software in the device. Considering the relatively small importance of this issue in practical circumstances, it would not surprise me a bit if many GPS devices calculated distances as spherical great circle distances. And if they do, then yes, you can treat them, just as well, as sight reduction calculations. It's a good trick, when it works. I suppose we on the list could do a good "service" to the navigating community by coming up with a decent test case... Got one?? Can we try it out empirically??