Originally Posted by ft
looking at Kayton and Fried, I see an figure from an avionics text book.... illustrating varius concepts and terms used, rather than a strict definition.
The strict definition, I think, follows from Kayton's definitions of Vnorth and Veast along with equations (2.5) (and reference 2.10 for confirmation of this interpretation.
I am generally wary of quoting definitions from engineering sources and taking them as "definitive", for in my experience many engineering definitions are messy and ultimately inappropriate.
See, for one of the most fundamental examples in safety-critical systems engineering,
Problems Calculating Risk via Hazard. See
this essay on the International Functional Safety Standard IEC 61508 for some (notorious) problems with the definitions there, and
Definitions For Safety Engineering for my suggestion of how to do it right. (Full disclosure: I wrote all of these.)
I read Heflin as wisely leaving it open which definition to use. For many purposes, they will agree, especially when the margins of error due to non-uniformities of the geoid are taken into account.
You make the point that in geodesy, one references to what Kayton calles the "astronomic latitude", the direction of apparent gravity (where the force vector points where you are). That may be (I wouldn't know). But the WGS84 system is referenced to an ellipsoid. Kayton says that national ellipsoids may differ, and there are conversion tables. The differences may amount to hundreds of feet. He also points out (p23) that any ellipsoid is appropriate for "the navigator" as long as there is a one-to-one mapping between its surface and the earth's surface (also p23).
I think you are right about the hobgoblin!
PBL