NavList:

John Karl, you wrote:
"Expressed differently, we use all the CN accuracy perpendicular to the new LOP, while retaining all the previously estimated positional info along the new LOP. We use all available info, make no unreasonable assumptions, and make no contradictions."

But John, this is where you get into trouble. This is a HAND-WAVING argument. What is your mathematical justification for picking a perpendicular point? What are the lengths of the axes of the new error ellipse? The diagrams you've been drawing are suggestive, but they are incomplete.

You also wrote:
"How can we do better than that??"

By doing it right.

-FER
----------------------------------------------------------------
NavList message boards and member settings: www.fer3.com/NavList
Members may optionally receive posts by email.
To cancel email delivery, send a message to NoMail[at]fer3.com
----------------------------------------------------------------