[time-nuts] Positional accuracy of the M12+T

Poul-Henning Kamp phk at phk.freebsd.dk
Thu Jan 4 05:07:11 EST 2007


In message <bfipp2d774ahdfucser5lapmc76m1cj0tg at 4ax.com>, Rex writes:
>On Wed, 03 Jan 2007 22:18:58 +0000, "Poul-Henning Kamp"
><phk at phk.freebsd.dk> wrote:
>
>>But for a band closer to the poles, from roughly 66 to 56 latitude,
>>where we have no sats in half the plane and only occasionally pick
>>up signals across the polar hole in the constellation, the ellipsoide
>>actually isn't one, and its axis are not aligned with the coordinates
>>we care for.
>
>You say, "the ellipsoide actually isn't one".  I don't understand what
>that might mean. Care to elaborate?

It is not an ellipsoide, it is some other weird ballon-animal shape.

>To be more specific: Say at the equator, some GPS reciever determines
>its lat/lon position within +- 3 meters but height +- 10 meters. If you
>move the same receiver to Sweden, I assume the position accuracy gets a
>bit worse. Does the height accuracy change by the same relative delta or
>a different one.

Well, Sweden is a bit unspecific because it has a large N/S span,
but for Denmark, yes: the receiver will do a lot worse.

In practice the vertical uncertainty in Denmark is 10 times the
horizontal.

>While we are at it, for a decent quality positional GPS receiver, what
>fraction worse would be the typical height accuracy vs the position at
>some mid latitude like north US or mid Europe?

I would expect 3x to 5x worse, but that is just a qualified guess.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.



More information about the time-nuts mailing list