Re: [Paddlewise] GPS errors

Bob Denton wrote (regarding GPS discrepancies of 3 minutes of lat/long):
> 
> I suspect that the GPS may have been set up on a different Datum, perhaps
> from an offshore cruise in foreign waters.

Again, this can't explain errors of 3 minutes of latitude and longitude,
even if the offshore cruise was to very distant lands using a highly
aberrant datum. Consider the following example. The original post was
from southern California, so let's use as our example the Santa Monica
pier, and set a waypoint at this position. The pier lies at N 34deg
00min, W 118deg 30min using the WGS84 datum. (Actually this is a spot
near the pier.)  Changing the datum changes the reported position of the
waypoint, but not by very much (as we will see). 

First let's examine the "usual suspects", such as NAD27 (CONUS and
Mexico). These cause a difference in reported latitude of about .001
minutes (33d 59.999min and 33d 00.001min, respectively) and a difference
of less than .06 minutes of longitude (118d 29.946min and 118d
29.953min). We can get a larger error if we choose an aberrant datum
from a far-away location (we are assuming here a very distant foreign
cruise, and a failure to reset the GPS upon returning home). Checking a
variety of such datums--from Afgooye to Zanderij--shows that the larger
deviations for the position of our Santa Monica waypoint are in the
range of .1 to .3 minutes, with somewhat larger discrepancies (still
only a fraction of a minute) occurring in the case of a few exceptions.
The largest discrepancies that I was able to find are for the
Hu-Tzu-Shan datum and for the Tristan Astr. '68 datum. (Wouldn't a
cruise to Tristan be fun? It is one of the most remote places on earth.)
The Hu-Tzu-Shan datum yields a reported waypoint position of 34d
00.373min, 118d 29.808min.  The Tristan datum yields a reported position
of 34d 00.294min, 118d 29.504min. In terms of total distance, the
champion is therefore Tristan Ast. '68.  However this is still only a
fraction of the reporting discrepancy of 3 minutes lat and long that was
the focus of the original question. 

In short, I think that we can rule out an incorrectly set datum as the
cause of such large errors.  So that brings us back to other possible
explanations, such as multipath error, a "broken" receiver, or (as
someone else mentioned) interference from other electronic devices.

Dan Hagen

***************************************************************************
PaddleWise Paddling Mailing List
Submissions:     paddlewise_at_lists.intelenet.net
Subscriptions:   paddlewise-request_at_lists.intelenet.net
Website:         http://www.gasp-seakayak.net/paddlewise/
***************************************************************************