BoatBanter.com - View Single Post - October 26

Mark Borgerson

In article ,
says...
As to a GPS anchor watch I note these two incidents:

A couple of years ago my wife and I took a walk down an old power line
that was over grown but had a trail. I took my pocket GPS with me and
logged the track. The outgoing and return tracks were off by at least
150 yards in places. Not big jumps but just gradually wandering off.
As I was walking a very narrow path through otherwise dense forest I
know I was traveling the exact same path.

You've stated the problem pretty well: "very narrow path through
otherwise dense forest". GPS signals don't penetrate
well through dense forest canopies.

A couple of days later I unintentionally left the GPS plugged into my
car overnight with the track turned on. The next morning I had a
"scatter plot" for my track covering a couple of hundred yards.
Was the GPS inside the car, or did the antenna have a clear view
of the whole sky? If the only clear signals were coming in
through the windshield, it was probably only using 2 or
three satellites to obtain a fix.

With many of my GPS logger projects, I regularly see about 5 meters
drift in elevation but only 2 to 3 meters of horizontal drift.
Of course that stability only applies in the continental US, where
the WAAS corrections are effective.

The GPS (Magellan) has AIS and has been dead accurate in every other
instance. I use it to find deer stands in the dark when hunting - going
out in the morning before sunrise.

The point is that GPS SYSTEM, though phenomenally accurate MOST of the
time can fail SOME of the time. I have thought of this when using the
anchor alarm. I might be dragging or the GPS SYSTEM might be brain farting.

I think that if your GPS does not have a well-placed antenna, you will
often end up in situations where it is receiving signals from only
two or three satellites. When that happens, the horizontal dilution
of precision can increase drastically.

A final point, oft made by Nigel Calder, not all charts are normalized
to the same datum. For example, through first hand experience, and as
noted on the charts, the South coast of Newfoundland charts Lat/Lon can
vary by as much as 1/2 mile from the GPS position. It is not a simple
conversion factor, the CGS does not know what the origional datum was.
It baffles me how the CGS allows this to continue but it does.

If you are interested I took a picture of my plotter which I posted here.

http://ph.groups.yahoo.com/group/sho...os/browse/bd9e

Mark Borgerson