GPS accuracy - I see not much has changed

Those familiar with me over the years may recall my rants on GPS
accuracy a few years back - especially WRT elevation calculations. See
my, "GPS Accuracy?" page for an example. (See:
http://tinyurl.com/ymcvty )

Anyways, I decided to treat myself to a new Magellan Roadmate 1470 as
an early X_mas gift to myself. I decided to check the elevation
reading and lo and behold, it said that I was at an elevation of 40 to
50 feet as I drove past the salt marsh (basically at sea level) on US
17. In the years since I bought that hand held unit, I see not much
has changed except for the fancy displays etc...

--
necromancer - ECHM

On Wed, 18 Nov 2009 20:20:10 -0500, necromancer rg> wrote:
>Those familiar with me over the years may recall my rants on GPS
>accuracy a few years back - especially WRT elevation calculations. See
>my, "GPS Accuracy?" page for an example. (See:
>http://tinyurl.com/ymcvty )

>Anyways, I decided to treat myself to a new Magellan Roadmate 1470 as
>an early X_mas gift to myself. I decided to check the elevation
>reading and lo and behold, it said that I was at an elevation of 40 to
>50 feet as I drove past the salt marsh (basically at sea level) on US
>17. In the years since I bought that hand held unit, I see not much
>has changed except for the fancy displays etc...

Do you think having a fancier unit is going to put more satelites in
the sky?

necromancer wrote:
> Those familiar with me over the years may recall my rants on GPS
> accuracy a few years back - especially WRT elevation calculations. See
> my, "GPS Accuracy?" page for an example. (See:
> http://tinyurl.com/ymcvty )
>
> Anyways, I decided to treat myself to a new Magellan Roadmate 1470 as
> an early X_mas gift to myself. I decided to check the elevation
> reading and lo and behold, it said that I was at an elevation of 40 to
> 50 feet as I drove past the salt marsh (basically at sea level) on US
> 17. In the years since I bought that hand held unit, I see not much
> has changed except for the fancy displays etc...
>
> --
> necromancer - ECHM
I need to write a book on the adventures of following a Garmin.
I think they have reintroduced the error and a lot more.
It has tried to take be through old Indian fords, stage coach roads,
down sidewalks, goat trails, around several blocks and back on the
route, off bridges, down rivers and creeks, through fields, said we were
in fields, while we were on 5 decade old roads,told us turn a quarter
mile after th intersection, under water along a lake shore and so on. I
really like 60 and 100 mile out of the way tours also.

On Wed, 18 Nov 2009 21:45:23 -0500, elmer > wrote:

>I need to write a book on the adventures of following a Garmin.
>I think they have reintroduced the error and a lot more.
>It has tried to take be through old Indian fords, stage coach roads,
>down sidewalks, goat trails, around several blocks and back on the
>route, off bridges, down rivers and creeks, through fields, said we were
>in fields, while we were on 5 decade old roads,told us turn a quarter
>mile after th intersection, under water along a lake shore and so on. I
>really like 60 and 100 mile out of the way tours also.

Here's one to get you started:

http://www.ananova.com/news/story/sm_3063945.html

--
"I... Can't drive... FIFTY-FIVE!!"
--Sammy Hagar

On Wed, 18 Nov 2009 20:22:11 -0600, AZ Nomad
> wrote:

>Do you think having a fancier unit is going to put more satelites in
>the sky?

No, but I would have expected that in the years since I bought that
first unit that their calculations of elevation above mean sea level
to have improved.

--
"I... Can't drive... FIFTY-FIVE!!"
--Sammy Hagar

> In the years since I bought that hand held unit, I see not much
> has changed except for the fancy displays etc...

None of the several sources of error have changed, nor has the fact
that elevation is the hardest GPS measurement. A technical workup of
why this is so, I leave to some good websites you can find -- it is
long and gives me a headache! Suffice it to say that GPS elevation
measurement (well, flat-map measurement too, though the problem is not
as severe) is only going to get so good unless you bring in a high-
quality outside datum.

There are additional canned data sets that can be combined with
measurements to improve GPS elevation readings, as well as ground-
based and accurately surveyed-in "augmentation" schemes for the
approaches to airports and for lesser improvement across wider areas.
Some units incorporate a barometric altimeter as well (how you
compensate for the well known sources of error in *that*, I don't know
-- probably depends on the unit). The more money and space you can
throw at the problem, the more of those you can incorporate -- at one
extreme, you can precision-land airplanes with augmented GPS; at the
other extreme, it is generally considered unwise to trust consumer-
grade GPS receivers for accurate elevation readings when there's
something at stake.

Horizontal accuracy should be a lot better, and that's what most
drivers are most interested in. That and usability, extra features
not strictly part of the GPS scheme, etc. -- information display is no
small thing, especially when you're on the move.

Cheers,
--Joe

On Nov 18, 8:20*pm, necromancer
rg> wrote:
> Those familiar with me over the years may recall my rants on GPS
> accuracy a few years back - especially WRT elevation calculations. See
> my, "GPS Accuracy?" page for an example. (See:http://tinyurl.com/ymcvty)

That is h i l a r i o u s.

In 1962 USGS horizontal tolerance for 7.5 minute map data was ±100m -
and- based on the NAD27 datum. The default datum for your Explorist
200 is WGS84.
http://www.magellangps.com/assets/ma...ist_200_en.pdf

http://www.maptools.com/UsingUTM/mapdatum.html

> In the years since I bought that hand held unit, I see not much
> has changed except for the fancy displays etc...

But not your willful, nearly perfect and unabashed ignorance. It is
amusing to imagine you heading out to a library for information.

http://www.esri.com/news/arcuser/0703/geoid1of3.html
http://gps.afspc.af.mil/gpsoc/
http://edu-observatory.org/gps/height.html

http://www.ngs.noaa.gov/PC_PROD/DDPR...tion.html#A.12

That ought to keep you busy informing those with education, training
and experience that they don't have a ****ing clue for some time to
come.
-----

- gpsman

necromancer wrote:
> Those familiar with me over the years may recall my rants on GPS
> accuracy a few years back - especially WRT elevation calculations. See
> my, "GPS Accuracy?" page for an example. (See:
> http://tinyurl.com/ymcvty )
>
> Anyways, I decided to treat myself to a new Magellan Roadmate 1470 as
> an early X_mas gift to myself. I decided to check the elevation
> reading and lo and behold, it said that I was at an elevation of 40 to
> 50 feet as I drove past the salt marsh (basically at sea level) on US
> 17. In the years since I bought that hand held unit, I see not much
> has changed except for the fancy displays etc...
>

A couple of things that I know of work their way into this.

First, the published error of the box is not just a circular error but a
spherical error. That means that positions collected by the box will
fall anywhere on or within that sphere. Just because there is a surface
running through the sphere doesn't mean that the positions will migrate
to the surface. If the error of the box is 50 ft, that error will
usually include a vertical component, not just a horizontal one.

Second, all of the readings are coming from the same hemisphere. You
can't get a reading from a satellite below the horizon. This tends to
exacerbate the problem in the vertical direction. If all the readings
are a little short or long (for whatever reason), this will move the
position up or down. This tends to turn the error sphere into more of
an ellipsoid(?) with the long access through the center of the earth.

This follows the same principle as the one in celestial navigation
where, when choosing stars for a 3 star fix, you want to choose them
from all around the sky, preferably as close to 120 degrees of azimuth
as possible. If your sextant is reading a little high or low, your
triangle may get a little bigger or smaller, but the center doesn't move
around. If you choose stars from all in the same part of the sky and
you're reading high or low, your triangle is not only changing size, but
it's moving around, too, giving you a bad fix.

--Andy

In article >,
necromancer > wrote:
>Anyways, I decided to treat myself to a new Magellan Roadmate 1470 as
>an early X_mas gift to myself. I decided to check the elevation
>reading and lo and behold, it said that I was at an elevation of 40 to
>50 feet as I drove past the salt marsh (basically at sea level) on US
>17. In the years since I bought that hand held unit, I see not much
>has changed except for the fancy displays etc...

And, as I've explained before, this is an expected limitation of the
system. Did you really expect the geometry of the constellation to
change?

There have been two big changes in accuracy and precision with
consumer GPS units. One was when the number of channels was greatly
increased, first by Garmin and then by others. This allowed more
satellites to be used in the solution and thus provide a better
solution. The second was the elimination of Selective Availability by
President Clinton.

There has been one change since then which will allow more accuracy
and precision; wide area differential GPS. However, unless my
information is out of date, the satellites capable of that have little
coverage and most of it is on water. Many GPS units can use it if
available though.

Some consumer GPS units are going to 5Hz sampling (versus 1Hz); this
will provide more current data and likely better precision, if not
much more in the way of accuracy.

The next big change will be if the second civilian signal is made
available (I haven't followed developments so I don't know if it is on
track). This will allow ionosphere and troposphere corrections to be
made directly by the GPS units, greatly reducing a major source of error.

As to altitude... consider how GPS works. A satellite directly above
you gives you no information on your horizontal position, and maximum
information on your vertical position. A satellite on the horizon is
the opposite. So for getting your horizontal position, you'd like all
the satellites to be scattered about your visible horizon. For
getting your vertical position, you'd like them to be above and
bel... err, wait, the ones below you aren't going to help either. And
that's why GPS is not as good for altitude.
--
The problem with socialism is there's always
someone with less ability and more need.

"Matthew Russotto" > wrote
> There has been one change since then which will allow more accuracy
> and precision; wide area differential GPS. However, unless my
> information is out of date, the satellites capable of that have little
> coverage and most of it is on water. Many GPS units can use it if
> available though.

Not quite right. It's WAAS, Wide Area Augmentation System. This
system uses around a dozen ground systems that cover N. America
and send out corrections for atmospheric variations. It allows guaranteed
30-meter (IIRC) accuracy. This is used by airplanes equipped with GPS
and flight-control software to accurately follow constrained landing paths.
All of Alaska Airlines planes are equipped, and they use the RNAV
capabilities to fly into valleys on pre-defined paths; and into Washington
National because of it's security corridor.

Many hand-held GPSs have WAAS capabilities, but it hardly matters.

FloydR