| Author |
Message |
Dave
Guest
|
 Posted:
Thu Nov 24, 2005 7:06 pm Post subject:
HELP ! UTM errors ? (Distance between two UTM's) |
|
|
I was out walking with Vista and came to location 'exactly' 2.00km from
another UTM, when I got back I did some simple trig and I made the distance
shorter ! here's the numbers
UTM location 'A'
31 T 332519 5047402
UTM location 'B'
31 T 332826 5049326
Both with WGS84 Datum's
subtracting one from the other comes to XY of 307,1924 using simple trig
comes out as 1948.339 meters 9.06 degrees
Where I'm I going wrong to get a ~50 meter error !!!
Dave. |
|
| Back to top |
|
 |
Dave
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
"David Lee" <davidlee_malvern@dont.use.this.bit.hotmail.com> wrote in
message news:E8CdnXVZU_m5rBreRVnyrA@eclipse.net.uk...
| Quote: | Dave wrote...
My Garmin Vista and Mapsource made the distances 2KM +- a couple and I
make
it 1948m from the numbers mentioned, just seemed a large error ? - think
you
must be right and the rounding errors mount up !
Are you trying to compare the distance derived from the coordinates of the
start and end points with the distance your Vista tells you that you
actually travelled? If so then I think that agreement within 2.5% over
this
distance is pretty good!
It's highly unlikely that you were walking in a perfectly straight line so
that the distance travelled will always be greater than the separation of
the points. On top of that are errors in the coordinates of the start and
end points. The Garmin spec for your Vista's accuracy is <15m typical -
however in less than perfect conditions the errors can be much larger than
this and can be very different at two different points due to different
reception conditions giving yo a net error between your start and end
points. Bear in mind that the accuracy that your Vista reports on the
satellite page is a statistically estimated precision and cannot directly
be
interpreted as the error bounds of your position, which can easily be very
much greater. I have frequently seen real positional errors of the order
of
3 times the reported EPE. In addition fluctuations in satellite reception
can add extra wandering to the reported total distance, as the estimated
position fluctuates around your actual track during your walk. Taken all
together these errors could easily account for a couple of percent of the
direct distance between the two points measured by your Vista.
Try examining your tracklog data to see exactly where the GPS thought that
you went betwen the two points.
David
|
No not walking but in a straight line (as the crow fly's) both
using Mapsource's measure and Vista's 'Navigation' window both say 2 km, try
it! ....Oh
well looks like as you EPE and rounding errors...
Both locations are more or less the same altitude !
Dave... |
|
| Back to top |
|
|
Jack Erbes
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
Dave wrote:
| Quote: | I was out walking with Vista and came to location 'exactly' 2.00km from
another UTM, when I got back I did some simple trig and I made the distance
shorter ! here's the numbers
|
Was that the distance was 2.00km determined from a sign? If so, there
difference is in the rounding off of the value.
And if the sign showed it to two decimal places, the sign designer was
conceptually handicapped. It should have been either 2km or 1.95km
Sometimes interstate highways with mile markers on the shoulders are the
best place to get comparative measurements.
Jack
--
Jack Erbes in Ellsworth, Maine, USA - jackerbes at adelphia dot net
(also receiving email at jacker at midmaine.com) |
|
| Back to top |
|
|
David Lee
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
Dave wrote...
| Quote: | My Garmin Vista and Mapsource made the distances 2KM +- a couple and I
make
it 1948m from the numbers mentioned, just seemed a large error ? - think
you
must be right and the rounding errors mount up !
|
Are you trying to compare the distance derived from the coordinates of the
start and end points with the distance your Vista tells you that you
actually travelled? If so then I think that agreement within 2.5% over this
distance is pretty good!
It's highly unlikely that you were walking in a perfectly straight line so
that the distance travelled will always be greater than the separation of
the points. On top of that are errors in the coordinates of the start and
end points. The Garmin spec for your Vista's accuracy is <15m typical -
however in less than perfect conditions the errors can be much larger than
this and can be very different at two different points due to different
reception conditions giving yo a net error between your start and end
points. Bear in mind that the accuracy that your Vista reports on the
satellite page is a statistically estimated precision and cannot directly be
interpreted as the error bounds of your position, which can easily be very
much greater. I have frequently seen real positional errors of the order of
3 times the reported EPE. In addition fluctuations in satellite reception
can add extra wandering to the reported total distance, as the estimated
position fluctuates around your actual track during your walk. Taken all
together these errors could easily account for a couple of percent of the
direct distance between the two points measured by your Vista.
Try examining your tracklog data to see exactly where the GPS thought that
you went betwen the two points.
David |
|
| Back to top |
|
|
Phil Wheeler
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
Dave wrote:
| Quote: | I was out walking with Vista and came to location 'exactly' 2.00km from
another UTM, when I got back I did some simple trig and I made the distance
shorter ! here's the numbers
UTM location 'A'
31 T 332519 5047402
UTM location 'B'
31 T 332826 5049326
Both with WGS84 Datum's
subtracting one from the other comes to XY of 307,1924 using simple trig
comes out as 1948.339 meters 9.06 degrees
Where I'm I going wrong to get a ~50 meter error !!!
|
I agree with your calculation. But how do we know the 'exactly' 2.00 km
is really 2000 meters? What is your basis for that number (which you
stated to the nearest 10 meters, while doing you calculation of 1948 to
the nearest meter)? If measuring on a 7.5' topo, a meter is a very
small distance on a ruler.
Phil |
|
| Back to top |
|
|
Rick Scott
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
(Dave <me@me.com> uttered:)
| Quote: | I was out walking with Vista and came to location 'exactly' 2.00km
from another UTM, when I got back I did some simple trig and I made
the distance shorter ! here's the numbers
UTM location 'A'
31 T 332519 5047402
UTM location 'B'
31 T 332826 5049326
Both with WGS84 Datum's
subtracting one from the other comes to XY of 307,1924 using simple
trig comes out as 1948.339 meters 9.06 degrees
Where I'm I going wrong to get a ~50 meter error !!!
|
I don't suppose the two points have a different altitude?
About 450m in altitude difference will give you your 50 extra metres.
At first I thought it might have been the curvature of the earth
coming into play, but the difference between the chord length and
arc length is only about 4m.
Cheers,
Rick
--
key CF8F8A75 / print C5C1 F87D 5056 D2C0 D5CE D58F 970F 04D1 CF8F 8A75
A mathematician is a machine for turning coffee into theorems.
:Paul Erdos |
|
| Back to top |
|
|
David Lee
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
"Dave" <me@me.com> wrote in message news:dm6hl3$qql$1@news.tiscali.fr...
| Quote: | I was out walking with Vista and came to location 'exactly' 2.00km from
another UTM, when I got back I did some simple trig and I made the
distance
shorter ! here's the numbers
UTM location 'A'
31 T 332519 5047402
UTM location 'B'
31 T 332826 5049326
Both with WGS84 Datum's
subtracting one from the other comes to XY of 307,1924 using simple trig
comes out as 1948.339 meters 9.06 degrees
Where I'm I going wrong to get a ~50 meter error !!!
|
Sounds like a rounding error to me - how did you determine that the distance
was EXACTLY 2.00km?
The planar separation of of your points is indeed 1948m (please don't quote
insignificant decimal digits!) and entering the coordinates into a route on
my own Vista gives a leg distance of 1.95km.
David |
|
| Back to top |
|
|
Dave
Guest
|
| Posted:
Fri Nov 25, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
"David Lee" <davidlee_malvern@dont.use.this.bit.hotmail.com> wrote in
message news:eaGdnUzB2ovIYRvenZ2dnUVZ8qOdnZ2d@eclipse.net.uk...
| Quote: | "Dave" <me@me.com> wrote in message news:dm6hl3$qql$1@news.tiscali.fr...
I was out walking with Vista and came to location 'exactly' 2.00km from
another UTM, when I got back I did some simple trig and I made the
distance
shorter ! here's the numbers
UTM location 'A'
31 T 332519 5047402
UTM location 'B'
31 T 332826 5049326
Both with WGS84 Datum's
subtracting one from the other comes to XY of 307,1924 using simple trig
comes out as 1948.339 meters 9.06 degrees
Where I'm I going wrong to get a ~50 meter error !!!
Sounds like a rounding error to me - how did you determine that the
distance
was EXACTLY 2.00km?
|
My Garmin Vista and Mapsource made the distances 2KM +- a couple and I make
it 1948m from the numbers mentioned, just seemed a large error ? - think you
must be right and the rounding errors mount up !
| Quote: |
The planar separation of of your points is indeed 1948m (please don't
quote
insignificant decimal digits!) and entering the coordinates into a route
on
my own Vista gives a leg distance of 1.95km.
David
|
Dave |
|
| Back to top |
|
|
David Lee
Guest
|
| Posted:
Fri Nov 25, 2005 11:58 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
Dave wrote...
| Quote: | No not walking but in a straight line (as the crow fly's) both
using Mapsource's measure and Vista's 'Navigation' window both say 2 km,
try
it! ....Oh
well looks like as you EPE and rounding errors...
Both locations are more or less the same altitude !
|
Now I see what you mean!
The answer is it's the usual crap standard of Garmin software!
Indeed MapSource gives the separation as 2.0km - note NOT 2.00km as you
reported - and that's significant.
The distance reported by MapSource is rounded to one decimal place - ie
nearest 100m. Since the real distance is 1.948km the value displayed should
be 1.9km. It would appear that MapSource may be working internally to two
places of decimals - using the extra decimal place as a "guard digit" to
minimize rounding errors. Whilst the idea would be correct the number of
guard digits is not sufficient since the internal representation would round
to 1.95km and then rounding to 2.0km on display.
If you import exactly the same waypoints into GPS Trackmaker and create a
track between them then the length is correctly given as 1.948km.
I can't speak for the navigation display on the Vista since I'm about 765km
away from your waypoints! However the Route Page display on my instrument
correctly gives a leg length of 1.95km for a route including your two
points. For a length about 2km the Vista should give a distance on the
Navigation display to two decimal places and I would expect this also to be
1.95km. If you get 2.00 and you have not made any errors reporting the UTM
coordinates then the discrepancy is yet another Garmin bug. Check your
receiver and let us know what values you see for the leg distance on the
Route definition page and exactly what the navigation display was telling
you to see if your unit is internally consistent.
Cheers
David |
|
| Back to top |
|
|
Dave
Guest
|
| Posted:
Sat Nov 26, 2005 4:00 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
"David Lee" <davidlee_malvern@dont.use.this.bit.hotmail.com> wrote in
message news:pYydnVy2q7pdzRreRVnyuw@eclipse.net.uk...
| Quote: | Dave wrote...
No not walking but in a straight line (as the crow fly's) both
using Mapsource's measure and Vista's 'Navigation' window both say 2 km,
try
it! ....Oh
well looks like as you EPE and rounding errors...
Both locations are more or less the same altitude !
Now I see what you mean!
The answer is it's the usual crap standard of Garmin software!
Indeed MapSource gives the separation as 2.0km - note NOT 2.00km as you
reported - and that's significant.
The distance reported by MapSource is rounded to one decimal place - ie
nearest 100m. Since the real distance is 1.948km the value displayed
should
be 1.9km. It would appear that MapSource may be working internally to two
places of decimals - using the extra decimal place as a "guard digit" to
minimize rounding errors. Whilst the idea would be correct the number of
guard digits is not sufficient since the internal representation would
round
to 1.95km and then rounding to 2.0km on display.
If you import exactly the same waypoints into GPS Trackmaker and create a
track between them then the length is correctly given as 1.948km.
I can't speak for the navigation display on the Vista since I'm about
765km
away from your waypoints! However the Route Page display on my instrument
correctly gives a leg length of 1.95km for a route including your two
points. For a length about 2km the Vista should give a distance on the
Navigation display to two decimal places and I would expect this also to
be
1.95km. If you get 2.00 and you have not made any errors reporting the
UTM
coordinates then the discrepancy is yet another Garmin bug. Check your
receiver and let us know what values you see for the leg distance on the
Route definition page and exactly what the navigation display was telling
you to see if your unit is internally consistent.
Cheers
David
I can report that the Navigation window and the route window 'correctly' |
reports 1.95km so the 'error' is in the measure distance tool on the Vista,
guess I'm 'dividing rabbits' or is it splitting hairs ;-) I would have
thought Garmin would have these sort of bugs sorted !!
So anything less ~100 meters I won't assume the Garmin can tell the
difference !!!
Looks like your in the Midlands Dave ?
Dave. |
|
| Back to top |
|
|
Phil Wheeler
Guest
|
| Posted:
Sat Nov 26, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
Dave wrote:
| Quote: |
I can report that the Navigation window and the route window 'correctly'
reports 1.95km so the 'error' is in the measure distance tool on the Vista,
guess I'm 'dividing rabbits' or is it splitting hairs ;-) I would have
thought Garmin would have these sort of bugs sorted !!
So anything less ~100 meters I won't assume the Garmin can tell the
difference !!!
|
Dave,
I would not be too quick to agree with the Garmin-bashing troll.
How did you determine that the difference between the two points was
2.000 km (exactly 2000 meters) in the first place? If measured on a
7.5' (24,000:1 scale) map, a measurement error of 10 meters would be a
very small distance on a ruler:
o 1 km = 41.7 mm
o 100 m = 4.2 mm
o 10 m = 0.4 mm
Of course, a more precise map would let you measure more accurately
(assuming the map itself is not in error).
Or you could have based your 2.000 km on signposts -- not a good source
of info.
Certainly there will be GPS measurement errors. But a net error of 50
meters in two measurements seems excessive to me. I've used Garmins for
a long time and, since the cessation of Selective Availability, I've not
seen errors of that magnitude.
Phil |
|
| Back to top |
|
|
Dave
Guest
|
| Posted:
Sat Nov 26, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
"Phil Wheeler" <w6tuh-ng4@yahoo.com> wrote in message
news:Qq_hf.83363$QM5.907@tornado.socal.rr.com...
| Quote: | Dave wrote:
I can report that the Navigation window and the route window
'correctly'
reports 1.95km so the 'error' is in the measure distance tool on the
Vista,
guess I'm 'dividing rabbits' or is it splitting hairs ;-) I would have
thought Garmin would have these sort of bugs sorted !!
So anything less ~100 meters I won't assume the Garmin can tell the
difference !!!
Dave,
I would not be too quick to agree with the Garmin-bashing troll.
How did you determine that the difference between the two points was
2.000 km (exactly 2000 meters) in the first place? If measured on a
7.5' (24,000:1 scale) map, a measurement error of 10 meters would be a
very small distance on a ruler:
o 1 km = 41.7 mm
o 100 m = 4.2 mm
o 10 m = 0.4 mm
Of course, a more precise map would let you measure more accurately
(assuming the map itself is not in error).
Or you could have based your 2.000 km on signposts -- not a good source
of info.
Certainly there will be GPS measurement errors. But a net error of 50
meters in two measurements seems excessive to me. I've used Garmins for
a long time and, since the cessation of Selective Availability, I've not
seen errors of that magnitude.
Phil
More the other way around - The Vista told me that when I was at point "B" |
that point "A" was 2Km away I logged the location of point 'B'. When I got
back I did some maths and came up with the 1.95km simple.I've been using GPS
since about 1994 when a I got my first a Garmin GPS75 the vista is my second
GPS!
Dave.. |
|
| Back to top |
|
|
Phil Wheeler
Guest
|
| Posted:
Sat Nov 26, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
Dave wrote:
| Quote: | "Phil Wheeler" <w6tuh-ng4@yahoo.com> wrote in message
news:Qq_hf.83363$QM5.907@tornado.socal.rr.com...
Dave wrote:
I can report that the Navigation window and the route window
'correctly'
reports 1.95km so the 'error' is in the measure distance tool on the
Vista,
guess I'm 'dividing rabbits' or is it splitting hairs ;-) I would have
thought Garmin would have these sort of bugs sorted !!
So anything less ~100 meters I won't assume the Garmin can tell the
difference !!!
Dave,
I would not be too quick to agree with the Garmin-bashing troll.
How did you determine that the difference between the two points was
2.000 km (exactly 2000 meters) in the first place? If measured on a
7.5' (24,000:1 scale) map, a measurement error of 10 meters would be a
very small distance on a ruler:
o 1 km = 41.7 mm
o 100 m = 4.2 mm
o 10 m = 0.4 mm
Of course, a more precise map would let you measure more accurately
(assuming the map itself is not in error).
Or you could have based your 2.000 km on signposts -- not a good source
of info.
Certainly there will be GPS measurement errors. But a net error of 50
meters in two measurements seems excessive to me. I've used Garmins for
a long time and, since the cessation of Selective Availability, I've not
seen errors of that magnitude.
Phil
More the other way around - The Vista told me that when I was at point "B"
that point "A" was 2Km away I logged the location of point 'B'. When I got
back I did some maths and came up with the 1.95km simple.I've been using GPS
since about 1994 when a I got my first a Garmin GPS75 the vista is my second
GPS!
|
Sounds to me that you really have no "truth points" in your experiments
... and are using multiple measurements to check one another. Poor
experimental process.
Phil |
|
| Back to top |
|
|
Dave
Guest
|
| Posted:
Sat Nov 26, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
"Phil Wheeler" <w6tuh-ng4@yahoo.com> wrote in message
news:Tl%hf.21598$2k6.5702@tornado.socal.rr.com...
| Quote: | Dave wrote:
"Phil Wheeler" <w6tuh-ng4@yahoo.com> wrote in message
news:Qq_hf.83363$QM5.907@tornado.socal.rr.com...
Dave wrote:
I can report that the Navigation window and the route window
'correctly'
reports 1.95km so the 'error' is in the measure distance tool on the
Vista,
guess I'm 'dividing rabbits' or is it splitting hairs ;-) I would have
thought Garmin would have these sort of bugs sorted !!
So anything less ~100 meters I won't assume the Garmin can tell the
difference !!!
Dave,
I would not be too quick to agree with the Garmin-bashing troll.
How did you determine that the difference between the two points was
2.000 km (exactly 2000 meters) in the first place? If measured on a
7.5' (24,000:1 scale) map, a measurement error of 10 meters would be a
very small distance on a ruler:
o 1 km = 41.7 mm
o 100 m = 4.2 mm
o 10 m = 0.4 mm
Of course, a more precise map would let you measure more accurately
(assuming the map itself is not in error).
Or you could have based your 2.000 km on signposts -- not a good source
of info.
Certainly there will be GPS measurement errors. But a net error of 50
meters in two measurements seems excessive to me. I've used Garmins for
a long time and, since the cessation of Selective Availability, I've not
seen errors of that magnitude.
Phil
More the other way around - The Vista told me that when I was at point
"B"
that point "A" was 2Km away I logged the location of point 'B'. When I
got
back I did some maths and came up with the 1.95km simple.I've been using
GPS
since about 1994 when a I got my first a Garmin GPS75 the vista is my
second
GPS!
Sounds to me that you really have no "truth points" in your experiments
.. and are using multiple measurements to check one another. Poor
experimental process.
Phil
|
You're properly right ! If that a ham callsign I see mines G6ZKC (all lived
in US for a while!)
Dave.... |
|
| Back to top |
|
|
Dave
Guest
|
| Posted:
Sat Nov 26, 2005 5:08 pm Post subject:
Re: HELP ! UTM errors ? (Distance between two UTM's) |
|
|
| To much wine! (SP) ;-) |
|
| Back to top |
|
|
|
|
|
|
FAQ
Memberlist
Register
Profile
Log in to check your private messages
Log in