Cycling and running with GPS

John

www.GPSrunner.net the new website for runners and walkers with GPS
systems

www.GPSbiker.net the new website for cyclists with GPS systems

Tom Kunich

"John" <news@bikecommute.com> wrote in message
news:4h2t80956jqe15b50j6ukrep9d49jpum2i@4ax.com...
> www.GPSrunner.net the new website for runners and walkers with GPS
> systems
>
> www.GPSbiker.net the new website for cyclists with GPS systems

Remember that GPS systems aren't particularly accurate as anything but maps.
For instance, they do not measure altitude gained very accurately because of
the angles involved don't make this dimension particularly accurate. And it
will read out distance in "as the crow flies". This is fine on flat terrain
by on routes with a lot of climbing you'll get errors in distance since the
GPS will not show the road covered but only the flat line distance. This
last weekend one of the riders had a GPS system and it produced a 2 mile
error on a 60 mile ride.

Richard Adams

"Tom Kunich" <cyclintom@yahoo.com> wrote in message news:<naXjc.16600$e4.971@newsread2.news.pas.earthlink.net>...
> "John" <news@bikecommute.com> wrote in message
> news:4h2t80956jqe15b50j6ukrep9d49jpum2i@4ax.com...
> > www.GPSrunner.net the new website for runners and walkers with GPS
> > systems
> >
> > www.GPSbiker.net the new website for cyclists with GPS systems
>
> Remember that GPS systems aren't particularly accurate as anything but maps.
> For instance, they do not measure altitude gained very accurately because of
> the angles involved don't make this dimension particularly accurate. And it
> will read out distance in "as the crow flies". This is fine on flat terrain
> by on routes with a lot of climbing you'll get errors in distance since the
> GPS will not show the road covered but only the flat line distance. This
> last weekend one of the riders had a GPS system and it produced a 2 mile
> error on a 60 mile ride.

Depends on the GPSr unit. I've got one of those Garmin ETrex Vistas
and it's pretty good, particularly for climbing as the sampling is
pretty accurate at the speed going up. Descending, OTOH, would work
better with more frequent sampling as a lot of ground is covered
between plot points and can miss turns (just cut across them), small
rises, etc. Same goes for the HAC4, though I don't believe the HAC4
allows changing the sampling interval.

Largest gripe about GPSr is it blanks out in canyons, heavily wooded
areas (notably among redwoods) and anywhere a 2.24GHz signal is
emitted for some wireless purpose.

They're a bit weighty, too, the ETrex Vista is one of the smaller and
lighter units with a handlebar mount available, but still about 6 Oz.

DirtRoadie

"Tom Kunich" <cyclintom@yahoo.com> wrote in message news:<naXjc.16600$e4.971@newsread2.news.pas.earthlink.net>...

> And it
> will read out distance in "as the crow flies". This is fine on flat terrain
> by on routes with a lot of climbing you'll get errors in distance since the
> GPS will not show the road covered but only the flat line distance. This
> last weekend one of the riders had a GPS system and it produced a 2 mile
> error on a 60 mile ride.

GPS inaccuracy is NOT the result of the difference between "slope
distance" and "map" or horizontal distance." That difference is
negligible for most routes that one might be travelling by bike. To
get a difference of 2 miles in 60 you would have to be on a very hilly
route having an average 26% grade for the entire distance. If you
finished where you started you would have climbed about 41,250
vertical feet.

That was quite ride wasn't it?
(Note that even on constant 8% grades for 60 miles the difference
between slope and horizointal distances would only be about .2 mi. a
tenth of what you described)

GPS errors in a recorded track are usually the result of momentary
loss of satellite reception for any of a number of reasons. But even
with this problem their accuracy is scary.

DR

Tom Kunich

"DirtRoadie" <DirtRoadie@aol.com> wrote in message
news:c6661314.0405040542.555e82e2@posting.google.com...
> "Tom Kunich" <cyclintom@yahoo.com> wrote in message
news:<naXjc.16600$e4.971@newsread2.news.pas.earthlink.net>...
>
> > And it
> > will read out distance in "as the crow flies". This is fine on flat
terrain
> > by on routes with a lot of climbing you'll get errors in distance since
the
> > GPS will not show the road covered but only the flat line distance. This
> > last weekend one of the riders had a GPS system and it produced a 2 mile
> > error on a 60 mile ride.
>
> GPS inaccuracy is NOT the result of the difference between "slope
> distance" and "map" or horizontal distance." That difference is
> negligible for most routes that one might be travelling by bike. To
> get a difference of 2 miles in 60 you would have to be on a very hilly
> route having an average 26% grade for the entire distance. If you
> finished where you started you would have climbed about 41,250
> vertical feet.
>
> That was quite ride wasn't it?
> (Note that even on constant 8% grades for 60 miles the difference
> between slope and horizointal distances would only be about .2 mi. a
> tenth of what you described)
>
> GPS errors in a recorded track are usually the result of momentary
> loss of satellite reception for any of a number of reasons. But even
> with this problem their accuracy is scary.

GPS measures the distance from point A to point B on a two dimensional
plane. If you have a climb in the interval it doesn't count the climbing
distance as part of the change in position. This means, for instance, that
there will be a 5% error climbing a 5% grade.

One descent was two miles of approximately 16% causing an error of about a
third of a mile. Of the 60 miles approximately 2/3rds was hilly or rolling.
Small errors build up rapidly. When I drew his attention to this error it
was on flat ground for the final 10 miles. Indeed there were no errors on
that return trip. The other two of us had mileage that agreed within 2/10ths
of a mile.

GPS errors caused by breaking lock would not cause significant errors if you
are proceeding in a more or less straight line on the two dimensional grid.

I think that you applied your math wrong. A 26% error in 60 miles is almost
16 miles.

The point isn't the accuracy of GPS which is about 10' on any position
between the artic and anartic latitudes, but whether the device is a
replacement for a bike computer. It isn't.

DirtRoadie

"Tom Kunich" cyclintom@yahoo.com wrote
>GPS measures the distance from point A to point B on a two dimensional
>plane. If you have a climb in the interval it doesn't count the climbing
>distance as part of the change in position.

This is true

> This means, for instance, that there will be a 5% error climbing a 5% grade.

This is not true. Your misunderstanding leads to the rest of your errors.

>One descent was two miles of approximately 16% causing an error of about a
>third of a mile.
Given a slope distance of two miles and a grade of 16%, the horizontal distance
is about 1.97 miles. That's a diiference of about 136 feet.
And and elevation difference of 1690 feet over that two miles.

> Of the 60 miles approximately 2/3rds was hilly or rolling.
>Small errors build up rapidly. When I drew his attention to this error it
>was on flat ground for the final 10 miles. Indeed there were no errors on
>that return trip. The other two of us had mileage that agreed within 2/10ths
>of a mile.

The accuracy over flat ground is more likely the result of constant satellite
reception.
If you have a course that measures 60 miles horizontally but 62 miles along the
slope, you can apply the Pythagorean thereom from that junior school math
(that you apparently slept through) to determine the "rise" i.e. change in
elevation.
In this case its the square root of 62 squared minus 60 squared
or using some progamming notation
SQR(62^2-60^2)
This is a rise of 15.6 miles (82,476 feet) in 60 horizontal miles (the "run")
or about 26%

>GPS errors caused by breaking lock would not cause significant errors if you
>are proceeding in a more or less straight line on the two dimensional grid.
>
>I think that you applied your math wrong. A 26% error in 60 miles is almost
>16 miles.

Yes but who said we have a 26% error? I was merely pointing out that it would
be a 26% grade. It's your mistake to equate grade with error.

>The point isn't the accuracy of GPS which is about 10' on any position
>between the artic and anartic latitudes, but whether the device is a
>replacement for a bike computer. It isn't.

No, but it is can still be amazingly useful (substitute your own word) if you
like to tinker with data and numbers. I don't know of a bike computer that will
let you record a ride and then view a graphical representation of the ride
overlaid on a map or aerial view (in 3-D with some software) plus see data at
any point along the way. The biggest limitation I see for use of a GPS as a
general replacement for a cyclometer is that the GPS units are still pretty
large (even the tiny ones) and the battery life is too limited. The pretty much
require recharging or new batteries on a daily basis.

There is some great software (see: topofusion.com) that lets you record mutiple
rides over the same route and then playback a simulated race or comparison
between the rides.

DR

Kyle Legate

DirtRoadie wrote:
> "Tom Kunich" cyclintom@yahoo.com wrote
>> GPS measures the distance from point A to point B on a two
>> dimensional
>> plane. If you have a climb in the interval it doesn't count the
>> climbing distance as part of the change in position.
>
> This is true
>
>> This means, for instance, that there will be a 5% error climbing a
>> 5% grade.
>
> This is not true. Your misunderstanding leads to the rest of your
> errors.
>
You shouldn't question Tom's math, he's an engineer! Math is all he does. At
least, he used to be an engineer; now he's posting like he's "unemployed"
again.