Measurement of Elevation Gain

I went out on a metric yesterday....I had my Garmin Edge 305 with me. It
comes with Training Center Software. I also have accounts on
Motionbased.com and Routeslip.com. From Motionbased.com I can export a file
that I can then load into Routeslip.com. For the 65 mile course, this is
what these tools are claiming for the elevation gain:

Garmin Edge 305: 4025 ft
Training Center: 4770 ft
Motionbased.com: 5505 ft
Routeslip.com: 5130 ft

Which one is best to believe? I tend to take the word of the Edge, as it
was there with me on the ride and it's the lower number. Yet, I don't have
a clue if it is accurate or not, or how accurate it is likely to be. I did
do a ride the previous weekend which it measured at around 2600 ft and this
last ride definitely had more climbing (according to my legs and HR). So, in
a relative sense, the Edge seems to be making some sense. Of course, the
other tools also gave lower numbers than the ones above on that ride too.
So, relatively speaking, they seem to make sense.

But, as you can see, there is a different of ~1500 ft between the lowest and
highest estimate of elevation gain. Maybe I should take the average!

"Roger Zoul" wrote in
:
Garmin Edge 305: 4025 ft
Training Center: 4770 ft
Motionbased.com: 5505 ft
Routeslip.com: 5130 ft

Which one is best to believe? I tend to take the word of the Edge, as
it was there with me on the ride and it's the lower number.

A barometric altimeter (like in the Edge 305) will generally give you a much
more accurate cumulative elevation gain number compared to programs that use
free digital elevation maps.

"Roger Zoul" wrote in message
...

Routeslip.com: 5130 ft


My experience with these sorts of maps is that they often measure going up
and down a gorge, as opposed to you riding the bridge over the gorge. This
includes things like highway overpasses. So you often get a bigger elevation
gain/loss than is real.

--
Warm Regards,

Claire Petersky
http://www.bicyclemeditations.org/
See the books I've set free at: http://bookcrossing.com/referral/Cpetersky

"Roger Zoul" wrote in message
...
I went out on a metric yesterday....I had my Garmin Edge 305 with me. It
comes with Training Center Software. I also have accounts on
Motionbased.com and Routeslip.com. From Motionbased.com I can export a
file that I can then load into Routeslip.com. For the 65 mile course, this
is what these tools are claiming for the elevation gain:

Garmin Edge 305: 4025 ft
Training Center: 4770 ft
Motionbased.com: 5505 ft
Routeslip.com: 5130 ft

Which one is best to believe? I tend to take the word of the Edge, as it
was there with me on the ride and it's the lower number. Yet, I don't
have a clue if it is accurate or not, or how accurate it is likely to be.
I did do a ride the previous weekend which it measured at around 2600 ft
and this last ride definitely had more climbing (according to my legs and
HR). So, in a relative sense, the Edge seems to be making some sense. Of
course, the other tools also gave lower numbers than the ones above on
that ride too. So, relatively speaking, they seem to make sense.

But, as you can see, there is a different of ~1500 ft between the lowest
and highest estimate of elevation gain. Maybe I should take the average!


It is important to understand how you want to define elevation gained. The
amount of elevation gained depends on how small of a change in elevation you
want to add to the sum of elevation gained. To state it another way, it
depends on how much you want to smooth out the small elevation changes over
small periods of time.

If you travel over a bump up and down 6 inches and 12 inches long, do you
want to count that as 6 inches of elevation gained?

To give an absurd example, what if the road surface is level but bumpy,
perhaps a 1 mm bump every 10 mm of surface. If you count each of these
bumps as a 1 mm elevation gain, then you are gaining 100 meters elevation
every 1 kilometer of travel (if I kept the decimal places correct in my
calculation!)

To give another absurd example, suppose you only count variations of 50
meters or greater. You ride 10 kilometers over 25 hills that are each 49
meters high. You would count you elevation gained as 0.

Any method / device that measures elevation gained is going to give a
different result depending on how much it smooths out the smaller elevation
changes over time. Hence, it should be no suprise that the four ways you
measured elevation gained gave different results. The range of "correct"
values ranges from the difference between the starting elevation and the
finishing elevation to infinity depending on how much the method adds up
small variations in elevation over small periods of time.

BobT

On 2007-10-07, Claire Petersky wrote:

"Roger Zoul" wrote in message
...

Routeslip.com: 5130 ft

My experience with these sorts of maps is that they often measure going up
and down a gorge, as opposed to you riding the bridge over the gorge. This
includes things like highway overpasses. So you often get a bigger elevation
gain/loss than is real.

The other problem that you can run into is that those services usually
only measure the elevation at the points that you click to plot the
route. The elevation gain/loss is computed as if there is a constant
grade between the plotted points, so you lose the effect of any local
highs and lows. For instance, consider this 0.2 mile piece of my PM
commute. If I plot only the endpoints, I get no climb:

http://veloroutes.org/bikemaps/?route=4718

However, if I add a few points in between I get 38 feet:

http://veloroutes.org/bikemaps/?route=4719

I'm not trying to pick on veloroutes. The other mapping sites I've tried
do the same thing, and in any case I think it's a reasonable thing to
do. Making a web service call to get the elevation of a point can be
time-consuming, and you wouldn't want it done every few dozen feet if
you were plotting a cross-country tour. There are various ways to decide
how often to get elevation data over a route, but I think that using
the points plotted by the user is at least as good an approach as any.

Steve Gravrock wrote:
The other problem that you can run into is that those services usually
only measure the elevation at the points that you click to plot the
route. The elevation gain/loss is computed as if there is a constant
grade between the plotted points, so you lose the effect of any local
highs and lows. For instance, consider this 0.2 mile piece of my PM

There is in fact no correct answer. You can get any increasingly large
elevation gain by going to finer and finer scales.

What you'd want is some effort-weighted altitude gain. Tiny hills
take no effort, even if they add up to thousands of feet.

The actual effort effect is in the increased time you're working, though,
if you want to measure effort. Uphills slow you down, is all, and lower
your average speed more than downhills raise it. Gearing makes you
indifferent to uphill or downhill as to effort, except for the increased
time taken.
--


On the internet, nobody knows you're a jerk.

"Ron Hardin" wrote in message
...
Steve Gravrock wrote:
The other problem that you can run into is that those services usually
only measure the elevation at the points that you click to plot the
route. The elevation gain/loss is computed as if there is a constant
grade between the plotted points, so you lose the effect of any local
highs and lows. For instance, consider this 0.2 mile piece of my PM

There is in fact no correct answer. You can get any increasingly large
elevation gain by going to finer and finer scales.

Not if you have a suitable means to determine elevation gain over such small
scales. Of course, IR you won't be able to do that, it seems.


What you'd want is some effort-weighted altitude gain. Tiny hills
take no effort, even if they add up to thousands of feet.

Really? That doesn't seem to make sense to me. Of course, you'd have to
cover a lot of linear miles to rack up mile gain if you're just doing tiny
hills.


The actual effort effect is in the increased time you're working, though,
if you want to measure effort. Uphills slow you down, is all, and lower
your average speed more than downhills raise it.

yes.

Gearing makes you
indifferent to uphill or downhill as to effort, except for the increased
time taken.

Really? Then why do I feel more tired when riding lots of uphill miles vs.
downhill miles even with gearing?

On 2007-10-08, Ron Hardin wrote:

The actual effort effect is in the increased time you're working, though,
if you want to measure effort. Uphills slow you down, is all, and lower
your average speed more than downhills raise it. Gearing makes you
indifferent to uphill or downhill as to effort, except for the increased
time taken.

That has not been my experience at all. I find that steep rolling
terrain takes a lot more out of me than covering the same distance on
flat ground, despite the fact that I run considerably wider gearing than
most road bikers. Just how hard do you push on the flats, anyway?

Roger Zoul writes:

I went out on a metric yesterday....I had my Garmin Edge 305 with
me. It comes with Training Center Software. I also have accounts on
Motionbased.com and Routeslip.com. From Motionbased.com I can
export a file that I can then load into Routeslip.com. For the 65
mile course, this is what these tools are claiming for the elevation
gain:

Garmin Edge 305: 4025 ft
Training Center: 4770 ft
Motionbased.com: 5505 ft
Routeslip.com: 5130 ft

Which one is best to believe? I tend to take the word of the Edge,
as it was there with me on the ride and it's the lower number. Yet,
I don't have a clue if it is accurate or not, or how accurate it is
likely to be. I did do a ride the previous weekend which it
measured at around 2600 ft and this last ride definitely had more
climbing (according to my legs and HR). So, in a relative sense,
the Edge seems to be making some sense. Of course, the other tools
also gave lower numbers than the ones above on that ride too. So,
relatively speaking, they seem to make sense.

But, as you can see, there is a different of ~1500 ft between the
lowest and highest estimate of elevation gain. Maybe I should take
the average!

Unfortunately the Avocet altimeter is not in production although it
holds an important patent for accumulating altitude on a ride.
Without this feature most altimeters add extra elevation gain because
they have no hysteresis, the ability to reject small variation such as
atmospheric oscillation while riding on level ground and while just
riding on rolling terrain with less than 10 meters difference (the
built in hysteresis of the Avocet instrument).

Typically a rise in the road of less than 10 meters is not added.
However, the altimeter does not forget the last low point and when
subsequently rising higher than 10 meters from that datum and all
additional gains are accredited. It is designed to, for instance,
ignore elevation gained from encountering many RR underpasses that are
less than 10 meters deep on an otherwise dead level course. Those are
not climbs and one generally doesn't need to shift to low gears to
climb their, otherwise, fairly steep grades.

You may not agree with the method but it is the only one that gives
reasonable and repeatable totals for riding over mountain passes.

Jobst Brandt

Steve Gravrock wrote:
On 2007-10-08, Ron Hardin wrote:

The actual effort effect is in the increased time you're working, though,
if you want to measure effort. Uphills slow you down, is all, and lower
your average speed more than downhills raise it. Gearing makes you
indifferent to uphill or downhill as to effort, except for the increased
time taken.

That has not been my experience at all. I find that steep rolling
terrain takes a lot more out of me than covering the same distance on
flat ground, despite the fact that I run considerably wider gearing than
most road bikers. Just how hard do you push on the flats, anyway?

I thought that a metric century in the "driftless" area of SW Wisconsin
was about as difficult as 100 mile century in relatively flat central
Illinois.

--
Tom Sherman - Holstein-Friesland Bovinia
Beer - It's not just for breakfast anymore!