Most accurate altimeter -- w/ exportable elevation profiles

Matt O'Toole said:

Who makes the most accurate cycling altimeter these days? I know the
old Avocet ones were great, but what about newer ones which can
export the data to a PC? The ones we've been fiddling with vary *a
lot* in their elevation data. So either some or all of them are off
by quite a bit. Is there even an accurate one being made these days?
Which one?

Matt O.

Cycling altimeters are barometric devices, depending upon the pressure of
the atmosphere for the altititude reading. Unless the altimeters are
constantly adjusted for the local pressure, they by their nature can't be
*that* accurate. When I use my cycling altimeter, I set my local altitude to
'zero' at the start, as I'm mostly interested in how much alititude I've
gained over the ride. Even then, localized pressure variations, such as a
cool valley, will give somewhat inaccurate readings.

"Matt O'Toole" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Who makes the most accurate cycling altimeter these days? I know the old
Avocet
ones were great, but what about newer ones which can export the data to a
PC?
The ones we've been fiddling with vary *a lot* in their elevation data.
So
either some or all of them are off by quite a bit. Is there even an
accurate
one being made these days? Which one?

GPS will fit the bill better with accuracy to within 5 or 8 feet with WAAS.

--
Phil, Squid-in-Training

Phil said:

"Matt O'Toole" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Who makes the most accurate cycling altimeter these days? I know the old
Avocet
ones were great, but what about newer ones which can export the data to a
PC?
The ones we've been fiddling with vary *a lot* in their elevation data.
So
either some or all of them are off by quite a bit. Is there even an
accurate
one being made these days? Which one?

GPS will fit the bill better with accuracy to within 5 or 8 feet with WAAS.

I use a GPS receiver on all my bike rides and have been happy with its
accuracy, but the above statement is overly optimistic. Under ideal
conditions WAAS can give horizontal position measurements accurate to
about 3m (10'😉 using a consumer-level receiver for a single measurement.
But altitude accuracy is somewhat less due to the inherent geometry:
best accuracy requires that the satellites be distributed in various
directions, so east-west accuracy is best if some are in the eastern
sky and some in the western; altitude would be best if some were above
and some below, but of course we can't receive signals from satellites
below us due to blockage by the earth. Bottom line is that altitude
accuracy is worse by about 1.5 - 2, so a single measurement would at
best be accurate to about 15 - 20'.

But that's under ideal conditions: an unobstructed horizon, no reflected
signals, etc. Typically on a bike ride a significant amount of the
sky is blocked by the rider's body and on hilly rides additional
blockage occurs due to the terrain. This also often blocks the WAAS
satellites so their correction data can't be used. Under conditions
where only three satellites can be received, the GPS will no longer
be able to determine altitude and generally assumes that it remains
constant until reception improves. My experience is that GPS by itself
can give a pretty accurate view of the elevation profile of a ride, but
calculations of cumulative elevation gain can be off considerably due
to random fluctuations even when riding on level ground.

Barometric and GPS altimeters provide a good complement to each other.
The barometric measurements have good short-term stability and are
unaffected by things like cliffs and dense tree cover that impair
GPS reception. But the barometric sensors are severely affected by
weather changes which occur on a longer time scale and can cause
errors of hundreds of feet. In comparison, the GPS measurements have
much greater short-term fluctuations, and are strongly affected by
anything that blocks the satellite signals.

The two methods can be combined to give better results than are possible
with either one. GPS is good when averaged over longer periods, like
30 minutes or more, while barometric sensors have better stability but
long-term drift due to changing weather. Using the time-weighted
average difference between the barometric and GPS readings to
automatically recalibrate the barometer takes advantage of the best
characteristics of both measurement methods.

Peter wrote:

(A bunch of stuff about GPS)

How hard is it to increase the accuracy of a consumer-level GPS unit by
sitting on a known point (USGS marker, ferinstance) and plugging in the
lat/lon?

--
BMO - former land survey tech

Boyle M. Owl said:

Peter wrote:

Quoted message said:

How hard is it to increase the accuracy of a consumer-level GPS unit by
sitting on a known point (USGS marker, ferinstance) and plugging in the
lat/lon?

If I understand your question, you're asking whether a more accurate
altitude could be obtained if you tell the GPS receiver its precise
location.

There are some receivers that allow the converse - the user can enter
a known altitude and thereby slightly increase the horizontal position
accuracy. Magellan is one manufacturer that has models with that
capability and it was most commonly used by ocean boaters who knew that
their altitude was pretty fixed at a few feet above sea level. The
improved accuracy with the demise of Selective Availability and the
introduction of WAAS has reduced the value of this feature.

But I don't know of any consumer-level GPS receiver that lets the user
specify the lat/long and use that for a more refined altitude
measurement. In principle that would be possible, but probably not of
much practical use.

Peter said:

Boyle M. Owl said:

Peter wrote:

Quoted message said:

How hard is it to increase the accuracy of a consumer-level GPS unit by
sitting on a known point (USGS marker, ferinstance) and plugging in the
lat/lon?

If I understand your question, you're asking whether a more accurate
altitude could be obtained if you tell the GPS receiver its precise
location.

There are some receivers that allow the converse - the user can enter
a known altitude and thereby slightly increase the horizontal position
accuracy. Magellan is one manufacturer that has models with that
capability and it was most commonly used by ocean boaters who knew that
their altitude was pretty fixed at a few feet above sea level. The
improved accuracy with the demise of Selective Availability and the
introduction of WAAS has reduced the value of this feature.

But I don't know of any consumer-level GPS receiver that lets the user
specify the lat/long and use that for a more refined altitude
measurement. In principle that would be possible, but probably not of
much practical use.

Dear Peter,

Speaking of practical versus principle, here's a truly
ignorant question.

What exactly is elevation measured from, anyway? High tide,
low tide, mean tide? Wave crest, wave trough, where the
surfboard was last seen?

That is, what is Pikes Peak 14,110 feet above?

Cluelessly,

Carl Fogel

Carl Fogel said:

What exactly is elevation measured from, anyway? High tide,
low tide, mean tide? Wave crest, wave trough, where the
surfboard was last seen?

Mean Sea Level (MSL). See:

http://www.esri.com/news/arcuser/0703/geoid1of3.html

Art Harris

"Matt O'Toole" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

Who makes the most accurate cycling altimeter these days? I know the old

Avocet

Quoted message said:

ones were great, but what about newer ones which can export the data to a

PC?

Quoted message said:

The ones we've been fiddling with vary *a lot* in their elevation data.

So

Quoted message said:

either some or all of them are off by quite a bit. Is there even an

accurate

Quoted message said:

one being made these days? Which one?

Matt O.

I have a ciclosport CM-436M ($210 @ Nashbar
http://www.nashbar.com/profile.cfm?sku=9010&SRCCODE=1067) with cadence
pickup (another $30). This unit downloads to a computer and the data can be
exported to a spreadsheet. The software is a bit of a puzzle but works just
fine once you figure it out. The altimeter works fine but suffers from the
same minor inaccuracies that all barometric altimeters do. The main error is
that from day to day as the weather changes, the altimeter should be
adjusted to compensate. In actual practice, I seldom do this since I am not
often interested in a actual elevation but in the gain/loss. All barometric
altimeters use a standard mathematical model for the relationship between
elevation, air pressure and temperature. This model is not perfect but it is
easily good enough for bike rides. I took mine on the Death Ride
(www.dethride.com) and during the day the elevation varied from about 5100
to 8700 feet while the temperature varied from 45 to 111 F (on the
handlebars) in 12.5 hours. After I got home, I checked elevations at various
points against USGS topo maps (www.topozone.com) and found a fairly linear
7.5% error between high and low points which is easy to compensate for in a
spreadsheet. At lower elevations, the error seems to be much less. The
elevation recorded at the start and the stop 12.5 hours later are 20 feet
different - pretty darn good I would say.

I also have a Garmin Forerunner 201 GPS wrist unit
http://www.garmin.com/products/forerunner201/ ($130) which I bought for
running but have taken on the bike from time to time. Garmin makes several
similar units some of which might be better suited to cycling. It provides a
speed readout, distance traveled and elevation. I have taken carried both
units together on a few rides and they agree very well on distance, speed
and elevation. It downloads to the computer but the data is not very handy
to other applications. Garmin could come up with better software. There may
be better software out there but I haven't looked for it. There is a web
site that will overlay your data onto a map http://www.gpsvisualizer.com/.

I seldom carry the GPS unit on a road bike as the 436 provides accurate
enough data. If you have the money, I would suggest the Ciclosport HAC4
since it has heart rate.

Art Harris said:

Carl Fogel said:

What exactly is elevation measured from, anyway? High tide,
low tide, mean tide? Wave crest, wave trough, where the
surfboard was last seen?

Mean Sea Level (MSL). See:

http://www.esri.com/news/arcuser/0703/geoid1of3.html

Yes, that link gives a good explanation of the differences between
Mean Sea Level (MSL), the ellipsoid surface, and the geoid surface.
However, it may be somewhat misleading in giving the impression
that GPS receivers will report height above the WGS84 ellipsoid.
While typical consumer GPS units will initially calculate that
height internally, most then apply a correction factor from a
lookup table to report the height above the geoid rather than
above the ellipsoid. So the reported altitude on most units
(including all Magellans and Garmins) should be quite close to
height above MSL at that location, but with some random measurement
fluctuations as described in my earlier post.

Matt O'Toole said:

Who makes the most accurate cycling altimeter these days? I know the old Avocet
ones were great, but what about newer ones which can export the data to a PC?
The ones we've been fiddling with vary *a lot* in their elevation data. So
either some or all of them are off by quite a bit. Is there even an accurate
one being made these days? Which one?

Any barometric altimeter will have a pretty large margin
of error. This is intrinsic in how they work. The only
thing to make sure is that it's temerature adjusted. The
other nice thing is ones that calibrate themselves to a
known elevation -- I use a Ciclosport HAC-4, and it has a
setting for your home altitude. Each time you start
recording, it calibrates its current altitude to this known
altitude, so the data are more accurate.

Larry Coon
University of California

Quoted message said:

I use a GPS receiver on all my bike rides and have been happy with its
accuracy, but the above statement is overly optimistic.

My experience has yielded results in this range.

Under ideal

Quoted message said:

conditions WAAS can give horizontal position measurements accurate to
about 3m (10'😉 using a consumer-level receiver for a single measurement.
But altitude accuracy is somewhat less due to the inherent geometry:
best accuracy requires that the satellites be distributed in various
directions, so east-west accuracy is best if some are in the eastern
sky and some in the western; altitude would be best if some were above
and some below, but of course we can't receive signals from satellites
below us due to blockage by the earth. Bottom line is that altitude
accuracy is worse by about 1.5 - 2, so a single measurement would at
best be accurate to about 15 - 20'.

I was actually trying to avoid the lengthy explanation of this part of GPS,
which is why I gave my figure, which was, in my opinion, rather
conservative.

Quoted message said:

But that's under ideal conditions: an unobstructed horizon, no reflected
signals, etc. Typically on a bike ride a significant amount of the
sky is blocked by the rider's body and on hilly rides additional
blockage occurs due to the terrain. This also often blocks the WAAS
satellites so their correction data can't be used. Under conditions
where only three satellites can be received, the GPS will no longer
be able to determine altitude and generally assumes that it remains
constant until reception improves. My experience is that GPS by itself

I used mine in my Camelbak and usually had a clear view of the sky.

Quoted message said:

can give a pretty accurate view of the elevation profile of a ride, but
calculations of cumulative elevation gain can be off considerably due
to random fluctuations even when riding on level ground.

I've never had fluctuations greater than 3 to 5 feet.

Phil, Squid-in-Training

Phil said:

Quoted message said:

I use a GPS receiver on all my bike rides and have been happy with its
accuracy, but the above statement is overly optimistic.

My experience has yielded results in this range.

On what basis are you judging the accuracy. I'd note that some
receivers do considerable filtering and averaging of the position
measurements so it isn't always easy to judge the actual accuracy
unless many observations are made at places that have been
accurately surveyed.

Quoted message said:

Under ideal

Quoted message said:

conditions WAAS can give horizontal position measurements accurate to
about 3m (10'😉 using a consumer-level receiver for a single measurement.
But altitude accuracy is somewhat less due to the inherent geometry:
best accuracy requires that the satellites be distributed in various
directions, so east-west accuracy is best if some are in the eastern
sky and some in the western; altitude would be best if some were above
and some below, but of course we can't receive signals from satellites
below us due to blockage by the earth. Bottom line is that altitude
accuracy is worse by about 1.5 - 2, so a single measurement would at
best be accurate to about 15 - 20'.

I was actually trying to avoid the lengthy explanation of this part of GPS,
which is why I gave my figure, which was, in my opinion, rather
conservative.

Hardly conservative since it indicates less variation than is expected
under the best possible conditions - totally unobstructed horizons and
an optimal position for the receiver antenna.

Dave Thompson said:

Quoted message said:

Who makes the most accurate cycling altimeter these days? I know
the old Avocet ones were great, but what about newer ones which can
export the data to a PC? The ones we've been fiddling with vary *a
lot* in their elevation data. So either some or all of them are
off by quite a bit. Is there even an accurate one being made these
days? Which one?

Quoted message said:

Cycling altimeters are barometric devices, depending upon the
pressure of the atmosphere for the altitude reading. Unless the
altimeters are constantly adjusted for the local pressure, they by
their nature can't be *that* accurate. When I use my cycling
altimeter, I set my local altitude to 'zero' at the start, as I'm
mostly interested in how much altitude I've gained over the ride.
Even then, localized pressure variations, such as a cool valley,
will give somewhat inaccurate readings.

That does not affect cumulative altitude gain since that depends only
on differences. What the altitude currently displayed is does not
affect the accumulator. Similarly, the things that are nice to know
are whether the rise ahead is higher or lower than the current
position. This is also unaffected by absolute altitude display.

Flutter rejection is Important in accumulating altitude gain, the 5 to
10 feet of atmospheric variation that occur often in an hour and the
slight rollers in flat terrain would otherwise indicate large gains.
For this, a rejection logic must be invoked, that remembers how high
and low the previous extremes were, otherwise flatlands accumulate
hundreds of feet of elevation in the course of an 8-hour ride.

Jobst Brandt
[email hidden]

Quoted message said:

Dave Thompson writes:

Quoted message said:

Quoted message said:

Quoted message said:

Who makes the most accurate cycling altimeter these days? I know
the old Avocet ones were great, but what about newer ones which can
export the data to a PC? The ones we've been fiddling with vary *a
lot* in their elevation data. So either some or all of them are
off by quite a bit. Is there even an accurate one being made these
days? Which one?

Quoted message said:

Cycling altimeters are barometric devices, depending upon the
pressure of the atmosphere for the altitude reading. Unless the
altimeters are constantly adjusted for the local pressure, they by
their nature can't be *that* accurate. When I use my cycling
altimeter, I set my local altitude to 'zero' at the start, as I'm
mostly interested in how much altitude I've gained over the ride.
Even then, localized pressure variations, such as a cool valley,
will give somewhat inaccurate readings.

That does not affect cumulative altitude gain since that depends only
on differences. What the altitude currently displayed is does not
affect the accumulator.

No, it doesn't - as was clearly recognized by the poster to whom you
responded. But the original post was not clear as to whether
cumulative gain was what they found to be inaccurate or if it was
the actual reported elevation or some other metric.

Quoted message said:

Similarly, the things that are nice to know
are whether the rise ahead is higher or lower than the current
position. This is also unaffected by absolute altitude display.

Surely that depends on how you got the information about the nature
of the rise ahead. If it was from looking at a map or guidebook and
seeing that the next pass tops out at 8700', then the actual current
altitude would be of interest for comparison and that would require
recent calibration of any barometric sensor for accurate results.

Larry Coon <lmcoon@nospam_cox.net> wrote in
news:4207D11F.2E38@nospam_cox.net:

Quoted message said:

Matt O'Toole said:

Who makes the most accurate cycling altimeter these days? I know the
old Avocet ones were great, but what about newer ones which can
export the data to a PC? The ones we've been fiddling with vary *a
lot* in their elevation data. So either some or all of them are off
by quite a bit. Is there even an accurate one being made these days?
Which one?

Quoted message said:

Larry Coon
University of California

I have National Geographic Topo software, and have downloaded my GPS tracks
into a map. It makes a really pretty topo map and profile with elevation
change. But, tree cover, cliffs and other obstacles cause me to lose
satellite signal, and Topo has no way to manually fix the errors.

Besides, I don't carry my GPS either. I'm reaching my minimalist stage of
cycling history. And, I like to talk about how hard the ride is, and maybe
stretch things a little bit to make me feel like I worked harder than I
really did! Ha ha ha!

"Dan" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

easily good enough for bike rides. I took mine on the Death Ride
(www.dethride.com) and during the day the elevation varied from about 5100
to 8700 feet while the temperature varied from 45 to 111 F (on the

Hey Dan; that would be http://www.deathride.com. "dethride.com" belongs to
a cheesy looking record company website.

I'm very much hoping that I can ride the (Markleeville) Death Ride next
year; employment, transportation and weather permitting.

H.

"HardwareLust" <[email hidden]> wrote in message
news:lsWNd.12877$uc.8864@trnddc05...

Quoted message said:

"Dan" <[email hidden]> wrote in message
news:[email hidden]...

Quoted message said:

easily good enough for bike rides. I took mine on the Death Ride
(www.dethride.com) and during the day the elevation varied from about

5100

Quoted message said:

Quoted message said:

to 8700 feet while the temperature varied from 45 to 111 F (on the

Hey Dan; that would be http://www.deathride.com. "dethride.com" belongs

to

Quoted message said:

a cheesy looking record company website.

I'm very much hoping that I can ride the (Markleeville) Death Ride next
year; employment, transportation and weather permitting.

H.

H-

I don't know, I think that I might like the dethride site better - seems to
have more content. I have some stuff here:
http://www.geocities.com/danmerrick/deathride2003/
I haven't yet, but am planning to sign up soon for this year. My wife and
one kid tell me they are up for it too - I doubt I finish if I ride with
them but what could be better?

Go heavy, go slow, go late... You'll know me, I will be the old, slow guy
with thin hair.

This excel spreadsheet comparison of diffferent altimeters is quite
interesting:

http://www.caltriplecrown.com/ClimbingDoubles.xls

I have the Polar 720i and find that it's cumulative altitude gain is
pretty decent. It does underestimate the altitude sligtly when going
from low to high elevation, say from 2000 ft. to 10,000 ft.

--
terry morse Palo Alto, CA http://bike.terrymorse.com/

Quoted message said:

That is, what is Pikes Peak 14,110 feet above?

Cluelessly,

Not so cluelessly as you may think. I believe it is from "Mean Sea
Level", a statistical measurment.

Cal