Electronics, Lighting, & Gadgets - Are any pressure based altimeters accurate?

I received a Garmin handheld GPS for Christmas and have taken it on several rides. When compared to my buddies altimeters we always have different measurements even though we all calibrate at the start of the ride. Our end results on total altitude climbed can be off as much as 25%.

Another issue I have noted is that when a large truck or vehicle passes you and you get that air blast, it causes a big spike in altitude which immediately calms down after the turbulance dissipates. But the total altitude gained totalizes it as if it were climbed. I even drafted in a pace line on a flat road and saw an 18 foot rise in elevation due to the negative pressure created by the guys in front of me. I would get out of line and immediately lose the 18 feet. My buddies saw similar results.

Are there any truly accurate altitude devices out there? GPS would be inaccurate because of GPS drift. I could always take the transit but that would put me over the weight limit for my bike. :rolleyes:

GPS is about as accurate as you can get as long as your device can receive data from at least four satellites (more than four is even better). Other devices that rely on atmospheric pressure to measure altitude can vary widely as you have observed. Even altimeters in aircraft that rely on pressure have to be recalibrated by the pilot frequently using known field elevations or current ATC info.

I received a Garmin handheld GPS for Christmas and have taken it on several rides. When compared to my buddies altimeters we always have different measurements even though we all calibrate at the start of the ride. Our end results on total altitude climbed can be off as much as 25%.

Another issue I have noted is that when a large truck or vehicle passes you and you get that air blast, it causes a big spike in altitude which immediately calms down after the turbulance dissipates. But the total altitude gained totalizes it as if it were climbed. I even drafted in a pace line on a flat road and saw an 18 foot rise in elevation due to the negative pressure created by the guys in front of me. I would get out of line and immediately lose the 18 feet. My buddies saw similar results.

Are there any truly accurate altitude devices out there? GPS would be inaccurate because of GPS drift. I could always take the transit but that would put me over the weight limit for my bike. :rolleyes:

GPS drift? GPS is capable of sub-centimeter accuracy (if the receiver is).

Mine says it is normally accurate to 15 feet. When you sit still and it drifts like that I would think that a GPS based altimeter would be pretty inaccurate. I don't think the ones that are available to the average Joe even approach that level of accuracy.

Mine says it is normally accurate to 15 feet. When you sit still and it drifts like that I would think that a GPS based altimeter would be pretty inaccurate. I don't think the ones that are available to the average Joe even approach that level of accuracy.

Fifteen feet accuracy is pretty good for most applications unless you are doing precision bombing! The accuracy is a direct factor of how many satellites are in view and the relative strength of their signal.

Fifteen feet accuracy is pretty good for most applications unless you are doing precision bombing! The accuracy is a direct factor of how many satellites are in view and the relative strength of their signal.

Yeah - the super-accurate receivers (for research and surveying and such) use some extra equipment that picks up a ground-based signal from relay satellites in addition to the normal GPS birds, or so I read once. GPS is a really fascinating and highly-complex system. It has to take relativistic effects of the proper motion of the birds and their position further out in the earth's gravitational field in order to adjust the clocks. If the adjustments weren't done the system would become useless within one day due to the difference in clock rates. Einstein was the man.

For people who have the Garmin305 units, how long does it typically take you to lock onto a signal? For me sometimes it locks on right away and other times it takes a few minutes. In some cases, I've seen some inaccuracies in the GPS when it gets confused by tree cover. Is there a way to get it to lock onto more sateliites if you leave it out in the open longer before moving around with it?

My Blackburn Delphi is accurate enough with two caveats: You have zero out the altitude before riding to the home altitude (press a button, it's built for this) and keep in mind that if a strong weather front is moving in (or out) that it'll have an effect as well. It's not like I'm using it to pilot aircraft fer chrissakes.

The interesting thing about the Garmin 305 (only) is that although it is based on a barometric altimeter, it uses the GPS altitude data to automatically adjust its readings. It is fairly accurate for civilian uses like ours. I would not be too concerned about sub-meter accuracy with a device like this.

Pressure-style altimeters are really just barometers. I have some experience with these as I'm an old airplane pilot.

The barometer as applied as an airplane altimeter has a setting window on it where the altimeter is to calibrated to local atmospheric pressure. This is done two ways:

1. For a "local" flight, taking off and landing at the same airport, the altitude of the highest runway surface is published for each airport. Before takeoff, the pilot just turns the knob on the altimeter until the hands on the altimeter read correctly for that figure.

2. On a cross-country flight, when you contact the tower by radio, for instance, for landing instructions, the current barometric pressure is given to you. You then set that pressure in the "window" on the altimeter, which calibrates it for the local atmospheric pressure.

For people who have the Garmin305 units, how long does it typically take you to lock onto a signal? For me sometimes it locks on right away and other times it takes a few minutes. In some cases, I've seen some inaccuracies in the GPS when it gets confused by tree cover. Is there a way to get it to lock onto more sateliites if you leave it out in the open longer before moving around with it?

Mine takes less than a minute usually.

I don't think that the elevation gained readings from pressure based altimeters are any good. With mine I can stay in the same place for several minutes and the variations in pressure during that time will end up with an "elevation gain". In other words, if if the elevation reading were 986 feet, it would vary from say 980 to 992 feet and everytime it went up, you would add to the elevation gain total. If you multiplied that over the length of a ride, you'll get several additional vertical feet.

I do think they are pretty accurate if you're figuring out the elevation gain by figuring out the elevation difference at two points. I'd say plus or minus 25 feet.

IIRC, the Edge 305 primarily uses the barometric altimeter while the GPS altitude data maintains the calibration during use and time of pressure fluctuation.

With my Garmin 305 I find that the altimeter function is usually quite accurate, it's the cumulative elevation function that gets screwed up. Garmin Training Center and Motion Based each take the elevation data and compute cumulative elevation and get substantially different numbers, off by say 20%.

But if I'm climbing a hill and want to know what elevation I've reached, I think it's accurate say to plus or minus 25 ft.

Does anybody have a more accurate way or are we stuck with crude and inaccurate total elevation data until someone makes a real cumulative altimeter?

ummmm, i don't get it. how can getting passed by cars and drafting people have any effect on the air pressure. i've had a suunto watch for years now that uses air pressure to calculate elevation and it is always highly accurate on my totals using the log book feature. i use it skiing and cycling and very rarely need to recalibrate it when i use it often. sometimes it gets out of whack when it sits around not being used for extended periods but thats it.

Isn't the calculation of cumulative elevation gain sort of arbitrary?

You go up a 10-ft rise - do you add that to total elevation gain?

What about a 1-ft rise in elevation?

One inch? A half an inch? If you sum up all the "gains" in elevation you make over a total ride, you'll get an unrealistically big figure. Somehow you have to filter out insignificant noise and only count real climbing.

Isn't the calculation of cumulative elevation gain sort of arbitrary?

You go up a 10-ft rise - do you add that to total elevation gain?

What about a 1-ft rise in elevation?

One inch? A half an inch? If you sum up all the "gains" in elevation you make over a total ride, you'll get an unrealistically big figure. Somehow you have to filter out insignificant noise and only count real climbing.


Yeah, the horizontal increments used in the calculation are probably larger. Like lets say 100 yds or so.

If you have an eTrex or similar...here are a couple things to check. In Setup...go to Altimeter Setup and turn off "Auto Calibration". GPS is accurate but not AS accurate initially for elevation. The eTrex will continually try to calibrate itself initially which can take up to 30 minutes (I've read). That said, even a loop I might complete seems to be +/- 20 feet when I return to that starting point. Is this due to change in barometric pressure or just inaccuracy in the device. Probably the latter.

The reason pilots update their altimeter data "frequently" is because air pressure is a local phenomenon and they need to sync to the area they're flying in...right?

Edit: Tonight I checked the starting and ending ascent/descent totals...after a 25 mile loop. The numbers were 750/752 feet...so that's pretty close...that was after 1.25 hours and no weather fronts were moving in that I know of. :)

Isn't the calculation of cumulative elevation gain sort of arbitrary?

You go up a 10-ft rise - do you add that to total elevation gain?

What about a 1-ft rise in elevation?

One inch? A half an inch? If you sum up all the "gains" in elevation you make over a total ride, you'll get an unrealistically big figure. Somehow you have to filter out insignificant noise and only count real climbing.

The eTrex Vista HCx has "Total Ascent" and "Total Descent" registers...and after riding a loop...the two should be near equal; and I've found them to be very close. As someone else pointed out above...having the unit sitting here in front of me on the desk...it's tallied 4 feet up and 4 feet back down over 5 minutes. In the car, changing the speed of the heater/ac fan will also change the indicated elevation because of the change in ambient pressure (with windows up). I don't use these numbers for anything scientific, but for my 32 mile loops here in the mostly flats, I get something near 900 feet ascended/descended...and that seems reasonable.

The unit does not offer a continuous GPS elevation option, except for calibrating.

Isn't the calculation of cumulative elevation gain sort of arbitrary?

You go up a 10-ft rise - do you add that to total elevation gain?

What about a 1-ft rise in elevation?

One inch? A half an inch? If you sum up all the "gains" in elevation you make over a total ride, you'll get an unrealistically big figure. Somehow you have to filter out insignificant noise and only count real climbing.

that's exactly the problem, that's why 2 different programs sponsored by the same company (Garmin) can give wildly different answers, using the same (fairly accurate) elevation data. I live on flat island. I can do a 20 mile flat ride and have the programs record 800-1200 feet of climbing. Seems flat but there are actually lots of 1% ascents and descents.

Someone at BF mentioned another web-based service that offers smoothing alternatives, to control the "roughness" of the calculation say 1 ft or 10 ft or 50 ft. Can't remember what that site is though.

I've climbed and relied on altimeters for a long time. In my experience a barometric altimeter is much more accurate than GPS. GPS does very well with position, but in my experience it's altimeter function tends to fluctuates a lot. I've found this when using my SiRF III gps or using an old handheld unit. A barometric altimeter is much more useful, and you'll find that most gps high end receivers offer barometric altimeters in their higher end models because it's simply more accurate. It does need to be calibrated, but when it has been, it's more useful, imo.

For people who have the Garmin305 units, how long does it typically take you to lock onto a signal? For me sometimes it locks on right away and other times it takes a few minutes. In some cases, I've seen some inaccuracies in the GPS when it gets confused by tree cover. Is there a way to get it to lock onto more sateliites if you leave it out in the open longer before moving around with it?

Less than a minute, usually closer to thirty seconds.
Make sure it's stationary and that you have the most up-to-date firmware from Garmin. Always let it lock on before moving around.

The reason pilots update their altimeter data "frequently" is because air pressure is a local phenomenon and they need to sync to the area they're flying in...right?

Correct! :)