GPS altimeters

Some websites like Strava/RwGPS etc. will let you use “corrected” elevation data, but yeah it’s silly when altimeter based readings are supposed to be more accurate in the first place.

From
http://support.garmin.com/support/searchSupport/case.faces?caseId={53d6ba60-e4d2-11de-d785-000000000000}

“The barometric altimeter is not temperature compensated. Temperature changes in the device will affect the barometric pressure sensor and altimeter readings.”

It seems temperature (and wind chill I guess) changes will have pretty bad effects on elevation data accuracy. My Edge doesn’t seem to read temperature that well in the first place, so I’d guess my barometer’s accuracy isn’t great either.

When I’m riding on Skyline Drive, or the BRP, the altimeter shown on my garmin is usually within 25-50 feet of the road-side markings, which I consider pretty good. Weirdly when I get down to the Potomac, say on Hains Point, it often shows at -150 feet or so.

@JimF22003 44814 wrote:

When I’m riding on Skyline Drive, or the BRP, the altimeter shown on my garmin is usually within 25-50 feet of the road-side markings, which I consider pretty good. Weirdly when I get down to the Potomac, say on Hains Point, it often shows at -150 feet or so.

I’ve had the same experience–not sure if it has to do with Skyline having a better lock when I start, or if starting amongst the tall buildings of Ballston screws things up (taking the elevator downstairs drops me 200 feet and sometimes the Garmin doesn’t register that right away).

Consumer GPS units are really geared much more towards accurate 2d positioning. My understanding is that getting elevations from GPS satellites is much more difficult than getting 2d positions, and accuracy largely depends on which satellites are locked onto. The barometric altimeters increase accuracy, but aren’t foolproof (since they measure off air pressure, which varies depending on weather). So, yeah, its a baseline number, but likely won’t be perfectly accurate. My commute tends to vary 10%-15% in elevation numbers.

As an aside, all this is the reason that a lot of back country trails (like the stuff in the GW forest) have elevation at some trail markers. Its to give a point of reference and to use as a point of calibration for the old school barometric altimeters some hikers use.

Altitude requires more satellites for accuracy, and consumer gps just doesn’t bother worrying about it much.

@DaveK 44820 wrote:

Looking at everyone’s tracks of my weekend ride, our various Garmins show up to a 2,000 ft difference in total climbing. Of course I got the short end of the stick…

I figured that Tim was just catching air shooting up and over the tops of all the hills.

I often go up or down 10 to 20 feet while riding the trainer. I checked this against weather reports and noted the rise and fall went along with the pressure readings. IIRC you can change the settings in the Garmin to ignore the baro. Overall I have found that I tend to end up about where I started so to speak and that the same ride such as the Arl loop produces about the same elevation.