about disparities in elevation accuracy Garmin 500

[eddief]

as I wrote prior, my friend and I both have Garmin 500. Mine older, her's newer. We do the same rides and she almost always ends up with about 3-400 feet less of climbing over an average 35 mile ride.

I spoke to Garmin support yesterday. They said one way to test accuracy is compare your on the bike number with the "enabled" number on Garmin Connect. Bike = barometer, enabled = survey data. If there is a significant difference, he said it might be due to a malfunctioning altimeter. One way in which the altimeters malfunction is the unit gets dirty and the little holes on the back of unit clog up.

He suggested cleaning the back of unit / holes with a toothbrush and denatured alcohol. I did that and today my friend and I did 35 miles and 3500 feet and we differed by less than 75 feet...within a realm that does not cause us to wonder what the F is going on.

[kramnnim]

Interesting, I wonder what would happen if your friend cleaned her Garmin as well.

[gary_a_gooner]

My 500 tends to go off with regards to elevation and accuracy after a few months. When I notice it, I hard reset (erase everything), and that seems to set it straight. I've also started setting elevation points before rides.

[jimoots]

I never had issues with my 500 with respect to elevation accuracy, but my 510 does not like getting wet. As soon as water gets near the unit, elevation goes haywire

[DfCas]

I did a pan flat rail trail ride and the Garmin site, with elevation correction enabled, said I gain over 800 feet. It seems the rail trail is not in the database and it defaulted to the roads alongside the trail. The barometric altimeter said 26 feet of elevation gain, which I believe.

[jimoots]

I did a pan flat rail trail ride and the Garmin site, with elevation correction enabled, said I gain over 800 feet. It seems the rail trail is not in the database and it defaulted to the roads alongside the trail. The barometric altimeter said 26 feet of elevation gain, which I believe.

Elevation correct relies on the elevation maps, which are usually at a resolution around 100m increments. The maps then extrapolate to fill in between the known points.

As a result, the data ends up being quite poor. For all their faults, barometric altimeters should do a much better job 99% of the time.

[kramnnim]

Elevation correct relies on the elevation maps, which are usually at a resolution around 100m increments. The maps then extrapolate to fill in between the known points.

As a result, the data ends up being quite poor. For all their faults, barometric altimeters should do a much better job 99% of the time.

I think the majority of the populated areas of the US have been mapped to much greater detail, perhaps 1m.

[jimoots]

I think the majority of the populated areas of the US have been mapped to much greater detail, perhaps 1m.

Apologies for the misinformation. I found this @ https://strava.zendesk.com/entries/20965883-Elevation-for-Your-Activity

For example, the USGS National Elevation Dataset (NED), which we consult for activities located in the US, generally has a 10 meter resolution though some small areas have a 3 meter resolution. For activities located outside of the US, we consult the ASTER and SRTM databases which have a resolution of 30 meters and 90 meters respectively.

I can't comment on how accurate the data is at those resultions and what sort of extrapolation errors you guys would see. But in Australia, it's really bad data.

[bjf]

My rides are short and always start and stop at one place -- where I parked my car. No matter where I am, whether the route is flat or hilly, there's always an elevation discrepancy at the end of the ride on my Garmin 500. Something like total ascent 279 feet total descent 231 feet, for example. I would have thought that to end up at the exact same place I'd have to descend the same amount as I ascended. The rides are so short that barometric changes can't account for this. I just see it as inherent inaccuracy. Thus I'm not surprised to hear that there are significant discrepancies on longer, more serious rides.

[OtayBW]

I think the majority of the populated areas of the US have been mapped to much greater detail, perhaps 1m.
I think 1m is optimistic. In the east, contour intervals - at least on 7.5 minute quads - typically ranges from 10 ft to 50 ft. It's greater on larger-scale maps.

[kramnnim]

It's been a long day. I was thinking of a 1m difference in actual elevation, rather than 3, 10, or 100 meters of distance.

[downtube]

My wife and I have identical Garmin 500's and when ride identical routes there is always a discrepancy in elevation. Can be a 50' on a 1500' ride. I just figure it is an acceptable deviation. chuck