Did a metric century with my wife today. She has a Garmin 510 and I have a Lezyne Macro gps.

Our distance data was within a mile but she recorded 3400 feet of ascent to my 2600. That’s a huge difference. Stays the same sort of discrepancy when the rides are uploaded to Strava too so I assume it’s a recording error somewhere but not picked up by Strava.
Anyone got any knowledge regarding the accuracy of either device?

Strava doesn’t correct elevation data uploaded unless you tell it to.

Little blue link will appear next to the altitude bit when viewing ride.

GPS elevation can have the effect of smoothing out the elevation from a ride. It can also vastly increase your elevation if your GPS traces blips your off the side of a cliff and back up again. Barometric isn’t perfect either so it’s best not to worry too much.

GPS is generally less accurate than Barometric (assuming barometric has been calibrated and you don’t have huge changes in weather during the ride). The way GPS is designed (and the fact that the satellites are a long way up) means horizontal accuracy is much better than vertical.

I road 40 miles at the weekend in a loop but according to my Garmin managed to descend 100ft more than I’d climbed.Treat it the same as calories burnt, top speed, average speed etc and just go with the highest number 🙂

I’ve done a few alps road bike trips with mates, each of our GPS units gives a different altitude gains reading on every ride….

We always go with the most…..

I rode W2 with my mate on the weekend, both of us recorded it on Edge 1000s. My GF used an Edge 25. I recorded 950m of ascent, apparently the correct amount, Jo had the same. My mate had about 1300m.

His Garmin showed a max altitude of 600m, the actual top of the trail is just over 400m

what they said tried it once with two phones and two GPS devices

All of them gave different readings on every variable!!

I did this because my phone and gps did it so I was intrigued not because I am so sad I cared

Treat strava as a rough figure basically

your legs tell you if you are getting stronger and it never records the wind all my fastest commute times [ and slowest] are a result of the wind not anything I have done for example

Fair one. I’ve told the wife we’ll use her figure as it’s the greater of the two!
I think both use barometric sensors for altitude, I guess there is too much variance in conditions, sensors, software etc to expect too much accuracy. I was just a bit surprised at how much difference there was.

There’s a segment just down the road from me in Harefield, called “Somerset Ave Climb,” which has dodgy elevation data that makes it a cat4 climb.

It claims there is 343 feet of climbing, when there is ~50 over Minstead and to the top of Beauworth according to my Nexus’s GPS.

GPS works on triangulation from satellites. It doesn’t take weather conditions into account and so can be up to 5 or 6m out. Imagine a 6m diameter bubble. You’re somewhere in that with GPS. Sometimes at the top, sometimes the bottom.

This is why you get the inaccuracies,

As said, GPS elevation is not very accurate.

On top of that, if your device doesn’t have a barometer, Strava will favour topographical data anyway, which isn’t terribly accurate if the elevation goes up and down a lot in a small area.

And then you have barometers. Strava will favour barometer data, but that is dependent on weather and air pressure. Really you need to calibrate it. On a Garmin that is typically done by telling it the known altitude at particular locations. I don’t bother and find it can be a fair bit off at times, but also vary during the ride and on a loop the end point (same as the start) can be higher or lower than when I started, often by 100ft!

p.s. You can tell Strava to ignore barometer data, and it will use topographical instead.

Main thing for me is consistency. So long as it’s roughly clocking the same elevation gain for me with the same device and ride, then it’s something I can compare. Comparing to other people is just plain inaccurate.

Even distance varies with device, especially if someone has a speed sensor on their bike, as it clocks actual distance rolled, not GPS.

I use ridewithgps to both plot routes and record rides – has an option in a recorded ride to refetch map data to correct gps error but never seems to work – recent route difference between plotted beforehand and gps recorded was 400m and this isn’t unusual
Have the paid up version any ideas other than logging as an issue

I like the take the higher figure but I’d rather not kid myself so much that I blow out on some big climb

Also remember in valleys you don’t see as many satellites so your ground position can be less accurate (the older GPS’s exhibited a ultra fast zig zag from one side to the other) if there are steep sides then it can think you are 10-20m lower or higher than you are by ground position.
On the barometric that starts to fall down on the longer rides when weather changes more, it goes mad when a storm rolls in.

The devices might be using barometer or GPS elevation or an interpolation of topographic mapping using GPS horizontal data. Each has their own inaccuracies.

But also think about what you are trying to measure. Do you want to measure every lump and bump you go up and down? Or an average on the scale of metres; or hundreds of metres? The total elevation gain (and loss) is an integral or sum of the elevation data that is being stored, so even if your devices use the same approach, say an interpolation onto a topo map from your horizontal position, if one calculate the height change every 1 second and the other every 2 seconds, you are going to get different results. Moreover, errors sum up, so it may just be that one device is noisier (less precise) than the other. Google fractals – the length that you measure depends on the scale you are looking at (i.e. the sampling rate).

dcrainmaker has covered this also (as you’d expect from him).

https://www.dcrainmaker.com/2010/05/understanding-sport-device-gps.html

Got quite a lot of elevation gain whilst canoeing down the river wye