I just put Strava on my iPhone. This weekend I went riding in the Arizona foothills and Strava said I did a 320 meter elevation gain. I find that a little hard to believe. Not impossible mind you, but hard to believe. If that's true then I'm a better cyclist than I thought.

I think the horizontal is accurate but I'm wondering if anyone else has experience with the vertical numbers.

Thanks.

The GPS app (cyclemeter on iphone) and the Garmin GPS (Edge 200) I have used do make errors. I can't say the extent of the errors. You could probably use a web page and paint your route along the road and see what kind of estimate you get.

I do not trust the elevation gain numbers from the cycling apps that I have tried. Even riding the exact same routes, I get widely different results in elevation gain from day to day (or even between morning and afternoon). I have mainly used the Cyclemeter app, but also have tried Strava and got similar results.

I rode the same local loop and turned on my iphone 6 strava app and at the same time I used my garmin 705. For a 2500' climb in the hills the app was off by more than 800' and the garmin was off by 200'. I was able to adjust the garmin by setting it to update more often and setting up a location. The next ride it was only off by 30-50' or so. However the app remained to report between 500-800' off each and every ride. I now ride with my garmin only and leave the phone in the seat bag in case I need to make a phone call.

What do you trust?

As the OP said - the horizontal numbers. The GPS location and timing and speed is very good. The VAM is good. Still, when you find a route that is done hundred-thousands of time - you get a very good idea of what the truth is.

GPS is pretty good at positional accuracy but largely sucks at altitude accuracy. Just the nature of the technology in terms of how the satellites are arrayed relative to your location.


Some web services such as Garmin Connect will correct elevations to topographic data for the locations. I don't know if Strava has that turned on by default or not.


Also have a look at Understanding Sport Device GPS Elevation Issues | DC Rainmaker

Are you folks saying these gadgets are trying to measure altitude? I would think that's a misguided approach. I assumed that elevation points are known for most points on roads and altitude is calculated from that collected data. But it's a big assumption of mine, and it could be totally wrong.

You would think they use collected elevation data, but my results on Strava indicate otherwise. One of the routes I cycle weekly is a mountain MUP and I get wildly different altitude results from week to week...2,200 feet to 1,500 feet. Distance is always consistent week after week and also agrees with that recorded on my basic Cateye non-GPS bike computer. My cycling buddy will ride the same route with me at the same time yet our recorded climbing results could differ 500' or more.

Keith

I love my iPhone and many of the apps.... but phone calls are what the phones were made for... and what they do best. I use the less popular (and far cheaper) elite cycling app to track my cycling. It's not perfect! If feet/meters climbed, or mile/kilometers cycled was REALLY important... I'd use something other than a phone app to track my data.

But as a non-competitive... unracer... the cycling app gives me a wealth of data for a handful of dollars a year. And I no longer need to transcribe to a clipboard and later type/enter it into a document. That hassle is gone! And everything is automatically backed too.

P.S. I still run bicycle computers on a couple of my bikes. I like being able to see the speed, time, and such when riding. And I keep my phone in my jersey pocket. AND... it's nice to NOT have a modern computer mounted on a vintage bike... yet still collect the (albeit less than perfect) data via the phone app.

I've been poking around the Internet with the keywords "google map elevation". I've been to some sights and I get quite a range of elevations. But the upshot is that Strava has almost doubled my elevation. The best I can determine is that I gained about 140-160 meters. That sounds more like it. Although I prefer what Strava said since it makes me more badass

I dunno. I'll do more research. In the meantime try this website and tell me what you think.

veloroutes.org - create a bike route map with elevation profile

I think you have to consider the altimeter on the phone. I have a Garmin Edge 500 on the bike, but use Strava on the phone for hiking. My friend uses two other apps to do the same thing and he has a different phone. Surprisingly, the distances often differ by 2-3% but the elevation is fairly close. I've never looked at the raw data to see if this is missing points that Strava fails to fill in or what. Stopping and starting and other motions can affect the data stream. You can probably find some forums with hard core GPS folks who can provide more details and assessment of various devices and software.

Overall, I'm happy with Strava. It's just recreation, not rocket science for me and I am happy with the data I get.

Agreed. But I've been taking a lot of multi-kilometer hills and it would also be nice to know my elevation gain. It's still a good program and I recommend it.

Just a point of clarification -- Strava or any other app is merely getting the info from the GPS chip built into the phone. While some apps/sites may apply different sampling/smoothing/correction algorithms to the data, ultimately the data is coming from the GPS chip not from the app. The exception is if the website then later applies elevation correction to the route using other data sources.

Ah! So it's the phone not the software. Interesting. Thanks for the clarification.

Technically it's a combination of both as different software can treat the data it gets from the GPS chip differently.

An analogy would be racers and racing bikes; the same bike will likely perform differently if ridden by different racers, and a particular racer may also perform slightly differently on different bikes.

In the end though, it would be difficult for software (Strava) to achieve greater accuracy than the underlying hardware can provide.

What's really interesting to me is that if I record a ride with cyclometer and then load it to strava, the elevation is typically very different. Presumably these are in fact using the single GPS dataset recorded during the ride.... It's been my impression that one or the other is reconciling the track with mapped elevTion data, and I've assumed that it's strava since cyclemeter elevations are much less consistent over repeated routes. Additionally, when I asked cyclemeter tech support they said elevation accuracy is limited to GPS data from the phone.

Anyway, that's been my experience.

SA

It's conceivable that all of the users are providing data to the apps which will, in turn, make elevation data more accurate, through aggregation. I don't know if this is happening, though.

Some of these gadgets have maps with a digital elevation model, so they can look up their elevation from their position. But a lot of these gadgets do not have maps. Some of them have barometric altimeters (which are more accurate than GPS by an order of magnitude) but not maps. I ride with an Edge bike computer and an ABC watch (altimeter, barometer, compass).

A lot of web sites that you upload your data to will do altitude corrections from the map. Garmin Connect automatically corrects elevation data from any device that doesn't have a barometric altimeter, lots of other sites do the same. They don't always treat things like bridges consistently; is it the elevation of the roadway over the bridge, or of the canyon below? Also if you're riding along the side of a mountain or valley or canyon, when the GPS gets your position wrong it will put you up or down from the road. There's a lot of side-to-side noise in GPS so that situation will exaggerate your elevation gain for the ride, and a barometer would be more accurate.

2% difference is acceptable. Two GPS devices will pretty much always disagree a little about how far they've traveled.

Sunday I did a ride around Camano Island with a Garmin Edge 800 and a Garmin Fenix 3, you'd think they would agree perfectly but one said 30.06 and the other said 30.15 miles. That's pretty typical. When you think about all the tree cover that the GPS signal traveled through, it's kind of amazing they agreed so closely.

Regarding Garmin cycling computers, I see a 11-12% difference in elevation tracking between the Garmin Edge 5xx/8xx series and the Edge 1000.

For whatever reason, the Edge 1000 shows less climb.

That's 0.2% difference which seems very good to me!

For gps altitude to be accurate it takes two receivers with very precise antenna distance measurement for them and some real hefty software. btdt
Cheers,
David in Alaska