You need to caveat this. They are only “crap” when the barometric altimeter is exposed to rapid changes in ambient air pressure. When the atmosphere is stable a properly calibrated barometric altimeter is far superior to GPS derived map datums, which are most accurate when the elevation changes are minor (useless in mountains). The Strava (or any app’s) evaluation data is a poor approximation of actual elevation changes, it is only superior to barometric data when the atmosphere is unstable or there is no barometric data.

Without a barometer, how would once record elevation data? Maybe I misunderstand what you’re saying.

An estimate may or may not be accurate, but it will definitely give you an approximation. In my experience Strava elevation data, when I have compared it to actually recorded elevation data, suffers from some accuracy problems depending on where you are in the world, and which map datum they are drawing from in that location. For the most part, the Strava elevation data is decent in most of the US and Western Europe.

2 measures = 2 results. I am guessing one of them is smarter than the other.

I always use Strava's, because numbers are higher (of course!).

​​​​​​From GPS. The file you record just has a field for elevation, it doesn't care how accurate it is. The file header identifies the device that recorded it, Strava keeps your elevation numbers if it thinks you had a barometer and it ignores them and uses its DEM instead of it thinks you didn't, because GPS elevation data is universally considered bad.

That’s what I thought you meant. Once point of clarification, without a barometric altimeter, there is no elevation data in the elevation field, and Strava fills it in from a pre-determined DEM, as you’ve pointed out, in the .fit file they generate when your activity is uploaded.

It's probably a distinction without difference, but the elevation is not being recorded or determined directly from the GPS. It is instead being derived by using your 2-dimensional (N and E) GPS positional data to derive your elevation by joining it to the DEM.

I thought GPS can be used to determine elevation reasonably well if the GPS is locked onto four or more GPS satellites?

I do not have a barometer or bike computer, but would the barometer reading remain accurate if there are localized temperature changes? For example, when riding from the bike lane in the sun onto a bike trail shaded by trees?

Reasonably well yes. But to your expectations, probably not.

In the .fit files that i save from my own application i store elevation data. Also the .fit files i extract from my phone using CycleMeter also contain elevation data. what makes you think elevation is not stored in a .fit file? i see elevation data in .gpx as well.

i should be clear that my phone does not have a barometric altimeter.

GPS is pretty good for lat/long but less accurate for elevation. But GPS can be used for elevation. I've been using a Garmin watch for years, I have a screen set up (for hiking) showing me the GPS altitude and the barometric altitude, along with some other things. You can record a gpx file on a device without a barometer or DEM and still get elevation data, it's just noisy.

​​​​​​The formula for turning air pressure into elevation takes temperature as an input. That's why the watches have a thermometer, it's useless otherwise because your body temp throws it off but it measures temp at the barometer. Got to be why most of the Edge computers have thermometers too.

To your question the temperature like in the sun or shade shouldn't have much effect. Differences in barometric pressure can be from other things than elevation though. Weather especially storms, fronts, riding from one valley over a mountain pass into another valley sometimes, things like that.

To this pspelger , correct…I went back and looked at the .fit file uploaded to Garmin from a few runs that were recorded on a device without a barometric altimeter, and there were still elevation data contained in it, but it was “noisy.” It was showing elevation changes where there was none. Bottomline, if using GPS to derive elevation, it will be subject to the same interference as the lat/long is subject to, which means, trees, terrain, buildings, your body, can all cause it to read inaccurately. This is probably why Strava, for example, automatically defaults to the DEM dat

Reasonably well, but not reliably, as the GPS signals are subject to a lot of interference at ground level. It’s a great technology, but it has limitations. Barometers will generally be much more accurate and reliable in determining elevation, with the ideal use case being when the atmospheric pressures are stable. Temperature is just one component of air pressure, for sure, in fact as I think about it, very watch or computer I’ve had in the last 10 years that had a barometer also had a thermometer.

Sooooooo, with all of this information, Strava is accuratER than Wahoo, right? :innocent:

Going across a bridge back and forth is a sure way of Everesting without all the climbing on Strava, ;-)

If the Wahoo doesn't have a barometer then yes, unless no. "It depends" is the heat useful but most truthful answer, like usual in this place.

Does the elevation graphic show an anomaly somewhere along your route? I encounter them every once in a while on my Strava workouts. For example, in the below screenshot of a Strava walk/jog I did along the beach at Punta Cana, DR in 2018 you’ll notice at ¾ mile and 3¼ miles on the elevation graphic (below the map) my elevation plunges from sea level to 1000 feet below sea level in a really short distance. When I finished that activity and saw that it recorded that much elevation change, I thought WTF? I didn’t leave the beach. I can assure you that I didn’t fall into an abyss. We vacationed there in 2019 also and it did the same thing in the same spots. (I should have checked it with another app, like MMR, but I didn’t.) I can’t explain it, and I’ve seen weird things like this in other places too. I don’t know if it’s a flaw in the Strava elevation function, the GPS system, or some sort of natural phenomenon. — Dan


https://cimg5.ibsrv.net/gimg/bikefor...adba1b99dd.png

Around Seattle, the Wahoo app on my smartypants barometer-enhanced iPhony never fails to yield much higher elevation data. 10k climbing w/ Wahoo app => Strava => becomes 5-6k. Others have suggested an actual Wahoo device would be more accurate than the app, but it doesn't matter to me either way so I'm cool with Strava's conservative (?) estimations.

This is exactly what I am experiencing. I did not know the iPhone has a barometer until you pointed it out.

Some do, some don't... as I understand

Supposedly iPhone 6 onwards have a barometer, but evidently Strava thinks the instrument is not good enough.

They actually put them in on the theory that having a barometer will get you a GPS lock faster. Knowing your approximate elevation means being able to throw out a lot of potential but wrong solutions while you're acquiring satellites.

What does the device name say in the files it records?

if I recall you can fi d the device name near the top of a .gpx file. It'll be in a .fit file as well but you will need a decoder. .gpx is just text.

I don't know how to get a .gpx file. The Wahoo iOS app can export a .fit file but I am still looking for a viewer or preferably a converter to Excel; can anyone please recommend such an app (preferably PC)?

I used both Strava and Wahoo apps on the same iPhone X while hiking. Wahoo gave unrealistically high numbers for climbing compared to Strava. These days I use Garmin Forerunner 935 watch - it is quite close to Strava numbers. So my take is that Wahoo lies regarding climbing numbers, I do not use it anymore.

If you want to see what a "huge" discrepancy looks like, this is the elevation profile recorded by my Garmin Edge 1030 - which has a barometric altimeter - during a heavy monsoon rain at the weekend:
https://cimg9.ibsrv.net/gimg/bikefor...08c96b2728.jpg

compared to the Strava-corrected elevation profile, which is a whole lot more representative of what the ride actually looked like:
https://cimg0.ibsrv.net/gimg/bikefor...7f6881aa8e.jpg

Golden Cheetah will show you the raw data in a grid and let you copy it in a way that Excel can understand. Might have an export, I've never tried.