Elevation gain in courses regularly overestimated by 10-20%

Both are incorrect 

Connect estimates altitude using a digital elevation model, your watch uses barometric altitude with DEM and GPS calibration. 

DEM

Supose that you are riding in an almost flat road  in a mountain side, the DEM samples are in the side of the mountain, if you are lucky with 10m horizontal distance but usually with 30m resolution. 

Unless you are very lucky the samples will not be in the road and DEM you will give you different altitudes for the flat road. 

Fenix

You are in the same road, however during the ride the barometric value changes due to weather changes or simply because pressure is not the same everywhere. Fenix will detect those changes  as altitude changes. Fenix will try to compensate this with calibration using DEM and GPS. 

Also temperature influences barometric readings and that's why fenix as a internal thermometer but of course is another error source. 

My experience tells me that Garmin connect overestimates in my country  and I use the fenix values.

So how do you do it? Do you have instructions? Or just when jogging in the barometer itself, set the altimeter value and then the altimeter will no longer rely on the weather? Explain

Anton Pochtoviy

hi,

because you asked so nicely, read here ... fenix-6s-altitude-becomes-incorrect-when-i-start-an-activity

acousticbiker

hi,

elevation during biking with any Fenix 6 has better result wearing the watch on the right wrist

or covering the baramoter holes

• some recent posts about the topic
• hohenmesser
• want-have-945-question-ist-there-really-a-baro-design-flaw-because-on-back
• fenix-6-pro-elevation-problems-when-running-and-cycling-since-10-10

happy & safe sporting

jose.cboliveira, does GPS elevation also come into play at all?

Former Member, is it fair to conclude that you are saying the GCM course data are more accurate than the activity data from the watch (when the GCM course says the elevation gain is 4000 ft but after the activity, it is recorded by the watch as having been 3000 ft)?

acousticbiker

out of context...

acousticbiker said:

GCM course data are more accurate than the activity data from the watch

here is another one to read ... edge - altitude-all-wrong

I personally would not trust GCM in a hundred years...

check this ...

make a note of elevation loss & gain from your activity

export your activity as GPX from GCM and import that GPX back into GCM

• it could say the activity already exists
  • just create a new account or wipe the original data from GCM (first export the original from GCM)

what is the elevation loss and gain from the imported GPX data ?

• when the same, I would trust the data from GCM
• when not ...

happy & safe sporting

acousticbiker

or does this procedure compare with your ride setup ? strava---gc-wrong-elevation-data-after-sync

the course you create from GCM could be wrong to begin with

update

• another one ... elevation-issue-for-routes-imported-from-strava

happy & safe sporting

I'm aware of the elevation issues with Strava course import but my issue is with courses created in GCM. I have an Edge 130+ and notice that the elevation at the end of a ride is different than that recorded by my F6. At the end of the day, what I'm wondering is whether there's a source of truth on the elevation I gained during a ride using a course?

ok

so wear the watch on the right side next time

happy & safe sporting

I think the underlying issue is the granularity of the DEM model the course generator gives, and the inbuilt accuracy of the GPS. Personally, I'll take the GPS over the DEM simply because I think the DEM's granularity is something like 5M - though I'll be happy to be proved wrong.

From the wiki article

"A much higher quality DEM from the Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) instrument of the Terra satellite is also freely available for 99% of the globe, and represents elevation at 30 meter resolution. A similarly high resolution was previously only available for the United States territory under the Shuttle Radar Topography Mission (SRTM) data, while most of the rest of the planet was only covered in a 3 arc-second resolution (around 90 meters along the equator). SRTM does not cover the polar regions and has mountain and desert no data (void) areas. SRTM data, being derived from radar, represents the elevation of the first-reflected surface—quite often tree tops. So, the data are not necessarily representative of the ground surface, but the top of whatever is first encountered by the radar."

So it's issue is not only the accuracy of the height (z axis),but how much it's sampled over x and y - i.e, how well does it map all the little up's and downs in a square related to it's x,y resolution.

DEM elevation will therefore suffer in more spiky terrain, and so on flat regions the DEM and real gain will agree well, but in mountainous territory or even just hilly may not.