Garmin elevation discrepancy.

[Hugh Manatee]

GNSS is a good part of my job. I too have noticed differences in heights when using my 800 model. I know the height using the OSGB1936(02) coordinate system of a nail in the driveway. I measured it using an RTK GNSS system to an accuracy of around 10mm. Let's say the height came out at 115m. With my Garmin it rarely states a height within 2 or 3 metres of this.

From what I can work out, there are a couple of reasons for this. First is the way GNSS works. All those satellites, be they the American GPS or the Russian GLONASS system are at the same altitude. This gives you nice large triangles in the horizontal plane but very small ones in the vertical plane. This means as a rule of thumb that the vertical element of the accuracy is always one and a half times less accurate than the horizontal element of the position. With the GNSS I use, I typically get around 10mm horizontally and 15mm vertically.

My Garmin reckons it gets 15 feet position accuracy so I can't hope for much better than 25 feet or so. If one of you is plus this error and the other minus, that would go a fair way towards your discrepancy.

Another thing I have noticed is that the 800 will allow you to start recording a ride as soon as it gets a positional fix. Again, the vertical element takes longer to settle. As soon as you start your ride, the vertical must fix itself and not adjust itself further as it receives more data. If it did adjust, your height gained would go to pot. I have noticed huge differences in level depending if I start riding the moment I get a fix or if I leave the bike outside for a few minutes whilst I get my shoes on or wait for the warm up balm to get hot!

If you were to overlay a series of vertical profiles of the same ride, I reckon the profiles would look the same but would by no means would they sit on top of each other. Ooh, now there's an interesting project.

 

[BrynCP]

All this talk of GPS, but the Edge 500 uses a barometric altimeter? This is why Garmin Connect doesn't make adjustments as it's supposed to be more trusted than the GPS readings for altitude.

Saying that, typically my house is 3 metres lower than when I started off! And sometimes my house is at sea level, sometimes I am at a couple of hundred feet. I know you can set the start altitude, I just haven't got around to it on this unit!

 

[bpsmith]

Went through this with my mates 510 always giving 300ft to 400ft extra over mine on a long ride with 2500ft of climbing. Mine even showed me underwater for many a ride, so got it replaced. We got our 510's from the same place in the same day and have both got all updates installed.

My new unit no longer has me underwater. My mates unit...still showing more climbing than me on the same routes. Academic really, as he does a whole lot more climbing than me, but strangely he is not bothered much like you @huwsparky. If it was lower than mine then maybe he would be more interested?

 

[huwsparky]

Maybe that's just how it is! I managed to find Strava's correction tool and it amended my ride from 1650 odd ft to 1350 odd ft.

Now, as I understand from the little research I have done the Garmin should be more accurate due to the fact it has a barometric sensor.

Either way, don't think I'll be amending any more of my rides!!!

 

[BrynCP]

Something I have read about before is people covering the underside of the Edge, such as with a silicone cover or a different mount; this can affect the barometric readings.

 

[bpsmith]

Maybe that's just how it is! I managed to find Strava's correction tool and it amended my ride from 1650 odd ft to 1350 odd ft.

Now, as I understand from the little research I have done the Garmin should be more accurate due to the fact it has a barometric sensor.

Either way, don't think I'll be amending any more of my rides!!!

Better to have honest rides though? In saying that, just stick to one way or the other in order to compare to yourself over time.

 

[bpsmith]

Something I have read about before is people covering the underside of the Edge, such as with a silicone cover or a different mount; this can affect the barometric readings.

Mine has no cover and is mounted using the standard mount and rubber bands on my stem. Exactly as manufacturer prescribes.

 

[huwsparky]

Mine has no cover and is mounted using the standard mount and rubber bands on my stem. Exactly as manufacturer prescribes.

Same as mine.

Better to have honest rides though? In saying that, just stick to one way or the other in order to compare to yourself over time.

Yep, I want my data as accurate as possible really but saying that, I have no definitive proof of which is most accurate so I may as well leave them all as they are.

I'm not really at all bothered. I was just wondering if maybe my Garmin was not setup properly. I have tried altering some settings so I'll just compare the same routes as and when I get round to doing them.

 

[outlash]

Who said Strava is particularly accurate anyway? Try UL'ing the same file to Strava and Ride with GPS, I constantly get a large difference in height, normally more with RWGPS.


Tony.

 

[BSRU]

I had some problems with different elevation gains on the same route caused by strong winds, turns out the silicon case was the fix for it.

 

[huwsparky]

I had some problems with different elevation gains on the same route caused by strong winds, turns out the silicon case was the fix for it.

Interesting. Was windy yesterday but the same for the both of us.

Our corrected Strava stats were about 20ft different from one another.

 

[T.M.H.N.E.T]

My house is more than 20ft high.

What is the point? OCD?

 

[BSRU]

Interesting. Was windy yesterday but the same for the both of us.

Our corrected Strava stats were about 20ft different from one another.

I assume the Strava correction just uses the GPS data to fetch altitude information and ignores the altimeter info on the Garmin.

The bad altimeter ride had 150 metres less elevation gain, over 15km, than on all the other good rides. The altitude graph made it look like a flat ride instead of hilly. Since getting the silicone cover the altitude values are very consistent on all rides.

 

[huwsparky]

My house is more than 20ft high.

What is the point? OCD?

Maybe,

I'm glad to hear the different theories as to why this seems to be happening. Not because it really matters, as it doesn't, but nice to know none the less. If for example my £130 bike computer was broken I'd rather know about it than not.