Temperature change causing elevation during activity

i am absolutely experiencing the same issue in all the sport activities (run, trail run, cycling etc).

it looks more relevant in the flat than in the mountains

when i start an activity i have a sudden spike then it drops below the sea level and after 30/40m i got the correct reading.

i tried everything (clean the sensor, change different settings) but still it's like that.

it started to happen 3-4 months back but i didn't figure out if it is due to the winter weather/temperature.

i was ready to ask for a replacement, but i see more and more people that is reporting this and i am afraid is more related to the FW. last year, same period i did not get this issue.

hi Simone,

Thanks for your reply. Could you include the elevation with temperature overlay plot? thanks!

sure, here you go

This looks very similar to what I observe (also since 3-4 months). 

I'm not 100% sure, but I think its been mentioned here before that altitude information logged during an activity is only based on the barometric altimeter, not GPS.  GPS altitude can be displayed during an activity, but the altitude logged for the activity is based on the baro altimeter.

The fact that two of your data displays are flat would seem to support that claim.

So, using barometer mode is not going to solve your issue.

Hi, sounds familiar!  Could you please include a screenshot of elevation with temperature overlay? 

I don't have anything as dramatic as those already shown, but I can contribute with this:

I started at 121 m and came back at the same spot at 128. In the middle, I stopped and sit down at a park: the autopause kicked in, and when I resumed walking the temperature had already risen from 16 °c to 28 °C, while the elevation dropped from 143 to 138 m. I DO NOT have autocalibration on, neither continuous nor at start, and this behaviour happens also when I never auto-pause.

Basically what I think is happening is that the altitude is set at the beginning of the activity, when the thermometer is still pretty much reading just the skin, and as the temperature drops, pressure drops as well and it is read by the watch as a rise in elevation. In fact, all my to-and-fro activities read more altitude gain than drop.

Having a diff of 7m altitude for a walk start and end isn't unusual.

How long was the walk.

Did the weather conditions change.

Any change in ambient air pressure will result in an altitude change of some sort.

The baro in the watch is accurate to approx +/- 10 feet.

The general lapse rate of air temperature is approx 1.98c for every 1000 feet of altitude under ISA conditions.

So if the watch was using the temp increase as a guide to altitude change then  (28 - 16) / 1.98 * 1000 = 6060 feet descent. (temp increasing = descent)

Which is clearly isn't doing.

So the watch isn't using temp as a guide to altitude changes and if you think about it, it cant. The watch temp reading will change dramatically depending upon what the person is wearing, T-shirt, base layer, fleece, Outer shell, gloves, that it could never reliably be used for guessing altitude changes.

Its more likely to be that the general air pressure over time is changing up or down, I have seen this a lot on my watch. I have every auto thing turned off  and so any pressure change the barometer measures is taken as a change in altitude. If I look at the baro graph over 12 hours and there is a trend up or down during that time I know the altitude is going to be wrong but in my use case I don't really care. To fix this, the trick is to recalibrate the baro every 30 mins or less if the have the information to hand. (its what pilots do all the time).

I know someone would post something like this. I said it was not impressive...

Anyway, it depends on your definition of unusual. If you have 1 activity with this behavior, it's unusual. If, like me, you have dozens and dozens activities with THE SAME behavior, that's not unusual, that's a pattern. All of them end at a higher elevation than the start. ALL.

That particular walk was 1 h, 5 km long, weather didn't change, but it's not that important.

I know perfectly the temperature lapse rate, and I never said the Garmin was bluntly implementing that. I just said that I think Garmin is using that information for some lower-level correction.

I know I am repeating myself a bit, but with my eTrex 30 I was performing 10-h-long tours, with thousands of m of elevation gain and loss, and when I came back to the start, it always gave me the same elevation (+- 3 m). And it didn't have a thermometer.

2572376 said:

Having a diff of 7m altitude for a walk start and end isn't unusual.

How long was the walk.

Did the weather conditions change.

Any change in ambient air pressure will result in an altitude change of some sort.

The baro in the watch is accurate to approx +/- 10 feet.

The general lapse rate of air temperature is approx 1.98c for every 1000 feet of altitude under ISA conditions.

So if the watch was using the temp increase as a guide to altitude change then  (28 - 16) / 1.98 * 1000 = 6060 feet descent. (temp increasing = descent)

Which is clearly isn't doing.

So the watch isn't using temp as a guide to altitude changes and if you think about it, it cant. The watch temp reading will change dramatically depending upon what the person is wearing, T-shirt, base layer, fleece, Outer shell, gloves, that it could never reliably be used for guessing altitude changes.

Its more likely to be that the general air pressure over time is changing up or down, I have seen this a lot on my watch. I have every auto thing turned off  and so any pressure change the barometer measures is taken as a change in altitude. If I look at the baro graph over 12 hours and there is a trend up or down during that time I know the altitude is going to be wrong but in my use case I don't really care. To fix this, the trick is to recalibrate the baro every 30 mins or less if the have the information to hand. (its what pilots do all the time).