Temperature

That has been there ever since got the 935. I am finding it about 4-5 deg C higher than the actual temperature so has some value if bear that in mind.

I have been sitting in my house all evening where the thermostat is set to 66 and my watch currently says 82.4 (current outdoor temp is 54). It can be a little closer on an outdoor run, but to me it's a fairly useless feature really. Body temp skews the reading too much. I've been thinking about getting a Tempe but I'm not sure I really need that either.

I have a Tempe but only use it if it's a really hot or really cold day and want to make sure that temp is recorded so that I have that info should I look at that run at some later date. Even then it can be really skewed by how much direct sun is hitting the device and where you're wearing/carrying it.

If you clip it on like a foot pod it's not affected too much by body temp, but when your direction of travel is such that the sun is beating on it, it will give artificially high readings.

It's been discussed in roughly a million posts already (give or take)...

The value is off because of your wrist. (I find mine to be off by about 10-20 degrees F, and that's indoors with no sun)
There is a Garmin support page that says as much.
If you really want an accurate temperature, you have to take it off your wrist or buy a tempe (which also has to be kept away from your body).

You might ask, "why bother including a temperature sensor in a WATCH if it doesn't work ON YOUR WRIST?" The answer to that is that the temperature is used in conjunction with the pressure to determine altitude. The question I have always had about that is "If the altitude is based in part on the temperature, and we know the temperature to be way off, what does that say about the altitude accuracy?" Maybe that's why we have to calibrate the altitude?

The question I have always had about that is "If the altitude is based in part on the temperature, and we know the temperature to be way off, what does that say about the altitude accuracy?" Maybe that's why we have to calibrate the altitude?

The way I understand it, is that the watch reads the pressure right where you are, be it at a 1 foot elevation, or at 1000' (given everything else is the same, it will be lower at 1000' as there's less air above you).

The temp is used then to adjust the observed pressure for sea level pressure. (weather reports etc are typically the sea level adjusted value, and cold air is more dense than warm). That's why calibrating the altimeter is important to start this whole process - use the calibrated elevation to adjust observed pressure with a known temp air for sea level, and use that known to look at observed pressure changes to determine the new elevation.

And the internal sensor is seeing the same temp as the air seen by the HW to do this adjustment, so it's OK that's the air is at a different temp inside the watch than outside, as long as it's known. It's a bit of a dance, but the actual temp of the measured air is involved in some way.

Moral of the story, if you want better temp data, a Tempe will do it! :)

Not sure with the 935 whether the wrist issue is the whole story.

My 935 consistently reads about 4-5 deg C too high. It continued to do so when I took it off my wrist when I rode my TT bike for an hour the other day. I did so as it improves the chances of getting a power meter reading from my "old" Rotor PM. Then again this was in the top tube bag so may have been "warmer" there.

Has anybody done a similar experiment? I did try to leave my 935 "outside" the other day before started a run.

What is interesting is that the temperature does typically drop for about 20 mins into an actual run before reaching a stable value. I can't see however that my wrist temperature would necessarily be following the exact same trajectory.

And yes AFAIK all Forerunners have always had "internal" temperature facilities. Very few of them have however exposed the value as a standard data field. I can only assume Garmin did so with the 935 as they think the values produced might mean something to us?

And the internal sensor is seeing the same temp as the air seen by the HW to do this adjustment, so it's OK that's the air is at a different temp inside the watch than outside, as long as it's known.

Ah yes of course. That makes sense.

And yes AFAIK all Forerunners have always had "internal" temperature facilities. Very few of them have however exposed the value as a standard data field. I can only assume Garmin did so with the 935 as they think the values produced might mean something to us?

Not that it really matters, but I think only Garmin tri/multisport/fitness watches with barometric altimeters include an internal temperature sensor, since it's only there to correct the barometer readings. Until the 935, the "runner-focused" watches did not have a barometer.

I would be surprised to hear that the 230 or 630 (or any Forerunner non-tri watch) have a temperature sensor.

If you find your way somehow to the diagnostics page on those ones then you might see a temperature...

Many folks use it as a bike computer (off wrist) and the temperature is accurate
Open water swimming is pretty accurate for water temperature.

For running, I use it as a reference to how my body felt, not as an ambient temp.