Calibrate local barometer

5693709 said:

I remember having this discussion with you a year ago or so, you're still not understanding how barometric altimeters are designed to work.  The fact that your eTrex was spot on after an 800m climb this time of year means that it was probably using GPS for the elevation information, and it happens to be very good at determining altitude with GPS.  If it wasn't using GPS, then its barometric altimeter is not working as designed, and just happened to be correct by pure luck.

The standard atmosphere used to calibrate all baro altimeters would expect the air temperature at 1900m to be 2.65°C.  That's based on a sea level temp of 15°C and a temperature decrease of 6.5°C per 1000m of elevation gain.  At 2700m the expected temp would be -2.55°C.  I'm guessing when you were hiking a few days ago the temperature was actually higher than that, correct?

If the air temp is higher than the standard atmosphere model expects, then the air is less dense than it expects.  In that case, pressure changes less as you change elevation than it would if the conditions matched the standard atmosphere model.  Let's say the actual temp when you were hiking was 15°C, and the average temp for your altitude range based on the standard atmosphere is 0°C.  A rough calculation in degrees Kelvin will tell us that the density of the air was 273/288 of what the standard atmosphere expected because air density is proportional to absolute temperature (ignoring minor influences like humidity).  Absolute temperature in °K is °C + 273.15.

Now, multiplying the fraction 273/288 by the 800m you climbed, says the altimeter should have interpreted that pressure change to be 758m rather than 800m because of the reduced air density.  That's a difference of 42m, which is very close to your observed 37m difference.  If the air temp were 14°C the difference would be 39m and at 13°C would be 36m.  Your altimeter is working perfectly, it was designed to work under standard atmosphere conditions, but when you go hiking those conditions don't actually exist and you have to know how to compensate for the differences.

Yes I remember the discussion: I didn't immediately realize it was you because your nickname is not explicative 

I understand the physics, but I guess it's you that don't understand how the eTrex works. It's not just this time that the eTrex was spot on: I bought the device in 2011 and used it in more that 200 activities, summer, winter, rain, hail and snow, and it was spot on every single time. (The only condition where the eTrex suffers is wind: I hiked to one summit in the Rockies where there was a 50+ km/h wind and there are lots of spikes in altimetry, I think due to sudden changes in pressure caused by the wind.)

By spot on I mean less than 5 m difference with respect to the real altitude, determined by geodesic points and other points for which I have the elevation value. I have tons of tracks to prove this.

It can't be pure lack, can it?

The only thing I know is that eTrex supposedly works similar to Suunto's FusedAlti: it combines GPS and barometric values with a proprietary algorithm, maybe with successive approximations, until they converge to a value.

So, it's stipulated that eTrex is always spot on: the issue is "why Fenix isn't?". Either the barometer is worse, its GPS chip is worse, or the algorithm is worse. I don't actually think it's the latter, it's just a piece of software that works well since more than 10 years, so why change it? So it's either the barometer or the GPS. This post was trying to inquire about the barometer...

As for the app you wrote, I remember reading it in another post few months back and yes, I'm intrigued. If you're still willing to share it, I would like very much to try it. I will PM you for details, thank you very much.

PS: For those who would like to ask me "why don't you stick with eTrex if it's so perfect?", well there is more than one answer. First, it has more than 10 years: I brought it to the Rockies, the Alps, to the Arctic, it took more than some beatings. I already changed its screen once and it would be time to replace it again. Sometimes it reboots itself for no reasons. Last, I took up running and it's not very easy to strap it to my wrist 

gspinoza said:

I understand the physics, but I guess it's you that don't understand how the eTrex works.

All barometric altimeters are calibrated to the standard atmosphere, and all will read low by the same amount that you saw in your hike.  You keep suggesting its a flaw in the Fenix, but that's not true.  If the e-trex is using only its barometric sensor, and shows exactly 800m of gain under the conditions you hiked, then that doesn't match with what every other altimeter in the world would report. 

So, the most likely explanation for your etrex is that its really using just GPS or a combination of GPS and barometer, and that combination always works well under any atmospheric conditions because GPS is obviously not tied to the standard atmosphere definition.  Also, because we know your hikes aren't going to be in conditions similar to what the standard atmosphere expects, the barometric altimeter data from the etrex should be off by the same amount as the Fenix, and when combined with the GPS data would likely cause more error than just using the GPS altitude data alone.  It may be that the e-trex is just going off of GPS data alone, but I really don't know, neither of us do.

The question I think you're asking is why can't the Fenix use its GPS/baro capability to produce similar results, that's a valid question.  I never use GPS with the altimeter on my Fenix, so I can't comment on how well the GPS/barometer combo works, and to be honest I don't even know if it can use both in combination or if you have to choose only one or the other.  That would be a question for Garmin.  I only use the barometric sensor for altitude during hikes, and have an app I've written to compensate for the non-standard atmosphere, so I get very accurate results using that.

The few times I've gone into an activity with GPS enabled and looked at the GPS reported altitude, it was very accurate, but I didn't have any trees, buildings, or other obstructions so the reception was perfect.  So, one function at a time, I'm convinced that the GPS altitude values are accurate, and the barometric altimeter values are accurate given what I know about the standard atmosphere.  Why GPS and baro can't work better together to provide performance like you see in your e-trex, I don't know, but its not because either of the subsystems in the Fenix have poor accuracy.

Have you ever tried using just the GPS reported altitude during a hike?  Given your typical hiking environment that would probably work a lot better unless you're in a forest of trees that block the reception.

I’ve learned a lot from reading this thread, even as a private pilot in the US that understands standard pressure, standard day temperature, etc. I think some of the confusion is the nomenclature the watch uses to describe the barometric pressure, and the ability to set both altitude and pressure in barometer calibration. Here’s what I’ve recently done that has worked for me.

The other day I was at my local airport, which reports local barometric pressure as part of its automated weather broadcast. Pilots pick this up via radio, but you can also typically call a phone number to hear the live broadcast. You can find those phone numbers for US airports (and the one for an airport hopefully near you) here:

https://www.faa.gov/air_traffic/weather/asos/?state=FL

Call one of those numbers and you’ll hear the automated transmission report a figure for “altimeter,” which is the current barometric pressure at that airport (in the US in Hg).

You can also find the altitude for that airport via SkyVector here (by clicking on the URL for the airport on the map): https://skyvector.com/

If there’s no airport nearby, I would assume you could use the pressure being reported by a nearby weather station ... but you’d have to also know the altitude of that station (at the least, it should not be significantly different from your own altitude).

On my watch I went to the barometer widget, clicked calibrate, and entered BOTH the local altitude for the airport AND the reported “sea level pressure,” using the pressure from the airport. This gave me correct pressure and altitude at that spot, and as I moved around after it stayed very accurate, within 10-15 feet altitude. When I noticed that the altimeter had wandered a bit, I entered the altitude at my location (my home is at 14 feet MSL, for example) and the barometer immediately reflected the correct pressure for my area. If you were not near your airport, I am presuming that you could enter the altitude and pressure for the airport, and then calibrate just the altitude to your known altitude, and it will adjust the pressure accordingly. 

I won’t do this all the time, but believe that calibrating both the known altitude and current pressure Hg from a nearby reporting station gets everything set straight. Once done calibrating altitude only seems to work, although after a flight in a pressured airline I would again re-calibrate both altitude and pressure.

Just my experience, but it has seemed to work.

5693709 said:

So, the most likely explanation for your etrex is that its really using just GPS or a combination of GPS and barometer

Well, that's what I've been saying all along. eTrex does use a combination of GPS and barometer, with proprietary algorithm.

5693709 said:

The question I think you're asking is why can't the Fenix use its GPS/baro capability to produce similar results, that's a valid question.

Yes, that's exactly the point. As I said, I don't think this is due to the software, so it must either be the GPS or the barometer.

5693709 said:

Have you ever tried using just the GPS reported altitude during a hike?  Given your typical hiking environment that would probably work a lot better unless you're in a forest of trees that block the reception.

I never tried that, but it would make no difference: it will be always the calculated GPS that goes ultimately in the .FIT file, not the GPS one. It may be beneficial to look at it while hiking, but since it will not be written in the file it won't be that useful.

Thank you, I'm already doing all that, even more. I set up a professional meteo station at work, calibrated by myself, and calculated sea pressure.

http://rime.inrim.it/luc/meteo/index.php

I do all the calibration in the watch as you said, sometimes even more than once per day, but after a while all the problems arise again.

Back to your original question on this thread, I think these baro sensors are typically spec'd to about 1.0mb accuracy, and since yours is off by about 0.7mb, that's within spec.  When I've compared my several sensors that read to 0.1mb over various altitudes, any discrepancies between them is relatively constant, so I'd suspect yours is going to be off by 0.7mb at just about any elevation, and would not have any meaningful effect on the displayed altitude.

I have a couple other devices that do allow the user to manually calibrate the raw baro reading from the device, but the Fenix doesn't seem to offer that option that I'm aware of.  Even if it did have that capability, that won't fix your most basic problem of it reporting elevation gain incorrectly during your hike because of standard atmosphere issues.  FWIW, the app I wrote to compensate for standard atmosphere error wouldn't change the values written to the .fit file, so that wouldn't help you either, same as using GPS for altitude information during the hike.

For what you want it to do the Fenix isn't going to match the performance you're seeing with your e-trex because it can't store the GPS altitude into the .fit file.  I don't think there's anything inaccurate in either the GPS or the baro altimeter features of the Fenix, I think if you tried a hike with GPS enabled, you'd probably get performance very close to the e-trex.  The e-trex likely has a larger antenna inside its larger form factor, so it might work a bit better in marginal reception areas, but in wide open reception areas they'd probably both display the same altitude.  Seems like that would be a good experiment to conduct just for another data point in your analysis.

You're just missing a software feature in the Fenix that you'd really like to be there, the ability to use the GPS derived altitude data and store it into the .fit file instead of the baro altitude data.  As has been mentioned over and over, what you're seeing with the baro altimeter readings is exactly what would be expected, so there's nothing wrong with it, and nothing that can be re-calibrated to make those readings act any differently.  If that's the only option for altitude data stored in the .fit file, you're never going to be able to fix that.

5693709 said:

I think if you tried a hike with GPS enabled, you'd probably get performance very close to the e-trex. 

What do you mean? Of course I hike with GPS enabled. Last time I hiked with both etrex and Fenix in GPS mode, and discrepancies arose there. I can post both tracks fwiw 

I enable GPS, i.e. recording track and all that, but there are two fields during activity: GPS raw elevation, and post-processed elevation (I guess using both GPS and barometer information) which is the info that goes in the FIT file. When you told me about GPS elevation I thought you meant the GPS raw elevation.

My answer basically meant: "during activity, i.e. with GPS enabled and track recording, I always look at the post-processed elevation data. I never did look at the GPS raw data, but I can add it as a field in one of the pages and look at the differences".

yeah, if the raw GPS elevation data matches more closely with the e-trex, then at least you'll know that the Fenix is potentially capable of tracking similar to the e-trex.  If its possible to log the raw elevation instead of the PP elevation (or in addition to it) and its similar to the e-trex, then would that be a solution to your problem?