[closed] Barometer based tracking

[CRider]

Hi again,
First, thanks for adding this feature so quickly. It works great and I just have some questions on how you use the altitude from the GPS.

I really like the fact that the raw pressure is recorded with each point. This allows some post processing if needed.

For the questions :

1. I've understood that the GPS is used for the initial calibration of the barometer. But how do you decide later to recalibrate regularly the barometer ?
For a tracking of several hours and less than a day, I don't expect that the pressure is affected by the weather condition or something else. Then using the GPS altitude regularly may decrease the altitude precision of the barometer.

2. What is the P0 ? What is this for ?

Best regards.

[CRider]

I just want to add that my main concern is on relative altitude (eg for computing precise ascending height) rather than absolute altitude.

[Psyberia-Support]

Hi,
The calibration of the barometer actually means the computation of P0, which is the pressure that you would have at the sea level (0m elevation), at your current location, with the current weather factors.
This P0 pressure is used as a reference for the barometric elevation computation: using the raw pressure, and regarding the current P0 pressure, a precise elevation can be computed.
The P0 pressure is computed at the beginning of the track, at the end, and approximately every 15 minutes. It's important to keep an updated P0 pressure, since the raw pressure is altered by the weather changes, but also the elevation change itself.
If you look at this graphic (almost flat area), you can see that the P0 is not so constant within one hour:


P0 gives you another information related to the weather. The standard P0 pressure is 1023hPa, so if P0 goes under this vales, you are in a low pressure zone, if it goes over, you are in a high pressure zone. It can give you tips about the coming weather: http://weather.about.com/od/pressureand ... essure.htm.

A good calibration is very useful for absolute altitude computation, it's less important for cumulated altitudes, which deals with much higher values.

[CRider]

Hi,

1. Do you use only the altitude based on barometer for cumulative height gain / loss ?
As you agreed, the absolute altitude is not really important in this case.

2. For computing the absolute altitude, relying on a GPS altitude measure every 15 minutes may introduce errors caused by the precision of this measure.

Regards.

[Psyberia-Support]

The barometric elevation, if present, is used for all statistics.
You can see which elevation is used using the "Elevation" title in the details, followed by 'Baro.' if the barometric data is used (like in the capture).

Again, the GPS altitude is only used to calibrate the barometer values on a regular basis, which actually mean the computation of P0 (sea level) pressure, which is plotted in light grey on the previous capture. The values are in fact means of few consecutive GPS elevations, to reduce the inaccuracy as much as possible. Then, the GPS values are no more used.

[CRider]

Hi again,
I'm sorry but I'm not convinced at all.
Your barometric elevation / altitude statistics rely on GPS altitude which introduces errors and inaccuracies.
Take a look at a recorded track I've just made around a very very flat area.
The raw pressure level is nearly flat which is correct.
But the geoide altitude has a lot of fluctuation with 75 m between the highest and lowest altitude.
Your baro altitude has less errors but follows the mean profile of the geoide altitude.
As a result, the statistics are not really relevant.

Can you please add an option to take only once the GPS altitude ?
This is enough to have a rough absolute altitude. This will not introduce inaccuracies and errors for the relative altitude based on the barometer.
Thanks.
Regards.

[Psyberia-Support]

The GPS elevations are not used for the computation of the barometric elevations, so they can't be related.
They are only used to compute the general trend of the pressure changes.

The important think to understand is that the barometric elevation is precise on a short term, but not on the long term, whereas the GPS is not much precise on the short term, but will remain stable on the long term.
What is done with the GPS elevation is only to compute the general pressure variation trend - p0, in order to keep the barometer calibrated.

What would be interesting to plot concerning your track is the GPS Elevation vs the Sea Level Pressure, with the Time as X.

Concerning the fact of having only a single calibration, it will only give you less precise results since this first calibration will not be more precise than the updates (the same GPS elevation based method is used for all points), and it will become quickly obsolete due to the atmospheric changes (if a storm is coming, the pressure can drop quite fast).

This feature is new, so will may be need few tweaks, based on tracks I'm gathering. Right now the pressure trend is updated every 15 minutes as explained before, using the mean of few GPS elevation values to reduce the inaccuracy. Between these points, the GPS elevations are not used at all.
This update time is based on an example I read in this very good study (pdf file available here).

[CRider]

Hi,
Thanks for the link.
I think that for computing the statistics, some settings / parameters may be useful like you have some settings for the tracking.
I'd like to experiment with some recorded tracks but at the moment I cannot (see my post on GPX with baro pressure and GPS precision).
Regards.

[Psyberia-Support]

Hi,
Yes I also had this idea of letting the user changing some parameters.
I'll try to do it for the next version.
Best regards

[CRider]

Hi again,
That would be awesome.
Thanks also in advance for the GPX export with baro pressure and GPS precision.
Regards.

[Psyberia-Support]

I have released version 1.3.12 with the ability to change some statistic settings, including the fact of using barometric elevations or not to compute them.
Best regards

[CRider]

Hi,
The new settings are awesome.
Would it be possible to have additional settings related with the calibration of the pressure with the GPS altitude (e.g. periodicity of the calibration) ?

I've noticed also that the barometric elevation setting is not disabled for a GPX track without pressure data. That's not a big problem though.

Regards.

[Psyberia-Support]

Hi,
Yes, all settings are always available (even the Barometric elevation one) since these are 'global settings'. They are not bounds to the tracks themselves, and they are shared with the statistics of the Location Tracker too.

Concerning the calibration setting I'll try to add this for the next version. You're right that it would be a great feature to have, as we have discussed for long about the best way to calibrate, it would allow more flexibility! But maybe this one should be saved for each track (so it would allow the user to enter manual calibration values).

Best regards

[CRider]

Hi,

Concerning the calibration setting (...) But maybe this one should be saved for each track (so it would allow the user to enter manual calibration values).Best regards

That's a very good idea. But then, the barometric elevation setting (on/off) in itself could also be saved for each track.
And if there is no barometric data for a given track, this setting could be disabled.
Thanks.
Best regards.