Vista: Using GPS Altitude instead of Barometric...?

[CheeseAndJamSandwich]

Hi,

 

I have a Vista HCx and used it the other day to track my transatlantic flight... but it reported and logged only the barometric altitude/elevation of the flight rather than the actual (GPS) altitude... i.e. it was reading the Cabin Pressure! :-(

 

Is there any way of switching this over to use GPS only for altitude/elevation?

 

I use to use my Venture Cx, which, when held up to the window for the entire flight, gave me a good track of the flightpath and altitude to import into google earth... The Vista tracks are useless in GE unless locked to the ground... The Vista HCx is a lot less fussy about being right next to the window and doesn't drop out like the venture sometimes did...

 

Cheers,

Rich.

Edited March 28, 2008 by CheeseAndJamSandwich

[oxothuk]

Hi,

 

I have a Vista HCx and used it the other day to track my transatlantic flight... but it reported and logged only the barometric altitude/elevation of the flight rather than the actual (GPS) altitude... i.e. it was reading the Cabin Pressure! :-(

 

Is there any way of switching this over to use GPS only for altitude/elevation?

 

I use to use my Venture Cx, which, when held up to the window for the entire flight, gave me a good track of the flightpath and altitude to import into google earth... The Vista tracks are useless in GE unless locked to the ground... The Vista HCx is a lot less fussy about being right next to the window and doesn't drop out like the venture sometimes did...

 

Cheers,

Rich.

1) There is no option on the Vista HCx that will give you a continuous real-time display of GPS altitude.

2) You can see an instantaneous display of GPS altitude with a menu option from the Satellite page.

3) You can record your tracks with GPS elevation rather than barometric elevation by making the following changes to the altimeter setup:

Auto-Calibration - OFF

Barometer Mode - Fixed Elevation

Edited March 29, 2008 by oxothuk

[AlunS]

3) You can record your tracks with GPS elevation rather than barometric elevation by making the following changes to the altimeter setup:

Auto-Calibration - OFF

Barometer Mode - Fixed Elevation

What does the real-time altimeter display show in this case?

[oxothuk]

3) You can record your tracks with GPS elevation rather than barometric elevation by making the following changes to the altimeter setup:

Auto-Calibration - OFF

Barometer Mode - Fixed Elevation

What does the real-time altimeter display show in this case?

It still shows the barometric elevation (e.g., "cabin presure altitude)"). But the GPS elevation will be stored in the track file.

[Hertzog]

3) You can record your tracks with GPS elevation rather than barometric elevation by making the following changes to the altimeter setup:

Auto-Calibration - OFF

Barometer Mode - Fixed Elevation

What does the real-time altimeter display show in this case?

It still shows the barometric elevation (e.g., "cabin presure altitude)"). But the GPS elevation will be stored in the track file.

If you fool around with "Zoom Ranges" and scrolling on the Altimeter page you may be able to get it to display the GPS elevation. I haven't actually done this on a pressurized aircraft. but it definitely changes the plotted elevation from the normally displayed calibrated barometric elevation to some other elevation.

[CheeseAndJamSandwich]

...

3) You can record your tracks with GPS elevation rather than barometric elevation by making the following changes to the altimeter setup:

Auto-Calibration - OFF

Barometer Mode - Fixed Elevation

This looks like just what i need, Cheers!

 

Shame that the real-time alt couldn't be gps too... But the recorded tracks is what i really wanted... :-)

[CheeseAndJamSandwich]

If you fool around with "Zoom Ranges" and scrolling on the Altimeter page you may be able to get it to display the GPS elevation. I haven't actually done this on a pressurized aircraft. but it definitely changes the plotted elevation from the normally displayed calibrated barometric elevation to some other elevation.

How do you get this to work?

I can only get it to change the, errr, zoom ranges...

 

Is it me or is the barometric altimeter more hastle that it's worth??? I kinda new where i was with the gps elevation on my Venture, allowing for the inherant errors...

Perhaps i just haven't got it set up right...

[roybassist]

Is it me or is the barometric altimeter more hastle that it's worth???

I guess that depends on your point of view. With Auto Calibration on, the barometric altimeter will be more accurate and more stable than GPS elevation. I prefer that.

[+Team CowboyPapa]

 

Is it me or is the barometric altimeter more hastle that it's worth???

Actually, it is not just you, CAJS. It's the both of us. Every time I note in a post in this forum that if one has a GPSr, any barometric altimeter is useless in the context that it adds nothing, I receive several responses indicating disagreement. Oh well, such is life. Although I understand the laws of physics and observe their manifestations in natural phenomena, I have not been successful in teaching them.

[CheeseAndJamSandwich]

 

Is it me or is the barometric altimeter more hastle that it's worth???

Actually, it is not just you, CAJS. It's the both of us. Every time I note in a post in this forum that if one has a GPSr, any barometric altimeter is useless in the context that it adds nothing, I receive several responses indicating disagreement. Oh well, such is life. Although I understand the laws of physics and observe their manifestations in natural phenomena, I have not been successful in teaching them.

So we need to hope they add a simple switch in a new firmware so that we can use the gps data insted if we choose to...

[Hertzog]

 

So we need to hope they add a simple switch in a new firmware so that we can use the gps data insted if we choose to...

Unfortunately, we have about as much "hope" for this as CowboyPapa has for us to come around to his point of view

 

I don't know why they won't add this, as you say "simple" feature; we've only been asking for it for the last 5 years or so, but in their infinite wisdom Garmin seems to be adament about saving us from ourselves.

 

As to your question about viewing the GPS profile, I'm not sure how it would work on the Vista; on the 60CSx the elevation plot (over time) seems to be showing me the barometric elevation on the far right and in the digiital readout at the bottom, but the rest of the profile looks to be the GPS elevation. If I push the left button to pan the profile, then I get the GPS elevation in the digital readout. The most recent value at the right of the profile always remains the barometric reading. My earlier comment on using the Zoom features may have been a red herring.

[+julianh]

 

Is it me or is the barometric altimeter more hastle that it's worth???

Actually, it is not just you, CAJS. It's the both of us. Every time I note in a post in this forum that if one has a GPSr, any barometric altimeter is useless in the context that it adds nothing, I receive several responses indicating disagreement. Oh well, such is life. Although I understand the laws of physics and observe their manifestations in natural phenomena, I have not been successful in teaching them.

Yes, it's me again, continuing my (possibly futile!) campaign to teach TCP that properly set-up auto-calibrating barometric altimeters on the newer Garmins really do work!

 

So, TCP, have you actually looked at a friend's unit with a barometric altimeter, or are you still relying on your view that you already know all there is to know about barometric altimeters, based on the experience with your (uncalibrated) Casio watch?

 

The reason that I (and a few other brave individuals) disagree every time you post on this subject is because we actually have first-hand knowledge that the auto-calibrated barometric altimeter is far more stable and accurate than the GPS elevation alone. Whether you NEED this accuracy depends entirely on your use of your GPSr - most people get by fine without it. In my work, I NEED accurate elevation traces as well as 2D location fixes.

 

And yes, I would gladly support a campaign to allow the user to select barometric altimeter or GPS elevation - but given the choice of one or the other, I will take auto-calibrated barometric elevation every time!

Edited April 19, 2008 by julianh

[roybassist]

Although I understand the laws of physics and observe their manifestations in natural phenomena, I have not been successful in teaching them.

Perhaps it's because what you say you understand and are trying to teach goes against the real-world experience of those who disagree with you, as well as the results of experiments.

 

For the record, I guess I should say that I’m only talking about Garmin units with auto-calibrated barometric altimeters (which is what is being discussed in this thread).

 

This graph http://timgiles.free.fr/garminalt.htm of measurements of GPS altitude versus auto-calibrated barometric altitude of a Garmin Vista over 5.5 hours not only shows that the error (whether average, RMS, or maximum) in the auto-calibrated barometric altitude was less than half that of GPS altitude, it shows how much more stable the auto-calibrated barometric altitude is. It also shows that auto calibration does not simply set the barometric altitude equal to GPS altitude at certain time intervals, as is often incorrectly stated (even by Garmin).

 

An ounce of experiment is worth a ton of pontification.

[+julianh]

Perhaps it's because what you say you understand and are trying to teach goes against the real-world experience of those who disagree with you, as well as the results of experiments.

 

For the record, I guess I should say that I’m only talking about Garmin units with auto-calibrated barometric altimeters (which is what is being discussed in this thread).

 

This graph http://timgiles.free.fr/garminalt.htm of measurements of GPS altitude versus auto-calibrated barometric altitude of a Garmin Vista over 5.5 hours not only shows that the error (whether average, RMS, or maximum) in the auto-calibrated barometric altitude was less than half that of GPS altitude, it shows how much more stable the auto-calibrated barometric altitude is. It also shows that auto calibration does not simply set the barometric altitude equal to GPS altitude at certain time intervals, as is often incorrectly stated (even by Garmin).

 

An ounce of experiment is worth a ton of pontification.

Thank you roybassist! This correlates extremely well with my own observations, and this is why I have been having this running feud with TCP for quite some time now. It might sound like "snake oil" to the sceptics, but the incredible thing is - it REALLY works! (One day in the field with a current model Garmin GPSr with a properly set-up auto-calibrated barometric altimeter would convince any sceptic.) Whether you actually NEED this improved accuracy is another matter, but if you want to know your elevation with a precision comparable to your 2D location - this is truly the way to go!

[+Team CowboyPapa]

Although I understand the laws of physics and observe their manifestations in natural phenomena, I have not been successful in teaching them.

Perhaps it's because what you say you understand and are trying to teach goes against the real-world experience of those who disagree with you, as well as the results of experiments.

 

For the record, I guess I should say that I’m only talking about Garmin units with auto-calibrated barometric altimeters (which is what is being discussed in this thread).

 

This graph http://timgiles.free.fr/garminalt.htm of measurements of GPS altitude versus auto-calibrated barometric altitude of a Garmin Vista over 5.5 hours not only shows that the error (whether average, RMS, or maximum) in the auto-calibrated barometric altitude was less than half that of GPS altitude, it shows how much more stable the auto-calibrated barometric altitude is. It also shows that auto calibration does not simply set the barometric altitude equal to GPS altitude at certain time intervals, as is often incorrectly stated (even by Garmin).

 

An ounce of experiment is worth a ton of pontification.

Hey, not to mischaracterize, I'm an open minded kinda' guy. So I'll meet you half way.

 

1. The linked data above is meaningless for drawing conclusions regarding the veracity of the method without the barometric (uncorrected or raw readings for each GPS data point. I'd like to re-examine by opinion based on complete data.

 

2. Please e-mail or PM to me your address in order that I may send you an self addressed, stamped box for you to return your GPSr to me for a brief testing period not to exceed 3 days at the conclusion of which I will return it.

 

3. In the meantime, take a visit to your closest college and seek out the math professor specializing in statistics and probability. Tell him that you have one hundred pairs of simultaneous temperature readings. One pair taken from an instrument with 3 digit accuracy. That is if the instrument reads 105, the temperature is somewhere between 104 and 106. The other pair is taken from an instrument with 2 digit accuracy. Tell him that it is your understanding that one can develop an algorithm that when applied to the readings from the 2 digit instrument, one can obtain readings of better than 3 digit (hopefully 4?) accuracy. Now, when he wipes away his tears of laughter,...........................

 

Now, on my way to half way...................

Edited April 21, 2008 by Team CowboyPapa

[dogwalkers2]

Tell him that it is your understanding that one can develop an algorithm that when applied to the readings from the 2 digit instrument, one can obtain readings of better than 3 digit (hopefully 4?) accuracy.

I really don't get what you are getting at, but what you mention above is pretty much what a Kalman filter does...

[+julianh]

Hey, not to mischaracterize, I'm an open minded kinda' guy. So I'll meet you half way.

 

1. The linked data above is meaningless for drawing conclusions regarding the veracity of the method without the barometric (uncorrected or raw readings for each GPS data point. I'd like to re-examine by opinion based on complete data.

 

2. Please e-mail or PM to me your address in order that I may send you an self addressed, stamped box for you to return your GPSr to me for a brief testing period not to exceed 3 days at the conclusion of which I will return it.

 

3. In the meantime, take a visit to your closest college and seek out the math professor specializing in statistics and probability. Tell him that you have one hundred pairs of simultaneous temperature readings. One pair taken from an instrument with 3 digit accuracy. That is if the instrument reads 105, the temperature is somewhere between 104 and 106. The other pair is taken from an instrument with 2 digit accuracy. Tell him that it is your understanding that one can develop an algorithm that when applied to the readings from the 2 digit instrument, one can obtain readings of better than 3 digit (hopefully 4?) accuracy. Now, when he wipes away his tears of laughter,...........................

 

Now, on my way to half way...................

Sure, I’ll meet you half way:

 

1. I have some similar test data of my own that I am totally prepared to share with the whole forum. (Does anyone know how to post an Excel spreadsheet on this forum, so I can share my data with everybody?) My test data covers two different sites, and shows very similar improvement in barometric elevation compared to GPS elevation as shown on the graph that roybassist linked. One of my sites was deliberately chosen to have a good (but not perfect) sky view, while the other was chosen to simulate a very bad sky view (as in deep tree cover or in a canyon or similar). Both show a similar degree of improvement. Anyone who has actually used a late-model Garmin with a properly set-up auto-calibrated barometric altimeter will have seen similar improvement in their elevation accuracy compared to GPS elevation.

 

2. Yeah, right! Why not just borrow a friend’s Garmin with barometric altimeter, or try the nearest GPS retailer, if you are REALLY prepared to try your own field tests? (A Garmin retailer should have no trouble giving you a “loaner” so you see for yourself that it really does work – you might actually become convinced and want to buy one!) And if you are NOT prepared to keep an open mind, and try some real tests, why the insistence on convincing others that you are right, when you have not actually seen the evidence first-hand?

 

3. And you point is …? The real-world accuracy of consumer-level GPS elevation is around plus or minus 20 metres at best, with a fixed location and a good sky-view; more like plus or minus 50 metres under real-world conditions of moving around with varying amounts of tree-cover, multi-path errors, etc. The precision of the barometer in my Garmin Summit HC (i.e. the smallest increment / decrement of air pressure the sensor can register) is 0.01 Pa / 0.01 mbar, which corresponds to a theoretical delta-elevation sensitivity of about 0.1 metres. Now, no-one is claiming that the auto-calibrated barometric altimeter has an ACCURACY of 0.1 metres (accuracy and precision are NOT the same thing!), and the unit only displays elevation to 1 metre increments. However, I simply fail to see why it should be impossible for an instrument with a precision of 0.1 metres to be able to maintain an accuracy of 5 to 10 metres when properly used and calibrated!

[+Team CowboyPapa]

And the point is:

1. Ok, I'll be serious about it this time.

2. Yes, the barometric altimeter on my Casio watch is sufficiently accurate for my purposes,

2a. I live at Sea Level and before leaving home I set it to zero,

2b. Climbing into the mountains 45 min later it will read 2960 feet when the roadside sign says 3000 feet (I think that most of the 40 feet discrepancy is time lag due to its sampling rate or interval between samples, note that I see about 40 feet too high on the way down),

2c. Similarly, 5 minutes later it will read 4960 when the road sign says 5000 feet,

2d. Accordingly, I am confident that it can read within several feet of sensed altitude at the time.

3. However, after calibrating to SL and staying at home, it changes due to weather changes,

3a. A day after calibrating to SL I've seen it read 180 feet,

3b. This resulted from a low pressure front having gone through,

3c. Currently, it is reading 160 feet, I must have last calibrated it to SL when the ambient pressure was higher than now.

3d. Sometimes I've calibrated it to SL only to see it display -100 ft the following day as higher pressure moves in.

3e. That is how barometers are used for weather prediction, at constant elevation, of course.

4. So what does it mean for outdoor excursions of more than a day?

4a. Calibrate to SL and drive up into the high Sierras and camp somewhere and another place a day later.

4b. Now if a weather front comes through, or not, how do I really know how high I am when the ambient pressure, at any given altitude, may have changed by the equivalent of 200 feet, up or down.

4c. How would I know if the pressure went up the equivalent of 100 feet or down 60 feet?

5. After a day or more from calibration, I'd feel lucky if the barometric altitude displayed were within 40 feet, up or down, but, how would I know that in a remote area of otherwise unknown elevation?

 

So, what is better to do?

1. Use your GPSr determined elevation directly. I feel that it is as good, or better, than the barometric (ignoring calibration enhancements from GPS sources).

2. Or, better yet as the horizontal GPSr data is basically one half the inaccuracy of the vertical, take a waypoint.

2a. Locate the waypoint on your Topo display.

2b. Now, the waypoint is between a contour of 6760 and one of 6800 feet and about 1/4 from the former and 3/4 to the latter,

2c. I would estimate my elevation to be within ±10 feet of 6770 feet.

 

About correcting the barometric readins with GPSr derived data, here is what I truly struggle with:

1. Go up the mountain and pause sufficiently, several minutes perhaps, at each elevation sign.

2. Note the signed elevation, the barometric displayed and the GPSr derived elevations for each.

3. Continue until a significantly sized population is aquired, 20 triplets should be sufficient.

4. Assume that the road signs are correct and calculate the standard deviation for both the barometric and GPSr.

5. If the GPSr is more accurate, it would have the lesser std. dev, assume 20 feet for it and 40 for the barometric.

6. Now, how can some correctional algorithm be designed to use one to correct the other and come up with predictions having a standard deviation lesser than either of the two measured. Say, 15 feet for the corrected?

 

Hmmmmmmmm..............

 

Is this that Rocket Science that I hear about all the time?

[+julianh]

About correcting the barometric readins with GPSr derived data, here is what I truly struggle with:

1. Go up the mountain and pause sufficiently, several minutes perhaps, at each elevation sign.

2. Note the signed elevation, the barometric displayed and the GPSr derived elevations for each.

3. Continue until a significantly sized population is aquired, 20 triplets should be sufficient.

4. Assume that the road signs are correct and calculate the standard deviation for both the barometric and GPSr.

5. If the GPSr is more accurate, it would have the lesser std. dev, assume 20 feet for it and 40 for the barometric.

6. Now, how can some correctional algorithm be designed to use one to correct the other and come up with predictions having a standard deviation lesser than either of the two measured. Say, 15 feet for the corrected?

TCP,

 

I (and others) have basically done this exercise numerous times when using our GPSrs (although not necessarily taking the triplet data sets, and doing the statistical post-processing that you describe). We don't need to do the statistical data processing to see the results with our own eyes - it's actually much simpler than that.

 

1. Start up a Garmin "S" series GPSr, get a good 3D fix, manually calibrate the barometric altimeter using known elevation (if possible), or known met bureau barometer pressure (if available), or use GPS elevation if all else fails. Make sure "auto-calibration" is turned on.

 

2. Go on a trip of what ever type you fancy - walk / bike / car / horseback (but not a pressurised aircraft!) Trip can be a few hours, or a few days, or a few weeks - you choose!

 

3. Whenever you get to a location with a known elevation (road sign, contour on good quality topo map, benchmark, whatever), check the auto-calibrated barometric elevation reading and the GPS elevation reading. The barometric elevation reading will be right to plus or minus 5 metres or so (based on my experience, and the experience of other Garmin users), and will be very steady while you stay at that location as long as you like. The GPS elevation will be anywhere from plus or minus 10 to 50 metres error (or significantly worse, depending on tree cover etc), and will fluctuate markedly if you stay put for an hour or so (and can even fluctuate in the very short term if your location has less than optimal satellite visibility).

 

4. At the end of the day, make a note of your current elevation before turning off your GPSR, so you can re-calibrate in the morning, to allow for any air pressure changes overnight.

 

5. At the end of the trip, when you get back home, compare your barometric elevation with what you started out at - chances are it will still be within about 5 metres, even if you have been gone for two weeks, and three cyclones have passed by!

 

If you actually go through this process, and diligently record the triplet data sets and post-process them, the auto-calibrated barometric elevation errors will have a MUCH lower S.D. than the GPS elevation errors.

 

Want to see independent corroboration? Check out this thread, for example, and in particular, look for "Paul Saunders" post of 10 April 2006 (based on a 60 CS, and not even using the latest software and algorithms):

 

http://forum.naviteq.net/gforum.cgi?post=20025

 

Out of 25 spot heights, 9 were identical, 12 were 1m out, and only 4 were 2m out.

 

Ever see that degree of repeatability on GPS elevation? Like it or not, THAT is what we see with Garmin units with auto-calibrated barometric altimeters!

 

With respect to your point 6 - you are missing the point. I never said the S.D. of the barometric altimeter is less than BOTH the S.D. of the barometer AND the S.D. of the GPS elevation. On the contrary, I said the barometer in my Summit HC has a precision of 0.01 hPa (0.01 mBar), which is equivalent to a delta-elevation sensitivity of about 0.1 metres, BUT the accuracy of the auto-calibrated barometric elevation is only about 5 metres or so BECAUSE it relies on the more erratic and inaccurate GPS elevation for auto-calibration. The resulting accuracy is intermediate between the precision of the two data sources - the barometer, and the GPS elevation. No big surprise there!

 

All the unit does in principle is assume that the long-term time-averaged GPS elevation is a good estimate of actual elevation, although in the short-term, it is erratic and can be very inaccurate. Using a heavily "damped" record of GPS elevation can be used to "anchor" the barometric elevation so as to compensate for changes in actual barometer pressure. The auto-calibration is not done by post-processing, it is done "on the fly". As best I can work out, the algorithm is something like this:

 

1) If at any instant, the GPS elevation is higher than the barometric elevation (say), this SUGGESTS an increase of barometer air pressure MAY have occurred since the last auto-calibration. In this case, increase the current assumed barometer pressure to partially compensate, but DON'T fully compensate - only apply 20% of the required correction (say). Also, make sure the algorithm applies a limit to the rate of change of barometer pressure to something physically plausible (say 1 or 2 hPa / hour?), so as to not make over-corrections when it happens that the GPS elevation is very erratic and / or inaccurate.

 

2) Repeat every 15 minutes (say).

 

Now think about what this means - if there really IS a significant change in background barometer pressure, repeated applications of this algorithm will make sure that the unit tracks accordingly, but will not over-react to short-term fluctuations in the GPS elevation, which are KNOWN to be erratic and relatively inaccurate in the short-term.

 

In the longer term, BOTH the GPS elevation AND the barometer elevation will oscillate about the true elevation; the difference is that the GPS elevation will be erratic and will show excursions of 20 to 50 metres (depending upon satellite observation conditions), whereas the barometer elevation will be smooth and will show MUCH smaller amplitude fluctuations - as I have REPEATEDLY said, the auto-calibrated barometer elevation errors can be SEEN to be only of the order of 5 metres typically, 10 metres worst-case, when properly used.

 

For example, if you stay at one location, the trace of barometer elevation will mimic the fluctuations of the GPS elevation - BUT THIS IS IMPORTANT: there is a noticeable time lag, AND the amplitude of fluctuations of the barometer elevation is much smaller! (I and others can show traces showing exactly this behaviour.)

 

I am not sure how much clearer I can make this, and will just repeat my plea:

Before you argue again that this is not possible, PLEASE make the effort to get your hands on one and try it out!

Edited April 22, 2008 by julianh

[dogwalkers2]

6. Now, how can some correctional algorithm be designed to use one to correct the other and come up with predictions having a standard deviation lesser than either of the two measured.

As I said before - the Kalman filter. Far too complicated to describe here. Google it and you'll get lots of math to explain what you ask, often in much more complicated situations than a simple one-dimensional (altitude) measurement. After a quick look, this link describes it fairly well, with a relatively easy to understand example: http://www.cs.unc.edu/~welch/media/pdf/maybeck_ch1.pdf See fig 1.6 for a good picture of the standard deviations...

 

edit: julianh posted while I was looking up some stuff, but what he describes is kindof the way a Kalman filter works - a recursive algorithm.

Edited April 22, 2008 by dogwalkers2

[+julianh]

And the point is:

1. Ok, I'll be serious about it this time.

2. Yes, the barometric altimeter on my Casio watch is sufficiently accurate for my purposes,

2a. I live at Sea Level and before leaving home I set it to zero,

2b. Climbing into the mountains 45 min later it will read 2960 feet when the roadside sign says 3000 feet (I think that most of the 40 feet discrepancy is time lag due to its sampling rate or interval between samples, note that I see about 40 feet too high on the way down),

2c. Similarly, 5 minutes later it will read 4960 when the road sign says 5000 feet,

2d. Accordingly, I am confident that it can read within several feet of sensed altitude at the time.

3. However, after calibrating to SL and staying at home, it changes due to weather changes,

3a. A day after calibrating to SL I've seen it read 180 feet,

3b. This resulted from a low pressure front having gone through,

3c. Currently, it is reading 160 feet, I must have last calibrated it to SL when the ambient pressure was higher than now.

3d. Sometimes I've calibrated it to SL only to see it display -100 ft the following day as higher pressure moves in.

3e. That is how barometers are used for weather prediction, at constant elevation, of course.

4. So what does it mean for outdoor excursions of more than a day?

4a. Calibrate to SL and drive up into the high Sierras and camp somewhere and another place a day later.

4b. Now if a weather front comes through, or not, how do I really know how high I am when the ambient pressure, at any given altitude, may have changed by the equivalent of 200 feet, up or down.

4c. How would I know if the pressure went up the equivalent of 100 feet or down 60 feet?

5. After a day or more from calibration, I'd feel lucky if the barometric altitude displayed were within 40 feet, up or down, but, how would I know that in a remote area of otherwise unknown elevation?

 

So, what is better to do?

1. Use your GPSr determined elevation directly. I feel that it is as good, or better, than the barometric (ignoring calibration enhancements from GPS sources).

2. Or, better yet as the horizontal GPSr data is basically one half the inaccuracy of the vertical, take a waypoint.

2a. Locate the waypoint on your Topo display.

2b. Now, the waypoint is between a contour of 6760 and one of 6800 feet and about 1/4 from the former and 3/4 to the latter,

2c. I would estimate my elevation to be within ±10 feet of 6770 feet.

TCP,

 

This discussion is SUPPOSED to be about how to calibrate the barometric altimeter on a Garmin GPSr for best results. How you calibrate your Casio watch, and the accuracy it can provide, are a whole different subject matter. (Perhaps you should go to the Casio forums for advice on that one? )

 

My Summit HC consistently gives me plus or minus 5 metres vertical accuracy (15 feet) when conducting similar tests. Also, it is truly "plus or minus" - it doesn't lag significantly, and it doesn't matter whether I am going uphill or downhill. If I walk in my front door and go upstairs, I immediately see an elevation change of 3 metres (even if I lose the GPS signal along the way). Presumably, the sensor is rather more sensitive than in your Casio watch?

 

Yes, if I leave my Summit HC turned off for any length of time, and then turn it back on, the altimeter can be out by 20 or 40 metres or more, depending on intervening weather changes. The point is - if I calibrate it manually when turning it on, it will then remain accurate to plus or minus 5 metres indefinitely, even allowing for weather changes!

 

No, sorry to disappoint you - the GPS elevation is NOT "as good, or better, than the barometric" - the auto-calibrated barometric elevation wins every time!

 

And as for "Locate the waypoint on your Topo display" - I have made this point before: what if you can't GET a decent topo map in the area of interest? (Yes, I know you Americans are lucky enough to be able to get good topos for the whole country. Some parts of the world aren't so fortunate.) My last trip into the wilds of Indonesia was a case in point - no digital topos at all, what paper topos we could get were very old, and had 50 metre contour intervals, and of very limited use when trying to do a number of preliminary site reconnaissances for a proposed multi-billion-dollar project. Having inspected a few sites, and narrowed down the options somewhat, our next phase will include detailed site survey, aerial and / or satellite photogrammetry, creation of a DEM (Digital Elevation Model), etc - but none of this existed when I went to site! However, I was able to come back with a whole swag of reasonably accurate elevation data (plus or minus 5 metres) which simply did not exist in any form before I went to site.

[+julianh]

[edit: julianh posted while I was looking up some stuff, but what he describes is kindof the way a Kalman filter works - a recursive algorithm.

Rats! And there was me thinking I had developed some sort of major fundamental break-through in digital signal processing!

 

I guess there's no chance I will get a Nobel Prize in Mathematics after all!

 

(And I guess there's no chance they will name the algorithm after me either! )

[+Team CowboyPapa]

 

3. Whenever you get to a location with a known elevation (road sign, contour on good quality topo map, benchmark, whatever), check the auto-calibrated barometric elevation reading and the GPS elevation reading. The barometric elevation reading will be right to plus or minus 5 metres or so (based on my experience, and the experience of other Garmin users), and will be very steady while you stay at that location as long as you like. The GPS elevation will be anywhere from plus or minus 10 to 50 metres error (or significantly worse, depending on tree cover etc), and will fluctuate markedly if you stay put for an hour or so (and can even fluctuate in the very short term if your location has less than optimal satellite visibility).

 

4. At the end of the day, make a note of your current elevation before turning off your GPSR, so you can re-calibrate in the morning, to allow for any air pressure changes overnight.

 

5. At the end of the trip, when you get back home, compare your barometric elevation with what you started out at - chances are it will still be within about 5 metres, even if you have been gone for two weeks, and three cyclones have passed by!

 

I agree that this is workable as long as long as all atmospheric, weather related changes occur during the evening with the GPSr off and none during the day when the GPSr is on.

I agree that I would have to see data taken with a combination of physical elevation changes during a period of significant ambient pressure change due to weather change.

[+Team CowboyPapa]

 

And as for "Locate the waypoint on your Topo display" - I have made this point before: what if you can't GET a decent topo map in the area of interest? (Yes, I know you Americans are lucky enough to be able to get good topos for the whole country. Some parts of the world aren't so fortunate.) My last trip into the wilds of Indonesia was a case in point - no digital topos at all, what paper topos we could get were very old, and had 50 metre contour intervals, and of very limited use when trying to do a number of preliminary site reconnaissances for a proposed multi-billion-dollar project.

Apologies, sincerely.

 

I missed that on the earlier pass. Absolutely, I am fortunate to have them, and for the entire US as you note. But, I don't leave with all that loaded in my GPSr all the time. I keep all those in my general area on several SD cards and can move them in and out as needed. I can see that one must use other methods when such maps are not available, quite a challenge, actually.

Edited April 22, 2008 by Team CowboyPapa

[+julianh]

I agree that this is workable as long as long as all atmospheric, weather related changes occur during the evening with the GPSr off and none during the day when the GPSr is on.

I agree that I would have to see data taken with a combination of physical elevation changes during a period of significant ambient pressure change due to weather change.

No, NO, NO, NO!!!! You are still missing the point!

 

The auto-calibration algorithm works in real-time, when the unit is turned on, and as long as you are getting a reasonable 3D satellite fix. It is correcting for changes in barometer pressure on a continuous basis. When you turn it off, and / or lose your 3D satellite fix, the auto-calibration will no longer work, and you will need to calibrate manually (just like your Casio watch).

 

I have any number of moving tracks that I have taken with significant weather changes - and I always maintain my plus or minus 5 to 10 metres vertical accuracy, even though I may be going through a barometer pressure rise or drop of 10 hPa (10 mBar) or more in a day, which would account for an apparent elevation shift of 85 metres, if not for the auto-calibration routine. In my line of work, I ALWAYS do a close-out of my elevation at the start and end of the day's records, because I NEED to know whether my barometric elevation record is reliable. I can tell you - IT REALLY WORKS!!!!!

 

(Apologies to other forum readers for the shouting, but I seem to be having some trouble getting my message across. )

[+julianh]

I can see that one must use other methods when such maps are not available, quite a challenge, actually.

Not really - buy a Garmin "S" series! That's how I solved my problem - works for me!

[gallet]

FWIW, by starting my HCx in Test Mode (ie with the "enter" key held down) I can see a continious elevation readout with 0.1m reported accuracy.

[CheeseAndJamSandwich]

How does the new eTrex 30 behave with this?

 

Can you get it to record GPS Altitude in the tracks?

And can it display the GPS Altitude, where the Vista cannot?

 

Cheers again.