Barometric Altimeter Sensor Errors

[Isaiah 40:31]

I've been wanting to have a GPS for a VERY long time. I finally have one now, so you'll probably hearing a fair bit from me for the next little while.

 

I am using a Garmin eTrex Visa Cx which has the magnetic compass and the barometric altimeter. I obtained the current altimeter setting from listening to the ATIS radio transmissions from the local airport and set the barometer settings on the GPS accordingly. I then went out for a short hike this afternoon to play with some of the features of my new toy.

 

As I was hiking I noticed thta not only did the GPS tell me what my altitude was at any given moment, it told me what maximum altitude I got to, and "Hey, isn't this cool!" I thought to myself, everytime I went UP at all, it would add it to a cumulative "Total Ascent" field on the display. I thought this was a great idea as it would tell you exactly how much actual climb you had done, and not just overall elevation change.

 

NOW FOR THE QUIRKS:

 

I ended up hiking out to a beach at a local lake on a stormy afternoon. When I got out of the tree cover onto a little peninsula I was met with the harsh winds. As anyone who understands Bernoulli's Principle would know, air flowing PAST a surface creates a low pressure area AT THAT POINT on the surface. As I was holding my GPS up, sometimes I had the sensor (later realized it was on the BACK of my unit) facing into the wind, and other times it was sideways to the wind. When the wind blew towards the GPS (increase in pressure) the GPS showed a descent, when the wind gusted past the GPS it decreased the pressure in that zone (point on the surface of the GPS) thus showing a gain in elevation.

 

So much for having an accurate accumulation of elevation.

 

PART II: Automobile Fun (?)

 

If you've ever thought about it, whenever you decelerate your vehicle the air pressure will increase in the front of the passenger compartment (inertia of the air causes it to continue to the front of the vehicle increasing the pressure) and whenever you accelerate the air pressure will decrease in the front of the passenger compartment. (Don't believe me? Put a helium balloon in a car and watch what it does when you slam on the brakes.) Sure enough I was watching my altimeter on the GPS and whenever I braked you could see a descent. The reverse will occur if you place the GPS toward the rear of the passenger compartment.

 

One more thing about the automobile. Have you ever been driving down the highway and you rolled down the window and your ears popped? Or closed the window and you felt the pressure build up in your ears? What do you think is going to happend to the readout on your GPS? (Rhetorical Question)

 

So, what do you think? Are there too many sources of error for the cumulative ascent total to be useful?

 

Cheers, all!

Edited January 3, 2007 by Isaiah 40:31

[SandyGarrity]

I ended up hiking out to a beach at a local lake on a stormy afternoon. When I got out of the tree cover onto a little peninsula I was met with the harsh winds. As anyone who understands Bernoulli's Principle would know, air flowing PAST a surface creates a low pressure area AT THAT POINT on the surface. As I was holding my GPS up, sometimes I had the sensor (later realized it was on the BACK of my unit) facing into the wind, and other times it was sideways to the wind. When the wind blew towards the GPS (increase in pressure) the GPS showed a descent, when the wind gusted past the GPS it decreased the pressure in that zone (point on the surface of the GPS) thus showing a gain in elevation.

An interesting point! As a mountaineer try to keep the unit as dry as possible and hence it is always covered, the west coast of Scotland has some spectacular rain!

 

I usually find that the unit is pretty accurate with total height climbed, but always use a map hence the GPS is a double check.

 

I wonder what it makes of all the hot air in the pub at the end of a long day on the hill?

[mtbikernate]

This general idea gets dealt with constantly on bike discussion boards, because cyclists like to know their total ascent for a ride, too. The point is, the information is useful, but don't take it at 100% truth. There are inherent errors (you pointed out two...there are more), so the readout will never be 100% spot on.

 

More sources of error:

Calibration errors - For example, if your GPS wasn't at the exact position as the airport sensor at the exact time the pressure reading was taken, then either the altitude or the pressure (or both) will be somewhat different between that position and your actual position.

 

Model of the earth - The earth is not a perfect sphere, and many altitude models make that assumption. Also, how are you defining sea level? Actual sea level varies quite a bit from location to location and from one time to the next due to tidal fluctuations. Realize that during tidal fluctuations, that ocean water displaces air, so air pressure at a given coastal spot can change ONLY because of tidal changes over the course of a day.

 

Weather systems - If a weather system is moving in, you will notice some drift in your barometric readings. This is another reason that the altitude readings might drift if your location has not changed. This change is not very perceptible if you are moving since it usually occurs slowly, but over the course of a day, it can add up.

 

Resolution - How often does the altimeter record a measurement? How fast are you moving? If you are moving quickly over rough terrain, you could move up and over a measurable height, but the altimeter may not record it because it is not taking readings quickly enough.

Edited January 3, 2007 by mtbikernate

[cmdigests]

Temperature changes will also affect altimeter/barometer readings.

 

If you have four or more satellites locked, then your GPS can determine your elevation from the satellites. You can calibrate your Vista's altimeter/barameter based upon that GPS lock rather than manually entering the altitude and pressure.

 

Supposedly some of Garmin's units automatically calibrate the altimeter/barometer every 15 minutes based upon the information from the GPS satellites.

[Isaiah 40:31]

mtbikernate, that is an interesting perspective I had not considered the ramifications to cyclists, or others using their units at greater velocities than walking speed with the unit exposed to the air.

 

As for sources of error, nate, those are all good ones, you forgot about in mountain saddles, or even in valleys where there is a wind blowing between the mountains (the wind need not be to the surface) there, too, will be a low pressure area. Barring those geographic anomolies, I was within 2 or 3 miles of the airport, so my settings were good in terms of barometric pressure - granted I could have been in the eye of a hurricane, but I wasn't!

 

However, your statement about my altitude is incorrect (unless I misread it). While it is true that if you have three different barometers at three different altitudes, the readings are going to be different (they will reflect the reduction in air pressure for each subsequently higher barometer - this is called STATION PRESSURE); the "altimeter setting" given by ATC (air traffic control) and the ATIS (Automated Terminal Information System) is a CORRECTED number. That is, it takes into account the current temperature, humidity and the elevation of the airport such that when you use that setting on an aircraft altimeter (or in this case in the GPS, calibrating the altimeter knowing the correct pressure) the altimeter will read your current altitude above MSL pressure (MEAN SEA LEVEL)

 

cmdigests, that is true, and is certainly useful information if you know niether your actual current altitude or the current pressure; however, we know that altitude is the least accurate reading supplied by a GPS, so you could be out 30 meters or more. One of the main thrusts of my argument is that the GPS has a barometer purportedly for the purpose of giving more accurate readings of altitude. And it will, provided you don't breathe on it!

 

Supposedly some of Garmin's units automatically calibrate the altimeter/barometer every 15 minutes based upon the information from the GPS satellites.

 

cm, I wonder which units do that? I set my gps around noon today (using ATIS - this time one mile from the airport with flat land the entire distance between me and the airport) with a reading of 29.76 Hg. By the time 19:00 came around, the display on the calibration screen had not changed from that number. I don't know if that means anything or not.

[Hertzog]

cm, I wonder which units do that? I set my gps around noon today (using ATIS - this time one mile from the airport with flat land the entire distance between me and the airport) with a reading of 29.76 Hg. By the time 19:00 came around, the display on the calibration screen had not changed from that number. I don't know if that means anything or not.

Yours for one Look for an "Auto Calibration" selection somewhere. On the 60/76 series it's on the Altimeter Setup page, but on my old (pre Cx) Vista it's on the System Setup page, under "Altimeter"; so it may still be there on the newer Vista models.

Edited January 4, 2007 by Hertzog

[Isaiah 40:31]

Yours for one Look for an "Auto Calibration" selection somewhere. On the 60/76 series it's on the Altimeter Setup page, but on my old (pre Cx) Vista it's on the System Setup page, under "Altimeter"; so it may still be there on the newer Vista models.

 

You're absolutely right, Hertzog, thank you. Unfortunately I don't see any indication of when or by how much it has 'auto calibrated' it.

[Hertzog]

You're absolutely right, Hertzog, thank you. Unfortunately I don't see any indication of when or by how much it has 'auto calibrated' it.

It's a continuous autocal; the barometric altimeter is incrementally adjusted toward the GPS elevation each second. You can see it by manually calibrating the elevation to a large error, say 1000 feet, then watching the elevation correct itself over a period of time. Assuming the Vista now has a decent elevation profile page (the original didn't, but they should have corrected that by now), you can see the elevation changing exponentially with a time constant of about 30 min. There is also a fast "catch up" mode, where the time constant is changed to about 30 seconds; this comes into play when the GPSr loses and regains track.

[+Alan2]

So, what do you think? Are there too many sources of error for the cumulative ascent total to be useful?

 

Yes!

 

Another problem I've noticed is when you lose the satellites. For example, I once drove through a tunnel under the East River that connects Queens to Manhattan. Naturally I lost the sats going through but the altimeter kept logging its correct barometric altitude. When I got home I transfered my Vista's tracks to my computer map and noticed that I was still minus -60 feet at the toll booth after exiting the tunnel. Apparently, the toll takers were collecting money 60 feet underwater.

 

What actually happened though was the program kept the last altimeter reading until the GPS re-locked on the satellites and then updated the track and altimeter to above sea level at a point long after leaving the tunnel. This kind of thing would happen on the trail too with heavy foiliage as satellites lock in and out. Of course, Garmin may have changed the program since I tried this a few years ago, but it still seems like satellite locks effect the final results.

Edited January 5, 2007 by Alan2

[kb9nvh]

For what its worth, I find that for short time periods (1/2hour) if in poor sat reception conditions the auto calibration should be OFF as it will cause variation that would not be there normally.

 

Also, yes you have to keep in mind the inaccuracies and what can cause them (pressure change by opening car windows, weather patterns and so fourth). Even the altimiters on airplans have to deal with this. You set them before you go and reset them when you get where your going. You reset all the time for barometric pressure.

 

Anyway, for best accuracy when I run I set autocalibrate to OFF, set the track to update as fast as possible and turn off lock on road.

 

 

quote name='Hertzog' date='Jan 4 2007, 12:24 AM' post='2644678']

cm, I wonder which units do that? I set my gps around noon today (using ATIS - this time one mile from the airport with flat land the entire distance between me and the airport) with a reading of 29.76 Hg. By the time 19:00 came around, the display on the calibration screen had not changed from that number. I don't know if that means anything or not.

Yours for one Look for an "Auto Calibration" selection somewhere. On the 60/76 series it's on the Altimeter Setup page, but on my old (pre Cx) Vista it's on the System Setup page, under "Altimeter"; so it may still be there on the newer Vista models.

[Hertzog]

For what its worth, I find that for short time periods (1/2hour) if in poor sat reception conditions the auto calibration should be OFF as it will cause variation that would not be there normally.

One potential problem to be aware of here is that when you turn off auto cal, the elevation readout reverts back to the last manual calibration, but the auto cal continues to work on elevation data as if it had remained on; when you switch back into auto cal it will be as if you had never turned it off (this is an interesting little aspect of the auto cal operation that I only recently discovered). If the reception was sufficiently bad during the period you had it off, then the GPSr probably will go into the fast "catch up" mode after regaining lock, so it may not present any major operational problems.

[Hertzog]

Anyway, for best accuracy when I run I set autocalibrate to OFF, set the track to update as fast as possible and turn off lock on road.

With the "x" model, I've gotten to the point where I just always leave the track log set to 1 second and forget about it. If I don't need the data immediately, every couple of weeks I just copy all the gpx files that have been created over to my computer and delete them for the GPSr. That way in the future I can always go back to a particular day and see what was happening (exactly where did I take that picture back on June 25th 2006?)