Jump to content

Elevation reading on a GPS


Guest Justin

Recommended Posts

Hey folks-

 

I have a Magellan GPS Pioneer and my elevation function is basically useless. If I have a strong signal (tracking satellites), then the elevation will fairly slowly change, jumping up or down 3 or 4 feet at a time, but to get an accurate reading can take an hour or more, especially in the mountains. Sometimes it stops moving, even when the reading is nowhere close to my real elevation. As a practical matter, I don't think it's very useful, especially for drastic elevation changes.

 

Is this common to all GPSs? Is there something wrong with mine?

 

Thanks a ton,

Justin

Link to comment
Guest klannamann

Hi Justin-

I pretty much ignore mine too. (Magellan 300) When I am on a boat at sea level I know my elevation isn't 90 feet, even at high tide. I think the GPS they use for landing airplanes is alot more accurate... Seems to me the rest of us can just be happy with the horizontal accuracy based on the angles to the satellites, but the elevation will only be good when you have several satellites nearly overhead, or if you have a very expensive GPS with a real accurate clock.

Ciao~

Ken

Link to comment

The height varies a lot because GPS works by crossing mathematical lines to determine your position. The lines all happen to be fairly close to vertical - which gives you good precision for horizontal position, but not such good precision for vertical position, since they're crossing each other at a pretty shallow angle.

Link to comment

The eTrex Summit has a built in altimeter, but even then you need to find a place where you know the exact elevation to calibrate it.

 

If it doesn't have an altimeter in it, I wouldn't trust it! I guess they put it in in the off chance you get close to your real altitude.

 

Jeremy

Link to comment
Guest Richard Amirault

The elevation data is the *least* accurate data from your GPS. I've been at local seashore and noticed a reading of a hundred feet or more at times.

 

However, I've heard that to get the most accurate altitude reading your GPS needs to see the *entire* sky of sats.

 

I was once on top of Mt. Cadilac in Bar Harbor, Maine and pulled out my (old Garmin 45) GPS. It gave me a reading within 10 ft of the height posted on a plaque at the summit!

 

Richard in Boston

Link to comment

I fired up my M315 on a commercial flight. 630 something MPH at 500'. I always knew the elevation was bad, but we were over 30000'. It could only get a lock on 3 sats, and one of them was intermittent at best. With it locked on to more satellites, the elevation improves a lot, but even with a sky full of good signals the error can still be a couple hundred feet.

Link to comment

I used to be the Army and we had what was called a PPS(Precise Positioning System) as oppposed to the commercial GPSes that are known as SPS(Standard Position System). Even with the precise system we still couldn't get any better vertical accuracy. We had horizontal accuracy up to a meter with the precise system. Basically the precise system had a crypto key that was loaded into it that made it so that specific timing information could be downloaded into the unit so that it could compensate for the Selective Availability errors that used to be introduced. Basically I think that the GPS system wasn't designed for vertical accuracy.

Link to comment

From what I have read GPS's are setup for horizontal accuracy. In most cases they have four or more satellites in view. Ideally you should only need 3 satellites to get a 3D fix if the clock in your GPS was perfect but it is not so you need at least four satellites for a 3D fix and 3 will get you a 2D fix at some assumed altitude. I know some Garmin units will let you enter an estimated altitude if it can only see three satellites. None-the-less, the GPS uses a least square fit to the available data, but since horizontal position is usually more important than vertical position the solution to the over-determined system is optimized to give you good horizontal position at a sacrifice of vertical position. In most cases if you EPE is, say, 20 feet then your vertical error will be about 1.5 times this. Since most unit report a pretty optimistic EPE you have to take even that with a grain of salt. I think I read somewhere that Garmin units report a 50% confidence number in the EPE slot. That mean that if it say you have an error of 20 feet you are going to be within 20 feet of the actual point reported 50% of the time. To get a say 95% confidence number you would need to nearly double the EPE number again. So when you GPS says the EPE is 20 feet your can bet that your altitude could be off by more than +/- 60 feet. This is again a very conservative look at the possible error.

 

Another point when looking at the altitude with respect to see level. Remember your GPS models the world as a great big ellipsoid. This ellipsoid minimizes but does not eliminate the error between what it call zero altitude and what sea level is. The WGS-84 put an altitude of zero at the average sea level. The actual planet is an ellipsoid but it is a very lumping ellipsoid. I was landing in San Diego and had my GPS on during landing. I know slightly illegal but? icon_wink.gif It reported -110 feet as the wheel touched down. If you ever been to San Diego you know the airport is only a few feet above sea level. Also remember that you unit does not change with the tides either. Lot of reason here for you GPS and sea level not to agree.

 

Hope that helps a little.

Link to comment

From what I have read GPS's are setup for horizontal accuracy. In most cases they have four or more satellites in view. Ideally you should only need 3 satellites to get a 3D fix if the clock in your GPS was perfect but it is not so you need at least four satellites for a 3D fix and 3 will get you a 2D fix at some assumed altitude. I know some Garmin units will let you enter an estimated altitude if it can only see three satellites. None-the-less, the GPS uses a least square fit to the available data, but since horizontal position is usually more important than vertical position the solution to the over-determined system is optimized to give you good horizontal position at a sacrifice of vertical position. In most cases if you EPE is, say, 20 feet then your vertical error will be about 1.5 times this. Since most unit report a pretty optimistic EPE you have to take even that with a grain of salt. I think I read somewhere that Garmin units report a 50% confidence number in the EPE slot. That mean that if it say you have an error of 20 feet you are going to be within 20 feet of the actual point reported 50% of the time. To get a say 95% confidence number you would need to nearly double the EPE number again. So when you GPS says the EPE is 20 feet your can bet that your altitude could be off by more than +/- 60 feet. This is again a very conservative look at the possible error.

 

Another point when looking at the altitude with respect to see level. Remember your GPS models the world as a great big ellipsoid. This ellipsoid minimizes but does not eliminate the error between what it call zero altitude and what sea level is. The WGS-84 put an altitude of zero at the average sea level. The actual planet is an ellipsoid but it is a very lumping ellipsoid. I was landing in San Diego and had my GPS on during landing. I know slightly illegal but? icon_wink.gif It reported -110 feet as the wheel touched down. If you ever been to San Diego you know the airport is only a few feet above sea level. Also remember that you unit does not change with the tides either. Lot of reason here for you GPS and sea level not to agree.

 

Hope that helps a little.

Link to comment
Guest Mike_Teague

the reason the altitude is less accurate than horizontal positiion is simply because of the geometry of the satellites. to get good vertical reading, you would need at least 1 sat "under" you, and since theres the big problem of the earth blocking everything "under" you, you only have sats above, and to the sides.... there was a java app on some website (might have been trimble's) that would let you drag sats around a simulated "sat status" screen, to experiment with satellite geometry and DOP.. it was interesting... cant find the site tho...

Link to comment

The previous poster had some good information, here's some more:

 

The closer the satellites are to the horizon, the more accurate the position (horizontal and vertical). Ideally, you'd want a good spread all throughout the sky, that will limit the error.

 

This is why you can't get good vertical when you're in a valley: the earth on either side of the unit prevents a good spread, hence a good read.

Link to comment
Guest Mike_Teague

actually if you only have 4 sats to work with, the best DOP will be 3 on the horizon spread out 120 degrees (so like one to the north, one to the WSW, one to the ESE), and one directly overhead..

Link to comment
Guest TG4W-CSMS

I have an eTrex regular model as well, and my altitude feature seems to be nearly as accurate as horizontal navigation. I had my eTrex on for the entire duration of a flight from buffalo to atlanta, and I noticed maximum elevation was 34,859. When I asked the pilot about maximum elevation, he checked for me and said it was 34,841. I'd say that is pretty good, especially considering I only had 5 satellites while in my seat.

Link to comment
Guest Mike_Teague

Airplanes dont use GPS altitude, or any _true_ altitude above-the-ground. They use barometric, or pressure altitude.. Completely based on air pressure, not real altitude...

 

 

[This message has been edited by Mike_Teague (edited 02-04-2001).]

Link to comment

If you are interested in the latest and the greatest in the technology of GPSs logon to

spacedaily.com

Motorola is producing cell phones with GPS capability! Yes, planes use them also.

Link to comment
Guest blackfrancis

The posts regarding geometric dilution of precision as being the cause of poor altitude accuracy are right on. However, I've found the altitude to be good to within one or two hundred feet since SA was turned off. When SA was still on, the altitude was off by more than 1000 feet much of the time, so we should be pretty happy with what we have. Whenever GPS is finally approved by the FAA for aircraft use, aircraft will be using some form of differential GPS like

Wide Area Augmentation System (WAAS) or Local Area Augmentation System (LAAS). The WAAS will be accurate to within about 5m, and the LAAS will be accurate to within about 10cm.

 

One other thing to note is that, to my knowledge, most of our hand-held GPS units are reporting altitude above the WGS84 map datum. The WGS84 datum is an ellipsoidal approximation to the true geopotential surface of the Earth, and it can be off by tens of meters. As I recall, the difference between WGS84 altitude and true sea level in the Silicon Valley area is about 30 meters.

Link to comment

I have access to a (supposedly) sub-meter accuracy backpack GPS which uses differential correction. I say "supposedly" sub-meter because we've tested it by sitting on known points throughout the southeast. Depending on where you are the horizontal accuracy varies from a foot or so to a couple of meters.

 

As far as vertical, forget it. As other posters have mentioned, gps's use the wgs84 ellipsoid as "0 elevation". Depending on where you are, the difference between actual ground elevation and the ellipsoid can be tens of meters. In addition, the vertical accuracy is, roughly, 1/10 of the horizontal accuracy.

 

Unless you use a kinematic setup (big $$$) don't expect to get decent vertical accuracy.

Link to comment
Guest Brassman

quote:
Originally posted by Mike_Teague:

Airplanes dont use GPS altitude, or any _true_ altitude above-the-ground. B]


 

Airplanes use radar altimeters to measure and display AGL heights during approaches for landing.

Link to comment
Guest Brassman

The P3s the US Navy uses for deploying sonobouys has a radar-altitude-hold function on the autopilot. But I don't want to start an aviation discussion here.

 

It will be interesting to see what happens to my altitude display when I seach for the cache hidden halfway up Stone Mountain, GA next weekend. If I find it I'll log the diplayed altitude and encourage others to do the same to see what kind of variations there are.

Link to comment
Guest Nostromo

I placed my eMap next to a Geological Survey Marker on top of a mountain. The Marker read 4027'

 

I set my eMap on Waypoint Averaging, walked away, let it take 47 readings, and it read the same as the marker. eMap estimated a 12' accuracy.

 

I'm not saying this is the norm. It was a clear day, and no overhanging trees, and I was at 4000 feet.

Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...