Eng-Tips is the largest engineering community on the Internet

Intelligent Work Forums for Engineering Professionals

Register Log in

Setting Datum Elevations

Status
Not open for further replies.
msquared48

msquared48

Structural
Aug 7, 2007
14,745
0
0
US
OK...

Recently got a Garmin for the car and have been using it some. I've noticed though that the elevation listed appears off by 20 to 30 feet or so as I know that I am not underwater when I drive 10 feet above Puget Sound and it registers -20 feet.

Anyone know a way to calibrate the Garmin to a local USGS known datum elevation?

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask
 
Replies continue below

Recommended for you

Sort by date Sort by votes
M^2:

Follows are coordinates for GPS90 control datums in our area. These were/are used in calibrations of the GPS network, so the datum heights are quite well known and established. These are vertical controls, with the coordinates converted to NAD83, a common gps receiver setting for our area. These are in very public places, though somewhat hidden from casual users. Take a 12" straight screwdriver along with you.

FWIW: most casual-use GPS receivers are set to find horizontal positions most accurately first, and the altitude gets found only if you have enough sats in view.

SY5759
N 47° 41.811 W 122° 08.472 (NAD 83)
Altitude: 34
Coordinates may not be exact. Altitude is VERTCON and location is ADJUSTED. Location: In KING county, WA View Original Datasheet
Designation: GPS90 3D4
Marker Type: horizontal control disk
Setting: set into the top of a round concrete monument
Stability: Most reliable and expected to hold position/elevation well.

SY5751
N 47° 43.054 W 122° 09.933 (NAD 83)
Altitude: 314
Coordinates may not be exact. Altitude is VERTCON and location is ADJUSTED. (more info)
Location: In KING county, WA
Designation: GPS90 1C1
Marker Type: horizontal control disk
Setting: set into the top of a round concrete monument
Stability: Probably hold position/elevation well.

You can find more info on the web about GPS control points across the globe. An easy way is to create an account on geocaching.com, a subset of that sport/game is locating survey monuments and benchmarks. A geocacher made them part of one of his geocaches, which made it even more fun.

As far as "calibrating" your gps - the biggest error source is having multipath errors due to signals from the sats bouncing off of nearby buildings, hills, trees, etc. Notice these control points are set in large open areas, specifically to minimize those errors. Also, surveyor-type instruments typically use averaging over long periods of time to achieve sub-1m accuracy, along with techniques like differential GPS. A subset of DGPS is to set your receiver to allow or enable WAAS, aka "wide area augmentation system", where local ground stations (air and sea nav radio stations and the like) broadcast signals that your gps receiver can integrate with sat info to better locate you.
 
Upvote 0 Downvote
Pud,

They can move (earthquakes, erosion, subsidence), or be moved (think construction and/or vandalism), or get covered by debris. "Recovering" monuments can be a PITA for a surveyor when the land conditions change, changing ownerships (monuments located on previously public property are converted to private land), and when reference points have been obliterated by subsequent construction and land use changes (forests turned to fields, rock outrcrops obscured by buildings or even eliminated by quarrying). Besides USGS monuments, there are thousands of state, county, city and even just local surveyor-placed monuments. Sometimes just figuring out how to get to the location is the challenge (I found one that was at the top of a fairly steep cliff, it won't be too many more years before erosion causes the marker to be rendered useless). There are several nearby located atop very steep peaks requiring fairly technical gear or helicopters to get access to them; I guess the original placement was made for the outstanding sight lines a surveyor would get from up there. Taking a photo of the marker is part of the game - to prove you really got there. The monuments I listed are sitting in the middle of open areas in public parks/walkways. But they are also hidden from casual passers-by under innocuous-looking and fairly heavy cover plates.
 
Upvote 0 Downvote
Note that the WGS-84 GPS altitudes are NOT altitudes above local sea level; rather, they are the altitudes about the WGS-84 geoid, which may, or may not, correspond to sea level at any given spot on the Earth. In general, the WGS-84 geoid is within about 30 m of sea level anywhere on Earth.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize
 
Upvote 0 Downvote
"In general, the WGS-84 geoid is within about 30 m of sea level anywhere on Earth."

OUCH! So, anywhere on the earth, the information could be off by as much as 100 feet? That's insane!

My real question though was ... "Does anyone know if a Garman can be recalibrated to change its datum elevation", which, considering the previous post, is now completely irrelevant if true.

Mike McCann
MMC Engineering
Motto: KISS
Motivation: Don't ask
 
Upvote 0 Downvote
The raw GPS psuedoranges result in an ECEF x,y,z coordinate, reference to the center of the earth's rotation. That is then translated into a lat/long/alt on the WGS-84 geoid, so this is the typical GPS basic output.

Because of the specific issue that M[sup]2[/sup] has with error relative to sea level, his GPS may be equipped with a further translation onto a localized geodetical datum, of which, there are literally hundreds. That may reduce the error, but I don't know if it would specifically eliminate M[sup]2[/sup]'s issues. However, bear in mind that the CONUS NAD83 datum is ostensibly calibrated onto some number of points on the North American continent, but there's still room for localized variances.

There is a further NAVD88 which is a vertical datum correction, but that doesn't appear to be part of the panoply of datums supported by the particular GARMIN receiver I looked at.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize
 
Upvote 0 Downvote
IR and btrue

I presume you are saying

1) The earth bis not a perfect sphere but a sat nav thinks it is.

2) The approximate sphere keeps changing shape anyway, especially in places with high seismic activity.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &
for site rules
 
Upvote 0 Downvote
Actually, not quite. The basic GPS uses an oblate spheroid (WGS-84), i.e., an ellipsoid of rotation (a=6,378,137 m, b=6,356,752.314 245 m) as the statistical "best fit" of the Earth's surface. Hoever, the actual surface obviously is not that cooperative. This leaves the basic GPS LLA sometimes above, and sometimes below, the actual usrface of the Earth.

Unless you apply additional fits to the basic navigation solution, like EGM96, you'll be at a certain statistical level of uncertainty, above and beyond the basic uncertainty of altitude, which is poorer than the uncertainty for lat/long. There are further corrections like NAD83, and NAVD88, that provide further refinement to the basic solutions.

And then, on top of that, things move. It's these things that give accuracy weenies sleepless nights.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize
 
Upvote 0 Downvote
M^2, you shouldn't need to worry about altitude, as long as your position is located well enough to find your way, and you don't notice any influx of seawater into your car.

"The approximate sphere keeps changing shape anyway, especially in places with high seismic activity. "

Well, it does change shape, but it does so pretty slowly in general, i.e. an inch or so a year (apparent spread rate of the Atlantic Ridge as measured near Reykjavik). More pointedly, "mean sea level" you would think to be a nice smooth shape, but it isn't. There are gravitational anomalies (lumps of heavy stuff way down deep under the crust) that make the sea surface lumpy, or at least lumpy in relation to the smooth curve of the assumed oblate spheroid. But a spheroid approximation is used because one can invert its equations to find positions via GPS timing signals; the more complicated you want to make your assumed earth surface, the more complicated the equations get, and the bigger the device you'd have to carry around to solve those equations..

Captain: "Lieutenant, where are we?"
Lt.:"Hold on whilst I decant this liquid helium into the GPS processor core..."
 
Upvote 0 Downvote
"The approximate sphere keeps changing shape anyway, especially in places with high seismic activity."

"Well, it does change shape, but it does so pretty slowly in general, i.e. an inch or so a year."

An inch a year may be the long-term average, but of course, it can be much more sudden. The mountains I see outside my window as I type this in LA rose two feet (600mm) in a minute in the 1994 Northridge earthquake, which wasn't huge.
 
Upvote 0 Downvote
CS,

Unless there is a benchmark used as a GPS datum atop one of those peaks, and even if there was, the sphere used to approximate the shape of the globe would still change very little, as it is computed from an average of many datum points across the globe. The altitude of the mountain, or even the whole range of them, can change, but the global average "spheroid" changes slowly, if at all. That was my point, not that local features don't shift suddenly and noticeably. And, as IRstuff pointed out, the GPS network is set up to give you positional (lat/long) accuracy in preference to 3d (lat, long and altitude). Changes to locations of the peaks would be noted on a topo. map somewhere, and altitude shifts too, relative to the datum spheroid.
 
Upvote 0 Downvote
The shape is actually an ellipsoid, not a spheroid, since the Earth is actually a bit squashed down at the poles. That approximation which is WGS-84, by definition, hasn't changed in 27 years, since it's only an approximation, and not subject to miniscule shifts in the crust. Where you are relative to ellipsoid, does change, and is reflected in the many corrections that can be applied to your position, depending on the type of GPS receiver you have. However, in most cases, you can also get differential GPS corrections via internet to get your position uncertainty down to a few centimeters, again, assuming your GPS receiver is capable of doing DGPS corrections. For most applications, current "standard" GPS receivers can get your position uncertainty down to a few meters, which is good enuf for most of us...

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize
 
Upvote 0 Downvote
Status
Not open for further replies.
Back
Top