I'm a little over my head understanding the vertical datums used by Google Earth and how well they relate to the real world. It's causing some mild consternation on the part of some of the folks who rely on my work establishing elevations and GP's on cell towers and the like. I'm having some difficulty in explaining to them why the elevations I've generated combining OPUS & conventional techniques don't sync with Google Earth. Anyone care to shed some light on the subject for me?
I've explained to them that Google Earth is not a surveying tool, but I confess to not understanding exactly how GE is generating it's verticals.
Thanks in advance. Willy
> I'm a little over my head understanding the vertical datums used by Google Earth ...
I only use the free version so I haven't seen GE elevations. Just guessing, but it would have to be (nominally) either ellipsoid elevations or NAVD88.
What are the magnitudes of the differences you are seeing? And compared to what?
Norman,
It's a fairly drastic difference. The tower I'm getting grilled on has an elevation of 355' in GE. Doing static/OPUS on two control points and conventionally transfering the elevation to the tower base produced an ellipsoid height of 390.11' and a ortho (NAVD88/GEOID09)elevation of 366.00'. Enquiring minds want to know why and if I've gobbed things up. I don't think so.
As best I can determine, GE's elevation model is seriously off up here, but without knowing the how and why, I'm keeping my mouth shut.
The free version does give me elevations.
Whatever datum they almost use, the values are smoothed and don't follow detailed terrain features. I'd almost guess they transferred the contours from the USGS topo maps.
I get the same problems here in Hawaii relating to 1-A's for cell towers. They always seem to question the elevations because it does not correspond to Google Earth. I also have to face the questions on why the datum I provide is not NAVD88 or NGVG29.
> The tower I'm getting grilled on has an elevation of 355' in GE... and a ortho (NAVD88/GEOID09)elevation of 366.00'.
11 feet doesn't seem like all that much to me, considering the likely source of the GE data. I've definitely seen GIS contours off by that much.
11 feet sounds like mind-boggling vertical accuracy to me! I would consider that a fluke and characteristically not that good by any measure.
Want reliable topographic elevations? Get a national government topo map.
It's my understanding that it would be like getting an elevation
off of a topo that failed to use enough breaklines. In some areas
it would be right on but in areas with greater "relief" you'll have
to be careful.
my guess would be they are using EGM96 geoid for their elevations. I get similar but not same values if I use the EGM96 geoid for my GPS processing.
Recently did a profile of a dam in UTM Zone 15N and the grades were about 4 feet higher than Google Earth shows.
😉
> 11 feet sounds like mind-boggling vertical accuracy to me!
I've just spent a few minutes comparing GE elevations with some NGS monuments I know exist here in the Tulsa area. The NGS elevations match the GE anywhere from 3 - 10 feet +/-.
SWAG
I don't know this for sure, but ....
I'd guess it's based on something like a 3 meter DTM that USGS developed for the country.
In fact a couple of years ago I took a FEMA workshop in WV and they were proud of the fact that they were generating new maps for the entire state based on a 3 meter DTM. And then go on to tell us that for eLOMA's we have to determine BFE's to 0.1'. As politely as I could I said "You're joking, right?". They really couldn't give a good response other than that's what you do. As you can imagine stream gradients are fairly steep and BFE's to 0.1' based on a 3 meter DTM are stupid!
Well, we should all have ......
JFGI.
Googleing it turned up a Wiki article and the following is there:
Google Earth uses digital elevation model (DEM) data collected by NASA's Shuttle Radar Topography Mission (SRTM).
Now I know Wiki isn't perfect, but this seems logical.
Funny this should come up...
Just yesterday I did a topo on a lot for a builder's site plan. I usually would just assume 100 on a point and knock it out. This particular HOA specified "elevations to be related to mean sea level". I know that that sounded good to the HOA, but I also know that they probably didn't realize what they were really asking for in that requirement.
I checked the County GIS contours and they seemed a little weird. Maybe pre-development. I went to Google Earth and moved my cursor to the intersection in front of the house. It jogged between 600 and 601. I decide to use the 600 which was actually darn close to the County contour. As long as I reference my elevation as assumed, I'm good to go. I get to the site and begin chatting up the neighbor, as usual, to glean any information he can provide about issues in general about the place. We discussed a drainage problem on my lot and some other items. He then asked how I would be establishing my elevations for the topo. I explain and he say's "wait a minute" and walks off. He comes back with a set of sanitary sewer as-builts and asks if they would be helpful. You bet. I look at the rim elevation for the manhole at the 600-601 intersection. Elevation...601.05.
Get some, Google Earth!
Funny this should come up...
that is interesting... where do you turn the elevations on for GE?? I'm not seeing them.
Funny this should come up...
I believe it is a part of the choices under tab VIEW and click STATUS BAR on
It should show up at the bottom right of the screen between lat, long and eye alt
😉
Funny this should come up...
Andy
I don't think I've ever turned elevations on. But at the bottom of the screen it is listed with the imagery date, the lat & long and the eye altitude. Maybe a setting in your >Tools>Options if yours doesn't show.
FWIW, it had a difference of about 7 feet compare to a manhole I had shot a few weeks ago with good vertical control.
> Enquiring minds want to know why and if I've gobbed things up. I don't think so.
I'm generally an easy going guy, but I'd be temped to have this reaction:
"Okay, let me get this straight.
You hired me, a professional land surveyor with twenty-five years experience, to determine the elevation of your site. Now you question my work because it doesn't agree with a free web based program that maps the entire f'ing world.
Here's your check back, don't ever call me again"
Just to satisfy my curiosity, I pulled up the data sheet for a benchmark (TT0658) I know to be out in the open and in a relatively relief free area, just to do a comparison with the published values and what GE is giving me, GE; 328', NGVD29;331.94, NAVD88;338.13'. The elevation GE is giving me is 4' lower than the original published NGVD29 elevation. This matches within a foot to the difference the folks grilling me 'found'. The area of the tower in question has substantially more relief and I agree that GE's elevation model appears to be smoothed and lacks break lines, making it a really piss poor choice for a vertical check.
I've already spent too much time already tryng to explain vertical datums to some folks that just have no clue.
Thanks all & have a great weekend!
Carry on.
