Geolocation has wrong elevation

Hi Huck,

The geographic coordinates of a geo-located model coincide with SketchUp’s world axes origin [0,0,0]
The elevation of GE Terrain at those geographic coordinates also coincides with SketchUp’s origin [0,0,0]

Given you know the real world elevation of the model’s geographic coordinates…

Using TIG’s plugin, you can assign a real world reference height to the Origin, enabling you to measure other points in the model in reference to that real world datum.

Coords-Tag from Datum — SketchUp Plugin by TIG

1 Like

Minor point @colin…a mile equals 5280 feet so the highest elevation quoted by @huckrorick is roughly a .25 mile distance and it’s even less when the initial Z coordinate of 319’ is subtracted. Though your point is well made the five mile reference is exaggerated.

I wasn’t referring to Huck’s example, but about the cabin I was planning to build on top of Mount Everest. It’s 5.49773 miles high, so I understated a little.

1 Like

Well, I suppose you’ll have a spectacular view but I doubt that you’ll get many visitors! :wink:


Hi Geo,

I don’t think I understand. If the world axes origin for SketchUp model and GE Terrain are the same, shouldn’t the elevations show the same in my model and on Google Earth?

I am using this information mainly to get differential elevations (for a gravity driven water system). I was assuming a could subtract the lower elevation from the higher and therefore get the elevation difference and gravity fed water pressure. That differential is more or less the same when measured on Google Earth and measured in my SketchUp model so for my immediate purpose the fact that the elevations are different in Google Earth and SketchUp doesn’t affect my current problem (only the difference matters, not the absolute elevation). But still, the discrepancy puzzled and bothered me. Why are they different?


Huck, you need to think of models as components. If you model a McDonald’s restaurant, later you’ll want to insert copies of it at many different elevations, all over the world.

The model has a local coordinate system based upon it’s local datum (the model’s internal origin,) which is the insertion point for the model, as a component, into the world coordinate system.

The elevation information, in the “GeoReference” attribute dictionary, “ModelTranslationZ” attribute is the z coordinate, where the model component is inserted in world space.

You cannot have objects miles from a local origin within SketchUp. It causes clipping issues, and makes entering distances into the VCB problematic.

See: Examples of the “GeoReference” Attribute Dictionary

And: About the Attribute Inspector plugin

Here is an example of someone having issues, because the model origin is far from the geometry:

1 Like

The elevation of a point on Earth is expressed as height above sea level.
The elevation (z) of a point in a SketchUp model is expressed as height above the Origin / ground plane.

Su and Google Earth both use the WGS 84 datum for the earth. The zero elevation is a mean sea level and the lat, long is based on a number of spherical harmonics to curve fit the Earth. See this for more info North American Datum - Wikipedia. The zero meridian no longer goes through Greenwich for this datum. Many locales use their own local datum for elevation like the Ordinance Survey in England and others for reasons you can search if interested. ( For example boats going under bridges, plate shifts, weather. etc.)
The elevation info in Google Earth is determined from satellite measurements from a number of different sources and of course will not all result in the exact same values just because or system design, weather, orbit accuracy ,etc. The GE data will usually show the source data on the image and I would guess Google Earth does not contract for the highest accuracy data just because of cost. Google Earth also has an terrain exaggeration(.01 to 3) used for display.
You as a user must under stand the accuracy of the data used and make sure it meets your requirements. I have not seen any error analysis but it must exist.
Just some thoughts for you.

Pascal would agree with your assumption.
In general terms:
1 column inch of water exerts 0.0361 PSI

Here’s a handy reference…
Hydrostatic Pressure — Engineering Toolbox

Personally, for the task at hand I wouldn’t place too much faith in heights derived from GE.
Field measurements will tell the true story, say, a few shots with a total station.

Then again, if distances aren’t too great, one could measure actual pressure developed on site, using nothing more elaborate than an ordinary garden hose, filled with water, fitted with a gauge at the bottom.

I ran in to a puzzling result in making measurements of water pressure. I have a pipe that runs from one of the relevant locations to a lower location. I put a pressure gauge on the pipe at the bottom. It measured 15 PSI. This didn’t make sense in terms of the elevation differential that I knew existed approximately. I opened a valve in the pipe and ran some water and shut the valve. The pressure then measured 46 PSI. This was close to the expected pressure. I drained the pipe and then filled it again. I got 15 PSI. this time nothing I did changed the pressure to the expected and previously measured 46 PSI. How could this happen?

The only explanation I can think of is that the pipe is not filling completely with water, i.e. there is trapped air. Bubbles in the tube are knows to cause errors in water-based levels.

Given I wasn’t there at the time, I can only guess.
My guess is the gauge was damaged.
How could that happen?

Most likely there is no surge control mechanism in the system.
Closing the valve briskly creates a surge in pressure.

Bear in mind that water, for all intents, does not compress.
Consequently, the surge force is akin to striking the gauge’s internals with a solid object, say, a hammer.

See … Water Hammer — Wikipedia

You might try replacing the gauge.

Thanks for thinking about this. It really puzzles me. The gauge is good. I have used it on other locations with good results.

It is puzzling to me how air in the line could affect the pressure. Wouldn’t it be compressed and pass that on down the line? I guess if there were a lot of air then It doesn’t weigh what water does so it wouldn’t contribute to the pressure, but I’m running water through the line (coming out where the gauge is, before attaching the gauge) so it doesn’t seem like there could be a lot of air.

I’m really baffled.

Laws of physics are well established in this area so one needs to question what you are doing in the test to account for result since you have establish measurement device is ok.
Water is non compressible while air is compressible. When charging an empty flow restricted pipe the air vol. will be compressed until its pressure equalizes and to first order follows the universal gas law pv=nRt. Very small vol. release can result in large pressure changes. I would suggest you have pressure gauges at both ends connected via t’s such you can bleed and will give you ref. at the head input and measure points in the test system. Homes will have pressure variations as supplied by your utility and some homes have pressure regulators. Spec is probably in 60 psi g range and low in the 30 range. What you measure can be an function of neighbor’s use ? If you have regulator in system make sure its ok?
Sorry for typing errors just had split put on thumb this PM. :joy_cat:

Thank you.

The system is gravity fed (rural, private). There is no one else on the system. It is a water tank at the top end with a 1 ¼” PVC line from the tank at the top to the bottom end where I am measuring pressure. No pressure regulator. There is no pressure at the top end except for the depth of water in the tank (about 4’). The gauge is placed on a hose bib at the end of the 1 ¼” pipe. I measured 46 PSI on several occasions. Then did a test where it showed 15 PSI. After opening and closing a valve it jumped to 46 PSI. After draining the pipe and then turning water back on (at the top) I measured 15 PSI again. Opened and closed the valve, but the pressure did not go up to 46 PSI as it had previously.

Really puzzling?

What could cause that?

FYI there RE many pressure calculators on the net you can use. If I could get thru my thick head exactly what you set up is I would take crack at what you should expect=> like I said physics does not lie so some thing is wrong in your approach and I am getting lost in top vs bottom, the pipe length and exactly how you are making measurement. You should be able to make a quick SU diagram. to let us all know. Typically pressure gauge accuracy is specified at full scale so a 100 psi g one at 10% it could have 10 psi-g error at the 15 psi-g level. To make accurate measurement you want to measure close to full scale.


Can you tell us the vertical height between the water level in the tank and the gauge, say, to the nearest foot.

do you really have 104’ of head?
[click here to see a water head pressure table][1] from our good friends at Engineering Toolbox…


There is over 875’ horizontally between the upper tank and the hose bib where I attached the gauge (it is at a lower tank which is shown on the attached SketchUp drawing) so I can’t easily determine the elevation with a survey and I don’t have sophisticated survey equipment. The most accurate measure I have of the elevation difference is the measured pressure which has been measured at various times in the past. The consistently measured pressure is about 46 PSI. It will vary a bit depending on the level of water in the upper tank. The elevation difference measured on Google Earth is 95’. Google Earth is not very accurate. 95’ corresponds to a pressure of 41 PSI which is reasonably close considering the uncertainties of Google Earth.

The big question is: What mechanism could account for measuring 15 PSI, and for that measurement being different at different times (sometimes 46 PSI) when nothing is physically different in the pipes and tanks? And for it changing from 15 to 46, again with nothing physically different in the pipes?

Forgot to attach the file. Here it is.