Google Earth and Lat/Long accuracy

When I take a property boundary, bring it back up to WGS84 and make a Google Earth line drawing out of it, it matches perfectly. That is Puget Sound though, I would imagine if it were very rural it would not match as well.

Eight years ago they had some panels that did not match our GPS well (20'+) and then the last work we did there it was fixed (2010). Other than that area, everything lines up perfectly. I have all our our GPS control points we have ever tied (1997 on) in a file I can pull up on my desk and the accuracy has always amazed me. The lines do plot "map of the earth" though and that might possibly be what you are seeing. Also check the date of the photo they are using, it might need an update

Rob,
Have you been exporting your TBC project to Google Earth ?
I am running version 2.81 and I import my Acad LDD 2004 dwg file (GRID!) into TBC, then export it, and my Trimble VRS collected points, and my optical data (total station) from TBC into GE. It creates a GE kml (or kmz?) file and opens up GE for viewing. You can toggle on/off the points, optical vectors, linework, etc. I have also been using the GE file to save 2 images, one with control points toggled on, and without moving the image, another with the points toggled off. This allows me to import the image file into cad, then translate, rotate, scale to my control in my cad file. Most of my projects are in my home county in cetral TX, so I havent seen too much distortion. However, my current tree & topo project has not been cooperating. The aerial image modeled was taken with a very oblique angle and I cannot get it to orient correctly. That is by simply importing the image, not trying to rubber sheet it.

there is no standard error budget for Google.I would tell that client to call Google and ask them to stake your property.

Or tell them that if they get the phone number for the resident RPLS at Google Earth, you would be happy to call him and discuss the problem. That should keep the client busy for a day or two.

James

Are you kidding me. The google earth stuff is no better than aerial photography, cause basically that is what it is.
Elevations are sometimes several feet off in some places. It's just hit or miss.

In my area the NAD83 2011 values have a clear shift compared to the Google imagery.

It's hit and miss, I've seen it be right on, at least as close as you can tell from the photos, and I've seen at least 20'. Depends on where you are I guess. I wouldn't depend on it for much of anything.

Bryan,

We are running TBC 3.11 and I have not tried the methods you are suggesting. I do appreciate you mentioning that though and I will use that as future reference for other work.

Thanks

> In my area the NAD83 2011 values have a clear shift compared to the Google imagery.

Kevin, how much shift are you actually seeing?

Around here, (Boise, Meridian, IE populated areas) I'm getting 1-3', maybe a bit more on average. This is on known, field collected survey data, not just a Garmin waypoint. 40 miles NW, I'm getting about 200', also on known survey data.

> It's hit and miss, ........I wouldn't depend on it for much of anything.

And you're right MM. I don't depend on it for anything more than a reference to get me out into the field. The catch is that other people do (falsely) rely on it and I have to, and want to, explaining to them why they should not use it for anything other.

I don't want to go into much detail but here's the history....

-I'm given northings and eastings (Y,X from others that aren't familiar with the difference) in NAD27.
-I enter that into Corpscon to get Lat/Long.
-Then I create a NAD27 project in TBC and enter those coordinates.
-I convert the project from NAD27 to NAD83.
-I then check those new Lat/Longs in Corpscon again.

Using the correct zones/datum throughout, everything jives great.

I then export those points to a TBC project that has "ground/project" coords so that I can stake it on the ground. My physically staked location falls on the wrong side of a fence and that concerns the client.

Bottom line is I know I can trust my work, based on the X,Y (eastings and northings) I was given by the client. I'm just going to have to chalk it up to less accurate imagery in this area of Google Earth.

On an 8 mile long project I did just a few years ago that involved aerial control topo and boundary. I find the Aerial Targets that now appear in GE are anywhere from 3 to 1 foot off in any RANDOM direction. Project coordinates were Nad83/91. GE's imagery is all rubber sheeted, tweeked and compiled from different sources. Its a tool in the toolbag, thats it.

I see the shift on known NGS control points in clearly identifiable locations... In a curb island in a parking lot, corner of a bridge, etc.

The NAD83 (2011) positions are about 9-10ft se'ly of the photo location on Google Earth.

I use GE to view control locations and as a planning tool. I don't rely on it for anything beyond that.

The CORS 96 values when I worked in the Portland area a couple of years ago were probably good to within 1-2ft! It just depends.

Plus it is not easy to find documentation on how GE is handling their data. I am sure they are projecting the best data they can on the fly, but it is all about metadata isn't it!?

:good:

>
> To make a long story short, I have a client that says a "point" I staked is in the wrong place because it doesn't fall in the right place on Google Earth went he/she creates a "placemark" with said NAD83 lat/longs.

This is where we are headed. Every nimrod with a code receiver will get the lat/long of their property from the local GIS expert and then tell us our boundary resolutions are wrong. It is our job as professionals to inform those ignorant (meant in the technical definition) to the science and art of measurements and boundary resolution.

I have done quite a bit of work with Google Earth in eastern NC. Every time there was a problem between my coordinates and where they should have been on GE, the problem was my coordinates.

So far everything we have done correctly has been dead nuts on. Can everyone everywhere expect that same level of fit? Probably not.

My advice is to check your work with GE and if they fail to match, double and triple check your work before you assume GE is off.

Is it possible your client has a units problem? Trying to insert DMS as DD.ddd for example?

Larry P

In TBC you can select objects and in view select Google and those objects will plot in google, you don't need to create a file or anything, can if you want but the Google plot takes seconds in the view pull down.

Google Earth uses a custom projection based on a modified WGS84 ellipsoid.

Every project we do I export the data into a google earth kmz file in addition to a CSV file. I have not been able to notice any discrepancies in the google earth imagery and what we have seen on the ground.