Moderated Discussion Areas
ContinuousWave: Small Boat Electrical
GOOGLE EARTH Accuracy Compared to GPS: Miles Off
|Author||Topic: GOOGLE EARTH Accuracy Compared to GPS: Miles Off|
posted 03-29-2010 04:57 PM ET (US)
While visiting a particular spot in my car I saved the location (N33 40.598 W78 57.824) from my GPS. When I got home I put a waypoint on Google map at this coordinate. When I look at the map the waypoint is miles from the actual location. The actual spot on the Google map appears to be N33 31.2499 W79 03.4234. I double checked my actual coordinate and I'm sure if I used my GPS it would take me back to the actual spot. Why the discrepency?
posted 03-29-2010 05:04 PM ET (US)
You can submit this problem to google on the google maps page. I have done this many times as related to addresses.
posted 03-29-2010 05:35 PM ET (US)
Is it possible that there is a difference in the way the information is being displayed/entered?
There are two different ways to show Lat/Lon coordinates: degrees, minutes, seconds, or in decimal form. Most GPS units use degrees, minutes, seconds. I believe Google Maps uses decimal form.
posted 03-29-2010 05:38 PM ET (US)
My Raymarine chartplotter displays the Lat/Lon coordinates for the vessel's location in the top right, UNLESS I move the pap to another location on the map. Then it displays the pap's location. I have often forgotten about this "feature" when attempting to plot my location on a paper chart.
posted 03-29-2010 05:46 PM ET (US)
It worked correctly for me.
I just copied N33 40.598, W78 57.824 and pasted it into the "fly to" and voila!
Please note that a put a comma between the 8 and W.
posted 03-29-2010 06:09 PM ET (US)
Oh yeah, go to "tools" , "options", 3D view, click on "degrees, decimal minutes", "apply".
posted 03-29-2010 08:34 PM ET (US)
Also guys - there are more than one "coordinate" systems that a GPS can be referenced to - but I forgot the details - maybe this will jiggle someone's mind - and I will go back and refresh my mind a bit. --- Jerry/Idaho
posted 03-29-2010 11:07 PM ET (US)
What you are referring to Jerry, is called the Datum.
Google Maps/Earth, like every other cartography product is not perfect.
posted 03-30-2010 08:44 AM ET (US)
There is no way on earth, and if you think about it, that is precisely what I mean, that there is a difference of several miles in the map datum being used by a GPS receiver and the map datum used by Google Earth for a location in the eastern coast of the United States United States. Look for some other cause for this discrepancy.
posted 03-30-2010 09:36 AM ET (US)
Actually Jim there could be,though I doubt it's the solution in this instance. I've seen points plot several miles apart that were recorded in NAD 27 state plane and plotted mistakenly in NAD 83 state plane.
Granted - State Plane is a projected coordinate system and not Geographic coordinate system like WGS-84 typically reported by GPS units but, my point is never say never....
Most GPS units have a huge catalog of datum's that could mistakenly be selected by the user. Usually - the numbers reported by the unit will be so different from what the user is used to seeing that the mistake will quickly be discovered but not always. when I get some time, I'll try to find an example to illustrate.
posted 03-30-2010 09:40 AM ET (US)
I suspect the problem in this instance is a discrepancy in how the reported numbers are being entered ie. degrees minutes seconds, degrees minutes, decimial minutes, or decimal degrees.
I always use decimal degrees. Makes recording and entry much easier. I've noticed most boaters use decimal minutes. I think that's tied to how charts were used prior to the emergence of GPS.
posted 03-30-2010 09:45 AM ET (US)
Most likely cause is degrees:minutes:seconds vs degrees:decimal minutes vs decimal degrees.
Google Like most modern mapping packages uses WGS84 datum. Even if your GPS was set for NAD27 datum it would only be off by tens of meters not several miles like your seeing.
posted 03-30-2010 09:48 AM ET (US)
And, since Google is only taking satellite photography and stitching it together and overlaying corrdinates on top of it there is all kinds of room for error...
posted 03-30-2010 09:48 AM ET (US)
I forgot about state plane(instead of Lat,Long), this could cause several miles of discrepancy.
Google is always going to be WGS84 with Lat and Long.
posted 03-30-2010 09:49 AM ET (US)
This is also why Google solicits corrections from users.
posted 03-30-2010 04:04 PM ET (US)
Google Maps won't be that far off, and it's way too far to
be a reasonable datum problem.
And the change in degrees from W78 to W79 says it's not the
posted 03-30-2010 05:11 PM ET (US)
Interestingly enough, I never suggested this was a Datum problem, I was only answering Jerry's question.
I don't expect google earth to be any more accurate than any other electronic cartography, and I certainly expect it contain errors.
My Garmin Nuvi is updated regularly and it often shows me "driving" where there are no roads even with a good view of the sky, three or more satellites in view and augmented by WAAS.
My Raymarine chartplotter often shows me "boating" where there is no water, or in water I know to be un-navigable. (I am primarily a river boater, so this is a bit more understandable.) This too is with a WAAS augmentation.
It's my position that google earth, or google maps are certainly not to be considered a "gold standard" for accuracy when it comes to applying a particular Datum to their satellite imagery.
They buy (or license, or whatever) their imagery from different sources and stitch it all together to make up their product.
I would expect more accuracy from Garmin/Navetch or Jeppesen's C-Map but I don't always get it.
posted 03-30-2010 09:37 PM ET (US)
The difference in position is described as "miles." That means at least two miles, and perhaps more. It does not sound reasonable to me that the east coast of the United States as seen via GOOGLE EARTH is two miles or more from its actual position. You might see that with some remote south Pacific Ocean island, but not in the coastal region of the United States.
I have GPS receiver position data for my living room. When I compare it with the GOOGLE MAP position I see a variance of about five feet.
posted 03-31-2010 12:53 PM ET (US)
I put the same waypoint position (N33 40.598 W78 57.824)in Garmin's HomePort and MapSource and end up at the same position as Google Earth which is ~11 miles from the actual location?
Is it something I'm doing incorrectly?
posted 03-31-2010 01:05 PM ET (US)
N33 40.598 W78 57.824 is an industrial location.
N33 31.2499 W79 03.4234 is a location on a river.
When you saved the location while in your car you might have moved the cursor to the river location before you saved it?
posted 03-31-2010 03:20 PM ET (US)
WT My Garmin Nuvi doesn't work that way. I simply SAVE 'Current Location'. I can't believe the ACTUAL location is wrong but that possibility always exists. Someday when I get to that location again I'll double check it.
posted 03-31-2010 05:26 PM ET (US)
If you are actually suggesting that GOOGLE EARTH is off by 11-miles based on one anecdotal observation with an inexpensive GPS receiver, you need to examine the probabilities of these two alternative explanations:
--Case One: You are correct, and the entire global population of GOOGLE EARTH and GPS users have never noticed this massive 11-mile offset before; or,
--Case Two: You are incorrect and your anecdotal data is in error.
posted 03-31-2010 09:57 PM ET (US)
I’m somewhat embarrassed! Puzzled by the discrepancy I went back and started playing with my auto GPS and on “My Locations” I noticed 2 locations that had almost exactly the same mileage from my house. Being suspicious I looked at the locations of both and they were the same.
I found that what I had thought was the ‘actual’ location was, in fact, a duplicate of another location I had previously stored under a different name. I have no idea how I did it but it appears I’m the cause of the problem.
My apologies for the confusion.
|Tom W Clark||
posted 04-01-2010 09:37 AM ET (US)
Ah yes, often the answer to really puzzling questions turns out to be something very simple.
Purchase our Licensed Version- which adds many more features!
© Infopop Corporation (formerly Madrona Park, Inc.), 1998 - 2000.