# How Accurate Are Google Earth Measurements?

I use Google Earth quite a bit, but I have yet to test their accuracy. Has anyone made a comparison of a known distance to the Google Earth ruled distances? The pic shows a yellow ruller line and the window shows the distance between the two points.

I only have experience double checking it around the office here, and it's been fairly accurate, within about 2-3 feet, generally. Obviously 2-3' can make quite a difference when measuring a 20' field of view. Generally though I've used it for longer distances, 100+ feet, up to about 700-800, so it's not a drastic difference.

It is generally within 5% of actual for me. I've notice that some of the image sets are slightly undersized according to the scale marker they display in the lower left. (It shows 10', but measures something like 9'-10".) Other image tiles do not seem to have this problem. (different satellite sweep?)

That's my experience as well. It's good enough for doing rough layouts, estimating things to the nearest yard, etc.

I wouldn't go and order exactly 518.25 feet of pre-terminated cable based just on a Google Maps estimate.

Remember that the Earth isn't really flat. Measuring the distance between two points on a sphere is different than on a plane. Probably for the IPVM crowd, this is not an important distinction. But it can make a difference in the outdoor radio link business.

Google earth images are not really from satellites. They are airplane images that are stretched as required to get them to match up. If you "fly" around in a city with 3D images of buildings like San Francisco you will see that sometimes you're viewing with a perspective from the South and sometimes from the North. Given the change in perspective, you can't expect the flat earth approx to be extremely accurate.

I've checked it against various football stadium fields at varying altitudes and I've found it to be within a few feet each time. That being said, I agree with some comments above that it should only be used for rough estimates. And to confirm that I'm terrible at eyeballing distances!

Google earth images are not really from satellites.

Some are.

Most land areas are covered in satellite imagery with a resolution of about 15 m per pixel. This base imagery is 30 m multispectral Landsat which is pansharpened with the 15 m [panchromatic] Landsat imagery. However, Google is actively replacing this base imagery with 2.5 m SPOTImage imagery and several higher resolution datasets mentioned below. Some population centers are also covered by aircraft imagery (orthophotography) with several pixels per meter. Wikipedia - Google Earth

Point taken though, sometimes its hard to know which is satellite and which is not. I've noticed that often images don't change for years, so you can start to rely on them. Til they change...

Thanks. I also checked it against some known dims and found it to be reasonably accurate. But like Undisclosed said, i would not base a terminated cable length on it.

I 'm also working with google earth in outdoor projects and some 2D/3D software to simulate camera positions and look at the correct focals. When I can't have a on site telemeter measurement, (or ask the customer to provide manually on segment) I usually take a car to calibrate the map. A small car = 5 meters , and most of the time the calibration is correct

In my software when I place a 3D car , it should match the 2D shape on the park place...