Aug 26, 2019

Calculated vs Measured distances

Hi there,

Using first principle trigonometry (Haversine formula), I calculate the location of a second point based on an initial Lat-Long.

the calculated distance between these two points appears to be different than the map distance measured using GE tool.

The discrepancy seems to vary as I move to other locations and is within 15m.

This has just recently appeared . It used to be spot on!

Is there any image processing calculations at the back end of Google Earth that is causing this discrepancy?

Thanks!

Details

Troubleshoot an issue,- Google Earth Pro (Desktop),Windows

Locked

Informational notification.

This question is locked and replying has been disabled.

Community content may not be verified or up-to-date. Learn more.

All Replies (4)

Alchemist251

Diamond Product Expert

Aug 26, 2019

Hi,

The satellite imagery is placed by an algorithm and may not be accurate. That imagery is taken from different angles and altitudes and that may cause errors too. You can often see the placement errors where roads go from tile to tile.

You can't use plane geometry over any more than short distances without getting an error. The Ruler tools use spherical geometry and great circles.

There are probably three or four more reasons why you might be 15 meters off.

Please tell me what your intended use is for the program and we maybe able to give you some suggestions.

js zerbini

Original Poster

Aug 26, 2019

Thanks for your prompt answer!

The haversine formula is calculating the great-circle distance between 2 points.

That's why I'm concerned.

What other reasons could explain this behavior according to you?

User Rocket Bob

Sep 24, 2019

I took a walking tour of Washington DC that had measured distances and then also created a path on GE Pro. The GE Pro distance was longer by a significant amount. I then decided to measure the distances between the Washington Monument and the Capitol rotunda as well as the distance between the rotunda and the Lincoln Memorial. These distances are well known and documented. In both cases the distances were close to 15% off and at least 10% off if one accounted for round off errors of the published distances. This seems like a lot more error than would be acceptable and I believe more than used to be the case for GE. I guess if the error is consistently 10 -15% off I could just reduce the GE supplied value by that much but that seems a bit brutish to me. Just for completeness, I'm running on an old Vista operating system. Are there other precise and well known distance I could try? Also, does anyone know if the error is there for longer distances?

User Rocket Bob

Sep 24, 2019

An update to my post just a bit ago. I decided to try measuring between Washington and London and Washington and Boston. The differences been GE and the known air mile distance were only a fraction of a percent. It seems odd to me that the longer measured distances appear to be more accurate.