I import a GPS .kmz into Google Earth (6.03.2197) and display the track elevation profile. The profile shows I gained 2339' in a 32 mile ride. The ride was pretty much flat with three hills. By flat I mean riding around the shores of Puget Sound about 5' above sea level - really flat!

The three hills were each one to two miles and gained 450', 250', and 350' based on elevations shown in Google Earth. That adds up to about 1050' and my GPS told me I gained 1400' which seems about right with the minor ups and downs.

MapMyRide shows an elevation gain of 950' for the same ride.

Google Earth, MapMyRide and my GPS all show a distance of 31.9 miles.

So how did Google Earth come up with 2250' gain?

Is this a known problem with GE

If a Garmin Edge 705 and a Garmin Edge 800 mounted on the same tandem can't agree (or come close) on elevation gain, I really couldn't tell you what to trust.

So - can I just make up any manly elevation gain and say that is what I did?

The problem with GPS data loggers is that they only log what the GPS tells them and for various reasons, GPS "solutions" are often off, especially in the up-down axis. Then you have your track file, which contains these errors, which are caused by the GPS computing its instantaneous position based on its distance from the various satellites it can see. Mounting location of your GPS and especially, the orientation of the antenna is critical to getting an accurate position fix. The antenna is supposed to be "right hand circular polarization" and it is supposed to be as close to omnidirectional in that direction, up, as is possible. That way, signals coming from above will be seen as the correct RHCP, and those coming from below, ie, first order (one bounce) reflections will in theory be seen as left hand polarized and so should to some extent cancel out.

Does that make any sense? The long and the short of it is that somewhere in your GPS device there is an antenna that needs to be pointed "up" and it should have as few obstructions as possible. The best place for an antenna would be on the top of abackpack facing up. If you really want to ake the GPS happy, get one of those magnetic patch antennas and a small sheet of ferrous metal at least the size of a CD or so, and mount the antenna on that.

If you go to rtklib.com then follow the link at the bottom to the authors home page, then you should find a page entitled "antenna evaluation" that evaluates a number of GPS antennas in the $50 - $5000 price range. There you can see what factors influence the antenna's quality for precise positioning. The antenna is very important.

There are cheap used GPS timing antennas on Ebay that are quite good, (removed from cell phone sites) for $27 shipped. But they look a little strange, like a little mushroom or missile nose cone. However, they will give you a nice clean signal relative to other antennas in that price range.



I think Google Earth has an option that lets you pin your track to ground level as seen by GE.

Maybe GE is simply taking every time it sees your altitude increasing and using that to iterate its counter upward, including very short stretches of up. Maybe its not subtracting the down portions from the upward ones as we mentally would do.

Computers just do what we ask them to do, in this case, its just looking at every time the altitude goes up and adding all that together.