[MHMarks]

Why is it that on almost all my runs I start and end in the same place, yet my elevation loss is ALWAYS greater than my gain. This shouldn't be possible. I'm talking an average of 15 feet. Could it be me dropping my arms on descents more than on climbs?

[wsjinames]

The elevation inaccuracy of Garmins is well documented, especially the non-barometric ones. There are a variety of technical reasons, but it IS interesting that it's always the same way.

Have you tried uploading to Garmin Connect and letting it apply the elevation correction to see what you get?

Wm

[MHMarks]

WS, I have uploaded, and sometimes I get a positive elevation gain, sometimes a negative net. But it is hardly ever the same. Even though I start/end in the same place. Perhaps there is some discrepancy on accuracy of even google earth

[Darren]

For reasons WAY beyond my mathematical comprehension, it is significantly more difficult to pinpoint (triangulate) elevation than it is location.

I _never_ trust Garmin data for elevation, especially when worn on the wrist.

The altitude correction capabilities are all based on known points and obviously the accuracy varies based on how many data points are used in the evaluation. If you ran down into a gully that isn't included in the map data, it won't be reflected in the correction.

I know there was a firmware update for the ForeRunner 305 that had to do with various elevation calculations (including one when paused) and satellite acquisition issues. Make sure you're on the latest GPS chipset and unit software.

[adam]

I use a .65-.7 multiplyer for my garmin elevation based on a few known routes. The hiller I use the .7 or even up to .75. The flatter I have to go down to .65. Done a couple checks on this and it is at least that much better. Makes a difference of 1000ft or more sometimes. Hill intervals with the lap counter on a steady grade actually seem to be fairly accurate. As you crest or start is where it gets wavy.

[wsjinames]

If you upload Garmin data to connect.garmin.com, you can use their elevation correction, and it comes out a lot closer to net change of zero. Their default is to apply it for Forerunner data and not for Edge data, because the latter have barometric detectors that are supposed to be more accurate. However, if you are out long enough and the air pressure changes as the weather does, you can see some pretty wild stuff on and Edge as well!

WKO+ is supposed to have correction data as well, but I don't know how to get it.

The correction data is just accurate elevations known for "every" ground point, so it ignores your elevation data.

I use Ascent, and it has an adjustable elevation filter, and with the default settings, I get pretty good data. it's just throwing away elevation changes that are smaller than a certain amount, basically

[MHMarks]

I have the elevation correction applied. IDK what the deal is. I think I get solid data, but its just off a bit

[wsjinames]

Sorry I'm not more helpful!