Discussion: GPX Upload & Climb

I was wondering how climb is calculated from uploaded GPX files -- wIth the usual GPS noise, this tends to suffer from scaling problems, and I would like to know how much trust to put into that number.
I've just uploaded my first track, and the result is a bit higher than what I got from counting, contours, but a lot lower than what QLandkarte (probably using a quite naive algorithm) gave me.

Attackpoint is replacing your GPX's elevation values with ground-survey height data before displaying the profile or computing the ascent. See here for some previous discussion.

We do also use a few smoothing and threshold methods in the ascent calculation to try to eliminate noise and errors. It gives reasonable results in most cases, though can be on the high or low side depending on circumstances.

Thanks, Ken. That sounds like a good way to deal with the problem. I'll take a look at the profile and compare it to my map...

on a related note: I have been using a Polar M400 for about a couple of months now. A week ago went for a hike in the Foothills. Polar said the elevation gain was 870m (with a descent of 785m, even though we started and stopped at the same place), but when I exported as a TCX file and uploaded into AP, it shows an elevation gain of 1091m which agrees much more closely with the topo map. So I guess what you are saying above explains this?

Here is an interesting blog post about Strava elevation calculation versus GPS elevation and barometric elevation. DCRainmaker mentioned it in his Week in Review for May 17th.

http://regex.info/blog/2015-05-09/2568