Page 1 of 1
[closed] Elevation data
Posted: Wed Dec 11, 2013 9:53 pm
by Guest
Hi, I use Alpine Quest for road cycling. It works very well but we have found an issue with the Height Gain/Loss values displayed on Localisation Tracker display. The values seem too low.
For example a recent route gave a gain of 593 and a loss of 600m. We knew it was more than that so I imported the track into Memory Map and that gave an overall climb of 1073m.
I then copied the track gpx into MS Excel and summed all the +ve elevation changes - amazingly it gave the answer 1074m - almost a perfect match for MM.
Can you explain this discrepancy?
Re: Elevation data
Posted: Thu Dec 12, 2013 9:44 am
by Psyberia-Support
Hi,
This is due to the fact that AlpineQuest smooths tracks before computing stats.
We believe (like a lot of people) that the gain should not take into account small increases/noises/errors of the elevation data.
In practice, when you sum all positive differences of the elevation profile, like you did with excel, the computed gain is highly over-estimated. If the elevation increase of +1m, then does -1m, then +1m, etc, the gain should be computed as flat (i.e. +0m gain), instead of +2m which is only due to the error of the GPS.
To compute a rough estimation of the "real" gain of your track, just sum the elevation gains between all summits. The real gain will be something between this value, and the one obtained by summing exactly all elevations (excel method).
If you consider that the smoothing is too high in AlpineQuest and the gain under-estimated, you can change it. However, the default settings have been extensively tested.
In the track details window, scroll to the section "Statistics settings", click on "Edit", then:
- set a smoothing of 1 (location);
- set an elevation threshold of 1m.
- select OK. The gain (and other data) will be updated.
Let me know if something is not clear.
Re: Elevation data
Posted: Thu Dec 12, 2013 6:33 pm
by Guest
Thanks for the response. It looks like we have not been climbing as much as we thought
The only thing that doesn't make sense is why Memory Map estimates a climb that is much higher.
We use MM to plot a proposed route then save it as gpx to use in Alpine Quest on the phone.
I then record the actual track in Alpine Quest (we usually manage to miss at least one turn) and I can then import that into MM. The MM climb for both the original route and the imported track are similar.
Re: Elevation data
Posted: Fri Dec 13, 2013 9:30 am
by Psyberia-Support
Hi again,
If you computed the same gain as MM did using excel, it means that it just sums all positive elevation differences without applying any error correction.
Concerning how much you climbed, it just a point of view. When starting at 1000ft and only going upward to 2000ft, you have a theoretical gain of +1000, but when you sum positive elevation differences recorded by a GPS, you get something much higher (as explained before due to GPS error but also small local decreases of the path), and both measures make sense and are arguable...
Edit: there is plenty of online visualization tools applying different method of error correction that your can test. For example
http://www.visugpx.com/ (in french) uses a default elevation threshold of 10m and no smoothing (but you can set a smoothing). You should have a look at the result it gives you.