Request: Elevation gain (Read 820 times)

I agree ... I use ANQUET mapping in the UK and it gives the elevation gain for each run. It is interesting info but there is nowhere to log it on the run-entry screens. One for the to-do list Eric? Thanks Neil UK

If you map your runs, it is there. When Eric adds GPS, it will be there too. You can add this to the comments too.

Hi Trent ... perhaps it would be better as a 'field' in the 'run entry' then it could be searched and graphed. It would also be good to have the facility in the graphs to total by year as well (at present we can select day, week, month). Thoughts? Neil

The elevation gain from GPS devices and mapping software (ie, Google) is wildy inaccurate.

This is a widely discussed and debated topic. And depends largely on what you consider the gold standard "true" elevation. Here is how the GPS measures. When running, the GPS device triangulates its position from satellites. The position it determines is supposed to be accurate to within 1-2 meters, but we all know that is hogwash in the hills and among trees. But usually it gets a trend from point to point, so while a single point may be off, the trend of several points are usually correct (e.g., they draw a line parallel to the road 20 feet north of the road). The GPS device records a point every 1-5 seconds. The point it records includes these data, derived from triangulation: Lat, Long, Elevation, Time. These Elevation data are fairly innacurate relative to the Lat and Long. To determine elevation, it simply adds all the change between Elevation points. This tends to overestimate elevation change substantially. Motionbased.com applies a correction when you upload the data. Motionbased essentially removes the elevation data from the points and replaces it with reference data from the USGS. The reference data is far more accurate than the data derived from triangulation. Again, elevation change is calculated as the total difference between points. RunningAHEAD.com calculates elevation change by looking at points only every 1/10 mile (~ every 45-60 seconds) and comparing them to reference USGS data. This tends to underestimate elevation change if the course rolls a lot since there can be big up/downs in a 1/10 mile. When I run a local hill that starts at around 670 feet elevation and peaks at about 870 feet elevation (this course, the hill at mile 3), there is 200 feet of net gain, but there are numerous small rolls that do not show up on my RunningAHEAD.com map, do show up on Motionbased.com (also google/USGS based) that says the total climb is about 250 feet. My legs feel like 30000 feet of climb, but that is besides the point. So, what is the truth? Should every 20 foot climb is added in the elevation calculations? Well, that is more a philosophical question than a scientific one. My legs say yes.

I'm definitely aware of how GPS works and calculates elevation. I'm also aware that it's completely horrible at it, especially in very mountainous areas like where I live. My GPS said my last 50 mile run across the Grand Canyon and back had 38,500' of gain!!! In my book that would qualify as "wildly inaccurate" as the true gain is around 11,000'. During most trails runs around here (Boulder, CO), my GPS is off by a factor of 2 or more, which just isn't acceptable, hence the reason I still use a barometric altimeter.

(Sorry if that was too much detail I provided, I tend to prefer to err on not assuming knowledge) It sounds to me like the GPS device that you used was not corrected. When using the direct GPS readings, it can terribly overestimate the true elevation. When you use a correction, such as the motionbased.com gravity service (which replaces the elevation from the device with reference USGS data), it tends to be much better. Do you use motionbased? MTA: I have had numerous runs where the corrected and the uncorrected GPS readings were 2-5 times different, just as in your example.

Back to my original question: how do you know what is true?

I agree. A route that only climbs (regardless of grade) can have elevation measured using just starting and ending elevations. Have you tried to map any of the routes here on RA? Just curious as to how it does.