Elavation Accuracy??

[paddison]

Hi Folks

Can anyone comment on this question I received recently...

I uploaded the route direct to Strava from Maprun6 and got 200ft more climb than my friend I ran with. She used a Garmin watch then uploaded to Strava. Do we know how accurate MapRun6 is?

I have seen the following on the Strava FAQs...

Why is my activity's elevation different than my friend's?
Even though you may have done the exact same activity as your friend, each GPS device, even if you’re using the same GPS device, will record its own unique set of data. This GPS data will always be different because of the inherent variables involved in recording GPS data: recording interval (the time between GPS points), signal strength, GPS hardware, etc. The differences in the GPS data may cause differences in the calculated elevation data.

Anything to add??

Thanks

Paul

[MichaelRaz]

I'm going to babble on here so be warned:  Few things on the various sport watch forums that causes more distress and confusion (beside plain old GPS accuracy) than elevation data.  This is not a Maprun6 factor rather related to to the device used to provide the GPS data, the way the GPS is then utilized to obtain elevation data (or not) and how the various apps such a strava then interpret the data.  It's easy to assume it should be easy but it is surprisingly complex and difficult.  Two primary ways to obtain the initial elevation data: Directly from the GPS data or, when available, a barometric altimeter (based on changes in air pressure).  The first is generally not especially accurate as vertical GPS accuracy will be, even in good conditions, up to or more than twice the error of horizontal (has to do with the math).  Given a good GPS signal may provide 5m horizontal (on a good day - realistically less), you can see how bad the vertical data can be. How is this data then used?  If directly from the GPS data, elevation calculations will not be accurate due to the normal error and, given the per second collection rate, will certainly not catch the shorter ups and downs on trail.  The primary way this is corrected is by using the available GPS horizontal location data and, after the fact, looking up the location against a elevation database (DEMS data).  The accuracy of this will depend on the initial GPS accuracy, the DEMS database being used, and the available resolution of the DEMS data for the location.  You can see how running along the side of a cliff, the GPS may show you a few meters over the edge, thus the "corrected" data will be way off! In addition, really high resolution DEMS data is not available for all areas (or not used), so consider 5 or 10m "pixels" and you can see how limited it may be.  Most sites (Garmin - I assume Strava but don't know as I don't use it), offer this type of correction (called Elev Correction on the Garmin Connect site).  

The second process using the barometric altimeter is the tool of choice for most higher end sport watches.  This will provide a much more accurate collection of the ups and downs on a trail so will normally be the default when available (on Garmin Connect, Elev Correction will be disabled by default for any watch that has a barometric altimeter).  Lots of issues with BA however, least of which is that changes in weather (even over a short period of time) will cause air pressure changes that the watch needs to try to determine if it is caused by elevation change or general air pressure changes. In addition, the sensor needs to be calibrated (generally prior to a run) if the user wants accurate absolute elevation rather than just the overall changes over the run - i.e. I don't care if I start at 100m altitude and go up 10, I may not care that the watch shows 250-260 as it still gives me the 10m change.  Calibration is handled differently depending on watch and options - most common is to us the initial GPS to set the start and sometimes check to make sure weather has not really messed up things.  Some newer watches will store location altitude data based on DEMS and use this - they get more fancy as new releases come out.
Bottom line, run the same route two days in a row with the same device and you will not get the same elevation change.  It is an art as well as a science.  Some will generally be higher than other device types, some less, some all over the place.

[paddison]

Michael

Thank you for taking the time to explain this complex issue. Really appreciated.

Paul