Related
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Why is Total Ascent of a course planned in Garmin Connect so low compared to actual?

lytleric
over 1 year ago

When planning a course in Garmin Connect, I pay close attention to the Total Ascent so I have an idea of how much climbing I'll have to do.  However, it is always significantly less than what the actual ascent ends up being (as recorded by my Edge 830).  Has anyone else experienced this? 

For my most recent ride, I created a course that Connect shows as having 1,920 FT of total ascent.  After riding that course, my Edge 830 recorded 3,458 FT of total ascent.  That's almost double.  This is consistent for almost every course I ride.

I have "Elevation Corrections" disabled, but turning that on only corrects the ascent total to what "data from professional surveys" calculate it should be, which in this case was 3,106 FT.  So much closer to what my Edge recorded using its barometric altimeter.  Why doesn't the course planner use the same "data from professional surveys" to come up with the total ascent for the course?

  • 0 over 1 year ago

    The difference comes from the number of track points. The more track points you have (the finer resolution), the bigger the the total elevation gain will be. The effect is called the Coastline paradox - Wikipedia. And since the planned route usually comes with many many fewer trackpoints, than the resulting track of the actual activity, the higher resulting elevation gain at the final activity is to be expected.

  • 0 over 1 year ago in reply to trux

    Interesting - that would have never occurred to me, especially since Garmin Connect plots out the exact roads of my course, rather than just leaving the course as track points in "connect-the-dots" fashion.

    I suppose it makes sense that that is the cause of the inaccuracy, kind of like why a wheel sensor racks up more accurate distance than a GPS odometer.  Still frustrating nonetheless.

  • 0 over 1 year ago

    This is weird. Where do you have elevation correction for courses planned in Connect? I only have this option for recorded activities. Connect should use the survey data in the first place, when you plan your course there, therefore there should be no point in correcting the altitude. 

  • 0 over 1 year ago in reply to #Matthias#

    Elevation correction shows up only on the uploaded activity, not on the course.  But when I follow the course during my ride, the total ascent for that activity (as recorded by my Edge) ends up being significant greater than the Connect said it would be when I planned the course.  If I disable elevation corrections on the activity, it adjusts the ascent total, but it's still way more than what the course calculated it would be.

  • 0 over 1 year ago in reply to lytleric

    Ah okay, I misunderstood your last paragraph. Would you share the gpx file of your planned course and of your recorded activity? I would be interested to take a look

  • 0 over 1 year ago in reply to lytleric

    The Course has 858 track points, while the activity has 2017 track points (more than double). As I explained, the higher total elevation gain is mainly the result of the finer resolution at the activity due to the effect called "Coastline Paradox".

    Sometimes there may be also some additional artefacts increasing the total elevation gain, caused for example by the wind, or by atmospheric pressure changes. Though, from the elevation profiles of the activity vs. the course, it is hard to guess in this case. However, since you wrote the corrected elevation profile was close to the recorded one, these artefacts are likely negligible here.

  • 0 over 1 year ago in reply to lytleric

    Okay thanks. I think that's just very difficult terrain for elevation gain measures, be it prediction or recording. There are many many ups and downs in your course and elevation data for prediction is not perfect. Because of the many ups and downs, the inaccuracies of every up and down is accumulating to very inaccurate value in the end. It seems like connect is applying a different smoothing in the planner and in the elevation correction and that's why it shows such a great difference. Looking at the elevation graph of your planned course in comparison to the corrected one, the smoothing is definitely too aggressive in the connect route planner. Small ups and downs are just "ironed out". Every app does the smoothing a little bit different. I did an elevation correction of your course in Locus map and I got about 2800ft of elevation gain. 

    I don't think that you can improve the prediction accuracy if connect. I would try other route planners and check how the prediction compares to the real result there.