Differing elevation readings (RidewithGPS & ELEMNT Bolt)

ajdobbin

Basically this is a question of which elevation readings are likely to be more accurate.

I plan my routes with RidewithGPS and use an ELEMNT Bolt.

As a couple of examples, I planned a 72mile ride and RWGPS stated the elevation gain would be 5674ft. After riding this the Bolt put it at 7100.

Also a 43 mile ride with RWGPS stating 2891ft but the Bolt putting it at 3822ft. Most of my rides vary by differing amounts.

Should I be concerned that the Bolt is not working properly or is the issue more likely to be with RWGPS? I know it’s not the be all and end all but it would be nice to know roughly what I’m actually climbing!

sungod

tools to calculate climb based on mapping data can only be as good as the data they use, if different datasets are overlayed (i.e. layering roads onto elevation data) the effects of misalignment on total error can be significant, i've even seen the derived road gradient show a climb instead of descent

on any non-trivial route, i'd treat anything a route planner says about total climb with great suspicion

some bike computers let you choose whether to use gps or barometric sensor, i think the bolt has barometric, if you can configure which is used, i'd choose barometric and assume it's the more accurate

reasons...

gps altitude data can be way out, if conditions are good expect 10-20m vertical accuracy, but when things aren't so good don't be surprised at errors of >100m, twisty roads on the side of steep terrain are not good conditions for gps

the error isn't constant, it could be low one minute and high the next, the measure of cumulative climb will also depend on how the raw data are smoothed

unless calibrated to a known start point altitude, barometric devices won't reach the absolute altitude accuracy of good gps, but for cumulative climb their inherent smoothing and the typically slow rate at which weather changes atmospheric pressure may mean they outperform a gps for the purposes of determining total ascent when cycling

in general they'll tend to slightly under-report total ascent as sharp reversals in gradient at, say, a short summit, get smoothed, though it depends how fast you go through these areas

ajdobbin

sungod wrote:

tools to calculate climb based on mapping data can only be as good as the data they use, if different datasets are overlayed (i.e. layering roads onto elevation data) the effects of misalignment on total error can be significant, i've even seen the derived road gradient show a climb instead of descent

on any non-trivial route, i'd treat anything a route planner says about total climb with great suspicion

some bike computers let you choose whether to use gps or barometric sensor, i think the bolt has barometric, if you can configure which is used, i'd choose barometric and assume it's the more accurate

reasons...

gps altitude data can be way out, if conditions are good expect 10-20m vertical accuracy, but when things aren't so good don't be surprised at errors of >100m, twisty roads on the side of steep terrain are not good conditions for gps

the error isn't constant, it could be low one minute and high the next, the measure of cumulative climb will also depend on how the raw data are smoothed

unless calibrated to a known start point altitude, barometric devices won't reach the absolute altitude accuracy of good gps, but for cumulative climb their inherent smoothing and the typically slow rate at which weather changes atmospheric pressure may mean they outperform a gps for the purposes of determining total ascent when cycling

in general they'll tend to slightly under-report total ascent as sharp reversals in gradient at, say, a short summit, get smoothed, though it depends how fast you go through these areas

Thanks for such an in-depth reply. I actually had no idea about the barometric sensor being an option, I just assumed it was all done through GPS. Makes sense though.

davebradswmb

Wahoo explain how the Elemnt units calculate elevation in this article.

Naturally they claim that their system will give the most accurate results, and on the face of it they could be right. I too have noticed greater elevation with my Bolt than Ride with GPS, and also compared to Strava and my mate's Garmin 520, but I haven't bothered investigating any further to come to any conclusions. Audax UK use contour counting on OS maps when calculating elevation on their rides, I was rather hoping someone would have done a comparison with this method to save me the trouble of doing it myself.