elevation accuracy on Bikely

abbeyboy

Does anybody know how accurate the elevation charts are on Bikely. I've saved a route which says I climbed a total of nearly 800 metres during a ride, yet it was recorded as just over 460 metres on my cycling partner's Garmin Edge 200 GPS cycle computer. Why the difference? Any ideas?
Thanks

tiredofwhiners

One uses barometric altitude and one uses GPS altitude calculations at a guess.

It also matters where you store the ride data and how its interpreted i.e. from the device data or using elevation charts as there are several sources and all have good areas and bad areas.

At a guess, its because the 800 is barometric and the 200 isn't.

rothers_bikeo

Post a link to the route so we can check it.

rolf_f

tiredofwhiners wrote:

At a guess, its because the 800 is barometric and the 200 isn't.

I think you are getting confused!

Bikely elevations are based on the digital terrain map. It will tend to give a small under-estimate but the number will always be the same for the same route (unless they change the DTM).

The 200 will be rather more random and I'd assume a lot less accurate. That said, I've not seen any serious attempt to prove the relative accuracy of different units and DTMs on here. It would be something that Bikeradar could do which would be a bit more useful than going on endlessly about how a particular wheelset really likes to climb hills!

abbeyboy

Thanks for your replies guys. Still not sure which data is the more accurate, but I'm tending towards Bikely as I assume they go by the elevation lines on the Ordnance Survey maps which are accurate.
One other thing I've noticed with the edge was that on a ride yesterday the data on the Edge 200 computer gave my ascent as 283m and my descent as 258m even though I started and finished at the same place. When I uploaded the data to my Garmin dashboard it showed my elevation gain as 201m and elevation loss at 200m; and a final figure of 200m elevation gain. Same ride, two different sets of data results. Very confusing!!

nicktheluddite

I've never really trusted the elevation on bikely very much. I don't think their background data is Ordnance Survey, I think it's google maps (although maybe that also ends up being ordnance survey).

How did you put the route into bikely? Manually entering points or a route logged by GPS?

283m vs 258m could be GPS errors? As I understand it's not always accurate to the nearest metre.

abbeyboy

Yeah, I think Bikely uses Google maps, which I thought was O.S. I input the route manually into Bikely. The route input via Garmin is GPS. Even so, 25 m out seems excessive.

briantrumpet

I think most of them use Google maps, but it then depends on the resolution of the elevation data that each site extracts from the data within Google. Ridewithgps and Bike Route Toaster both use Google data, but Ridewithgps consistently comes up with higher ascent/descent figures, as they (I assume) use a higher resolution, so pick up much more undulation. I also use the program Mapyx Quo, and they use the highest resolution of Ordnance Survey elevation data, but you can manually reduce the resolution within the program, and the effect it has on ascent/descent figures is very noticeable - much more than just a few percent.

abbeyboy

Thanks for all the comments. Much appreciated. Seems there is no definitive answer to my question as it all depends on the data used by the various mapping sites. I think the solution is to use just the one site, or take an average of both of them. However, I might just keep using the one that tells me I've climbed the most which will be good for my ego!!!

desweller

I seriously doubt the Google maps terrain data is derived from Ordnance Survey; why would OS help a competitor?

briantrumpet

DesWeller wrote:

I seriously doubt the Google maps terrain data is derived from Ordnance Survey; why would OS help a competitor?

I think the OS data is now freely available, but in any case I think the Google data is derived from satellite data - I did look the subject up once upon a time - I do remember that Google earth data is pretty accurate for certain parts of the globe.

robbo2011

I have just mapped the same 100 km ride i did a week ago on mapmyride (which I normally use) and ridewithgps. Mapmyride gives a total elevation gain of 1566m and ridewithgps gives the elevation gain as 2800m! I am shocked there can so much difference.

If you look closely at the descents on the ridewithgps elevation map it gives lots of little uphill spikes on long descents that I know there are no uphill parts on. My own scientific conclusion is that the ridewithgps elevation data is a load of bollox! (at least where I am)

briantrumpet

robbo2011 wrote:

If you look closely at the descents on the ridewithgps elevation map it gives lots of little uphill spikes on long descents that I know there are no uphill parts on. My own scientific conclusion is that the ridewithgps elevation data is a load of bollox! (at least where I am)

I noticed some funny spikes in the French Alps rides I did recently too - noticeably one where the road goes through a tunnel (Col de Rousset), but Ridewithgps takes the surface contours. It seems more accurate on GB roads though, comparing it with the figures I get from Mapyx Quo.