Mapping oddity (Strava question!)

1 watcher
Mar 2016
5:13pm, 7 Mar 2016
481 posts
  • 0
I ran probably the flattest run I've ever done last Sunday along the Wirral Way at the top of the Wirral, following the largely concrete coastal path from Leasowe to New Brighton.

As is my wont, uploading just to Fetch doesn't suffice, so I upload also to strava (and various other sites too). strava corrects altitude data on devices without a barometric altimeter by reference (I think) to a topographic map it has stored in its backend (don't know who provides it, possibly Google, possible OS, not sure).

Anyway, partly along the route looking on strava there are these odd elevation changes, which I've converted into a segment for isolation.

Does anyone here have access to topographic data that might show how strava gets this sooooo wrong?!

Just curious; a mild Monday pm diversion!
Mar 2016
6:00pm, 7 Mar 2016
1,016 posts
  • 0
I don't have any data, but I find the strava elevation is often rubbish!

Fetch is much more accurate.

(By that I mean I have got my paper map out and counted contour lines. Fetch matches much more closely than Strava)
Mar 2016
6:19pm, 7 Mar 2016
482 posts
  • 0
Does fetch actually do anything with the elevation data, or did it just take it add recorded by the watch?
Mar 2016
6:21pm, 7 Mar 2016
1,017 posts
  • 0
No idea! Hopefully someone more technical will join the thread...
Mar 2016
6:46pm, 7 Mar 2016
1,698 posts
  • 0
K5 Gus
Yep, pretty sure Fetch will use your GPS points to go and reference some backend map database and work out the altitude gain.

If you just load your data to Fetch, and don't do a Glasses View on it ( either you, or anyone else ) then the altitude gain shows as 0, he only works it out when someone does a Glasses View. No idea why he doesn't work it out when loading the activity - sure I've asked on threads and not had an answer, although to be fair I've never sent a Feedback to ask.

I don't use strava but used to use mapmyrun, and Fetch was more accurate that mmr. So there are obviously different reference maps being used by different companies, or perhaps the same reference maps, but different sampling rates ?
Mar 2016
7:07pm, 7 Mar 2016
6,981 posts
  • 0
My guess there would be that strava is using a simple rectangular grid of elevations, and you've got a shoreline running north-east and sloping up steeply, so the track keeps switching between squares which are at sea level in Strava's head and at 25m in Strava's head. You can see there's a bit of smoothing in the slopes, but not smart enough to take care of that kind of variation. I think most of the online data come from the Shuttle SRTM map, but different companies have cleaned it up in very different ways - Google's is the best one I've played with.

And Gus is right about how FE works. I can tell because I live in Cambridge, and my Fenix 3's track shows the railway bridges, but they all vanish when Fetch processes them and there is nothing left to see on the elevation plot :)

Got something to say?

To join the discussion, sign in or join us.

About This Thread

Maintained by larkim
I ran probably the flattest run I've ever done last Sunday along the Wirral Way at the top of the W...

Related Threads

  • strava

Buy a Fetch buff

Order now for immediate dispatch!

F'Laps 2024

Join us in 2024 for a part-virtual, part-gathering endurance event to celebrate 20 years of Fetcheveryone!

Training Heatmap

See all your training overlaid on one interactive map. A great way to remember everywhere you've been

Race Listings League

Which Fetchies have done the most to make our race listings as awesome as they are?

Back To Top

Free training & racing tools for runners, cyclists, swimmers & walkers.

Fetcheveryone lets you analyse your training, find races, plot routes, chat in our forum, get advice, play games - and more! Nothing is behind a paywall, and it'll stay that way thanks to our awesome community!
Get Started
Click here to join 111,744 Fetchies!
Already a Fetchie? Sign in here