3 ms·

SRTM data is nearly useless for that purpose. I had a run at this exact application a few years ago and it was quite an exercise in frustration. The resolutio

SRTM data is nearly useless for that purpose. I had a run at this exact application a few years ago and it was quite an exercise in frustration. The resolution of the data is adequate, but there are artifacts related to the way the data was collected and processed. There are discontinuous edges throughout the data. Also, of course, the data is not the elevation along a route. Bridges and other things over or under which one might ride a bicycle must be accounted for externally. Even roads in trenches or cuts need special attention. I think the only way to make such an application work is to combine the elevation data of numerous sources, SRTM being just one. The most valuable source would be the ambient pressure sensors in mobile phones, which are next to useless for absolute altitude but are excellent for relative altitude as you move along a path, and of course they have GPS so you also know the path taken.

carlob 12y ago
This is really really interesting. I was curious as to why I hadn't seen any execution of this pretty obvious idea, now I know.
- thrownaway2424 12y ago
  Not that I'm trying to discourage anyone from solving this problem. That was merely my experience. There probably exists some combination of filtering and cleverness that would get around the discontinuities. But I think you can see from Google's incredibly vague elevation-along-route bike directions that this problem is not easy.
- maxerickson 12y ago
  http://cycle.travel/ http://cycle.travel/ and https://www.komoot.de/ https://www.komoot.de/ both take elevation into account (and rely on OSM data for routing). I guess they are probably supplementing the SRTM data where possible though (but they are almost certainly falling back to it in places).
  - thrownaway2424 12y ago
    cycle.travel/map is really nice. I remember there was (years ago) a nifty cycling map of the city of Atlanta, which included a unique triangular control that allowed the user to trade off climbing vs. distance vs. safety (on-road or off, bike lanes, etc). It was based on Google tiles and routes but had its own elevation data.

carlob 12y ago

This is really really interesting. I was curious as to why I hadn't seen any execution of this pretty obvious idea, now I know.

thrownaway2424 12y ago
Not that I'm trying to discourage anyone from solving this problem. That was merely my experience. There probably exists some combination of filtering and cleverness that would get around the discontinuities. But I think you can see from Google's incredibly vague elevation-along-route bike directions that this problem is not easy.
maxerickson 12y ago
http://cycle.travel/ http://cycle.travel/ and https://www.komoot.de/ https://www.komoot.de/ both take elevation into account (and rely on OSM data for routing). I guess they are probably supplementing the SRTM data where possible though (but they are almost certainly falling back to it in places).
- thrownaway2424 12y ago
  cycle.travel/map is really nice. I remember there was (years ago) a nifty cycling map of the city of Atlanta, which included a unique triangular control that allowed the user to trade off climbing vs. distance vs. safety (on-road or off, bike lanes, etc). It was based on Google tiles and routes but had its own elevation data.