Erratic elevation in routes with OSM maps

Often when planning a route in Basecamp I notice that on the Graph tab the elevation will occasionally plunge down to zero. This is not accurate as the real elevation is nowhere near zero anywhere along the route. This happens with maps from two different sources although both were created using OSM data. Both maps also have contour lines that clearly show the real elevation although I believe the graph works from a DEM model and ignores these.

One of the maps I use is the excellent OpenFietsMap, the other is a selection of tiles from garmin.openstreetmap.nl. I also use a Garmin product based on Ordnance Survey data and this does not exhibit the problem, hence I am more suspicious of the OSM map data.

Does this indicate a defective DEM model, which I am guessing is part of the OSM data? Or is it a deficiency in the Basecamp calculations?

As an example, a route from Kintbury, Berkshire to Walbury Hill, Berkshire should climb steadily from 100m to about 260m, but the route graph shows a sudden descent to sea-level half way along.

No, the OSM database does not contain elevations at all. To show those contours, most map makers use SRTM/NASA data made available by http://viewfinderpanoramas.org/dem3.html which contain those gaps.

BaseCamp calculates elevation graph based on contour lines. This shouldn’t be a problem for OSM maps which include them, but maybe there is something wrong with contours on your map? Contours aren’t standard feature for OSM maps.

Hi popej, that is very useful. I went to the trouble of recreating that route with my own maps before I read ligfietser’s previous post which seems to explain that these gaps are to be expected in the SRTM DEM data. I would be interested to know which map you are using for that screenshot. If I create the exact same route with an OpenFietsMap map, there is a descent to zero elevation between the 150m and the 175m contours.

Your map shows that the gapless DEM data is available, somewhere.

For the Openfietsmap I have used older elevation data from 2014, maybe this is improved?

I think SRTM v3 with filled voids were published in 2014. But if you have older data then any problem should be directly visible on map. Please trace contours which cross this route and check if there are any unexpected heights.

The route, from North to South, crosses a 125m, 150m, 175m, 200m, 225m, 250m and 275m contours. The problem appears between the 150m and 175m contours. The zero elevation appears to be at or very close to N51.36718° W1.46131°. Not sure if this helps.

The OFM does not render the zero m contours, but it might be possible that Garmin takes this void into calculation although you cant see it on the map. When I calculate my altitude profiles I always plan it on ridewithgps.com those are more accurate.