Subject: RE: Map My Ride vs BT elevation profilesWhich is accurate? Probably neither. While measuring and calculating elevation at ground level would seem to be a pretty straightforward task, it's anything but. It's actually an extremely complex subject. There have been some pretty long threads in this forum and the Triathlon Talk forum regarding using consumer electronics to measure elevation changes. Some of the participants in those threads are professionals in the field who do a much better job of explaining the complexity than I could. There are three keys facets to this: - the elevation dataset being used
- what points on the map are being used
- the algorithm used to calculate differences in elevation between those points
I can't specifically comment on other sites, but I can address our implementation. BT primarily uses a web service from the US Geological Survey to get our elevation data, with another service used as a backup in case the USGS service isn't available. When you click on the map, we get the latitude and longitude for that point from the Google Maps service, send those values to the USGS service and that returns the elevation above sea level. We then compare that elevation to the previous point on the route, and if the elevation is greater we add the difference to the Total Climbing value. BTW, there are some conditions under which our maps will give erroneous total climbing values* (timing problems arising from calling Google and the USGS for two pieces of data on the same point. A simple description for a very complex problem ). This is why the "Reset Elevation" button is there. If a climbing values seems too high, click that button and it will retrieve the elevation values from the USGS and recalculate the climbing totals. This can also be used when uploading from some devices like older Garmins that report specious elevation data. * Another recent thread asked about planned upgrades to the Route Tracker. It is being totally rewritten from the ground up, in part to get rid of this problem. |