Originally Posted by
DaveLeeNC
Hmm, that is very interesting (re: route data). That does change my perspective somewhat. Do you have any idea what the interval is on a RideWithGPS 'route'? Thanks.
I do most of my riding in Moore, Richmond, Montgomery, and Hoke counties. The highest elevation that I have ever seen is 750' and the lowest is 210'. Those 1000' climbs are a bit tough to find :-)
dave
ps. True flatlands would actually make it easier to simulate a long climb, IMHO. In that case a steady, low RPM effort is doable. It is the descent side of rolling hills (like around here) that makes that hard. Around here, pretty much no matter where you ride it is about 1000' of gain every 25 miles.
Calculating grades
Basically, short steep pitches can be inaccurate. The overall climb is usually correct.
Locally, ridewithgps often misses the short, steep pitches on a climb. The overall elevation gain and average grade is quite accurate. I've seen climbs with "8% max" that are really 12% or more for a short section.
On the other hand, gps recordings can be quite incorrect too. The grade depends on road distance and elevation. And recorded data can be messy and noisy. All recording sites use different smoothing methods. I often see climbs online that go 8%-11%-7%-9%-10% where the road is actually a consistent grade.
The most accurate grades are from gps recordings using barometers for altitude and a rear wheel sensor for distance.