 |
Bikely originally depicted the route as having 3,600' ascent. However, though I was tired, I would have placed it at "about 3,000." John reported about 2,700' on his barometric-based bike computer.
Looking closer at the profile:
there are three spurts of precipitous dips and climbs that I don't "remember." One theory is that there is no topological data for that point and the Bikely/Google software calculating the ascent assumes a value of "zero." This would be easily testable if speed was recorded.
A factor I've observed on other rides like Tour de Blast -- and wouldn't be much of a factor here -- are some topological mapping software products do not handle cumulative ascent when bridges are involved. For example, consider a trip between points A and B on the diagram below. A human would see the cumulative ascent is 500' -- the rider descends down 500' to the bridge, crosses the chasm, and climbs 500' to point B. The program might rely on the geographical representation, erroneously reporting 1,000' of elevation gain, as if the bridge wasn't there.
Another source of error occurs when using instruments whose precision is comparable to the error. For example, when I did STP a few years ago, I came up with an accumulated elevation gain of around 4,000' feet using the simple barometric altimeter in my heart rate monitor. Those who have ridden STP know it's the flattest 200 miles one can do in this half of Washington, and 4k seems about right. However, there were other riders arguing passionately that their GPSes reported 6,000 - 10,000 in cumulative ascent. A 10,000' gain on STP is unbelievably high.
Just for kicks, I exported my barometric data into a spreadsheet. When I added up the raw ascents, I got 5,435'. As the Polar HRM only reported changes of 5 feet, it tempers the error somewhat. Polar wouldn't tell me what their smoothing algorithm was, though they claimed that their tests show it estimates to within 5%.
For comparison, during the first 12 hours of the first day of that same STP, my Garmin eTrex recorded 4,400' of ascent before its batteries went kaput. Unfortunately, I didn't retrieve the original data for comparison. However, I did observe that while the GPS is more precise, the error accumulation (at least on my eTrex) was pretty scary. For example, while I stood in line for the Sani-Can on a flat parking lot, the GPS recorded a climb of 20', taken in one foot chunks. Extrapolate that over 200 miles, and it's easy to see where that 10,000' number might have come from.
Anyone done a serious analysis of the elevation gains?
.