Sojourner,
Here's the reason I ask.
I was a beta tester for Mick Chawner who wrote the program for the BackUpFRunner, a feature pack logbook for the Garmin Forerunner. One of the features he added was total feet climbed (Ascended) and total feet descended. This was calculated from the GPS trackpoint log.
In evaluating the data produced, we discovered that the difference in these totals from the same information generated from the identical route (trackpoint log) using topographic mapping software was almost always in disagreement by at least 25% and more commonly 30-45%. On generally flat routes the discrepancies were even greater dispelling the presumption that perhaps tree cover and/or steep narrow routes were affecting satellite reception.
One of the specfic rides I analyzed was the first day of The Columbus Fall Challenge which was slightly less than 100 miles from Lancaster, OH to Marietta, OH. This was not the most extreme example but a good example of uneven terrain. Over the route, there was a general downward slope or drop to the Ohio River Valley at Marietta of about 450 feet.
Topographic software measured the feet ascended as 8,293.6 feet, whereas the Forerunner measured the feet ascended as 6083.9 feet, a divergence of 26.6%. This ride was broken up into 4 legs whose divergence range between -19.2% to 44.0%
Topographic software measured the feet descended as 8,754.4 feet, whereas the Forerunner measured the feet descended as 6,315.8 feet, a divergence of 27.9%. The divergence over the four legs ranged between -20.2% to -33.3%.
However, the linear distance travel is measured with a consistent accuracy that varied between +0.08% to +0.88% and was +0.44% for the entire ride.
What was discovered by comparison and analysis was that not only was the Foreunner consistently divergent in its readings vs topograpic mapping software regarding elevation changes, it was also similarly inconsistent in relative measurements of elevation data.
The conclusion drawn from these comparisons and analysises over a number of rides over flat, rolling and hilly terrain, was that the Forerunner's measurements of elevation data should be taken with a large grain of salt. If one desires relatively accurate information regarding elevation changes over a GPS trackpoint route, especially one logged by the Forerunner, one is far better off measuring this data using topographic software rather than the Forerunner's GPS trackpoint record.
This is what I termed as "significantly inaccurate." This is also the reason I raised my first question, how is the elevation calculated? These discrepancies are not the result of points being "off route" as one can determine from viewing the route overlayed on a road map. The trackpoints are consistently and almost without exception dead on the road. Your assurance that the algorythm used discounts any erroneous GPS trackpoints does little to reassure me in light of the data and analysis I have personally done. If anything, it leads me to believe that we have been bamboozled by MB and their assertion that their calculations are accurate because their calculations drop out GPS trackpoints that are decidedly off route. My personal experience, testing, data and analysis tends to demonstrate very clearly this is not the case with a GPS trackpoint record.
What is your reference that reassures you that the elevation data you have accumulated during your rides is accurate? The MB assurance regarding their algorythm and their calculation of this data? I contend it is merely relative and consistently inaccurate as my full data record has demonstrated to my satisfaction and the satisfaction of the developer I was working with.
If you have ever noticed on most handheld GPS units that offer altimeter information, they use barometric altimeters in the GPS unit to perform these calculations rather than GPS information. This is the case for even the WAAS enabled handheld GPS units. The reason for this is due to the significant (gross) inaccuracy of GPS records of this data.
Your data chart is impresive looking, however, as accurate or as inaccurate as its data may be.
Just my two cents, FWIW.
Ben Cline
Better to aspire to Greatness and fail, than to not challenge one's self at all, and succeed.
Here's the reason I ask.
I was a beta tester for Mick Chawner who wrote the program for the BackUpFRunner, a feature pack logbook for the Garmin Forerunner. One of the features he added was total feet climbed (Ascended) and total feet descended. This was calculated from the GPS trackpoint log.
In evaluating the data produced, we discovered that the difference in these totals from the same information generated from the identical route (trackpoint log) using topographic mapping software was almost always in disagreement by at least 25% and more commonly 30-45%. On generally flat routes the discrepancies were even greater dispelling the presumption that perhaps tree cover and/or steep narrow routes were affecting satellite reception.
One of the specfic rides I analyzed was the first day of The Columbus Fall Challenge which was slightly less than 100 miles from Lancaster, OH to Marietta, OH. This was not the most extreme example but a good example of uneven terrain. Over the route, there was a general downward slope or drop to the Ohio River Valley at Marietta of about 450 feet.
Topographic software measured the feet ascended as 8,293.6 feet, whereas the Forerunner measured the feet ascended as 6083.9 feet, a divergence of 26.6%. This ride was broken up into 4 legs whose divergence range between -19.2% to 44.0%
Topographic software measured the feet descended as 8,754.4 feet, whereas the Forerunner measured the feet descended as 6,315.8 feet, a divergence of 27.9%. The divergence over the four legs ranged between -20.2% to -33.3%.
However, the linear distance travel is measured with a consistent accuracy that varied between +0.08% to +0.88% and was +0.44% for the entire ride.
What was discovered by comparison and analysis was that not only was the Foreunner consistently divergent in its readings vs topograpic mapping software regarding elevation changes, it was also similarly inconsistent in relative measurements of elevation data.
The conclusion drawn from these comparisons and analysises over a number of rides over flat, rolling and hilly terrain, was that the Forerunner's measurements of elevation data should be taken with a large grain of salt. If one desires relatively accurate information regarding elevation changes over a GPS trackpoint route, especially one logged by the Forerunner, one is far better off measuring this data using topographic software rather than the Forerunner's GPS trackpoint record.
This is what I termed as "significantly inaccurate." This is also the reason I raised my first question, how is the elevation calculated? These discrepancies are not the result of points being "off route" as one can determine from viewing the route overlayed on a road map. The trackpoints are consistently and almost without exception dead on the road. Your assurance that the algorythm used discounts any erroneous GPS trackpoints does little to reassure me in light of the data and analysis I have personally done. If anything, it leads me to believe that we have been bamboozled by MB and their assertion that their calculations are accurate because their calculations drop out GPS trackpoints that are decidedly off route. My personal experience, testing, data and analysis tends to demonstrate very clearly this is not the case with a GPS trackpoint record.
What is your reference that reassures you that the elevation data you have accumulated during your rides is accurate? The MB assurance regarding their algorythm and their calculation of this data? I contend it is merely relative and consistently inaccurate as my full data record has demonstrated to my satisfaction and the satisfaction of the developer I was working with.
If you have ever noticed on most handheld GPS units that offer altimeter information, they use barometric altimeters in the GPS unit to perform these calculations rather than GPS information. This is the case for even the WAAS enabled handheld GPS units. The reason for this is due to the significant (gross) inaccuracy of GPS records of this data.
Your data chart is impresive looking, however, as accurate or as inaccurate as its data may be.
Just my two cents, FWIW.
Ben Cline
Better to aspire to Greatness and fail, than to not challenge one's self at all, and succeed.
Last edited by:
Wants2rideFast: Jan 28, 05 13:11