SUBMITTED by Joe Dabes, Finger Lakes Trail Conference (6-Jan-2010)
Greetings,
I have been over this 8.4-mile A Loop in the Citrus Tract of the Withlacoochee S. F. here in FL so many times now that my head is spinning. However, I have gathered a very large amount of data on the accuracy of various GPS units, and here now is wheel comparison data.
Some important information before the surprising results:
Trail: Rolling sand hills on hard packed sand mainly covered with pine needles and oak leaves and some sporadic wire grass, totally dry both days, gently curving trail with ~right angle turns at four trail junctions. There were no logs or rocks on the trail that the wheel had to go around or over. No switchbacks. Trail was mowed ~5' wide; I tried to stick to the centerline. Elevation changes: low 15 feet to high 140 feet above sea level. Overhead canopy light to moderate (live oaks and long leaf pines).
Wheel: Borrowed from the Florida Trail Association: a Rolatape Professional Series Wheel, Model 415, 4.000 feet in circumference (with 4 "click pins" along it each foot); diameter 2.546 feet.
GPS: Garmin GPSmap 60CSx with Gilsson amplified external antenna mounted on top of hat. Track points set to be taken every 4 yards (12 feet, which equates to ~3 seconds at 3 mph), WAAS enabled and WAAS corrected most of the time (little D's near bottom of satellite bars visible on satellite page). Track distances obtained from bringing the .gpx track files into TopoFusion. Overlaying the tracks from the two days finds that the maximum distance between them over the ~8.4 miles is 18' (measured in MapSource).
Results: Surprising to me as I had expected the wheel distance to be a little longer than GPS track distance, but the opposite was true:
Day 1: Wheel: measured 43,800' = 8.296 miles, GPS track = 8.35 miles, wheel low by 0.65% (GPS odometer read 8.39 miles, and is updated every 1 second)
Day 2: Wheel: measured 43,785' = 8.293 miles, GPS track = 8.36 miles, wheel low by 0.99% (GPS odometer read 8.39 miles, and is updated every 1 second)
Question: Why is the GPS track distance slightly longer? One possibility is that track points bounce back and forth (right and left) as I walked. For example by bouncing 1' left of true track and then 1' right (total of 2') over a distance of 12' would cause an increase in track length of 1.4%. Another possibility is that the wheel "slid" rather than rolled some of the time.
-Java Joe
Greetings,
I have been over this 8.4-mile A Loop in the Citrus Tract of the Withlacoochee S. F. here in FL so many times now that my head is spinning. However, I have gathered a very large amount of data on the accuracy of various GPS units, and here now is wheel comparison data.
Some important information before the surprising results:
Trail: Rolling sand hills on hard packed sand mainly covered with pine needles and oak leaves and some sporadic wire grass, totally dry both days, gently curving trail with ~right angle turns at four trail junctions. There were no logs or rocks on the trail that the wheel had to go around or over. No switchbacks. Trail was mowed ~5' wide; I tried to stick to the centerline. Elevation changes: low 15 feet to high 140 feet above sea level. Overhead canopy light to moderate (live oaks and long leaf pines).
Wheel: Borrowed from the Florida Trail Association: a Rolatape Professional Series Wheel, Model 415, 4.000 feet in circumference (with 4 "click pins" along it each foot); diameter 2.546 feet.
GPS: Garmin GPSmap 60CSx with Gilsson amplified external antenna mounted on top of hat. Track points set to be taken every 4 yards (12 feet, which equates to ~3 seconds at 3 mph), WAAS enabled and WAAS corrected most of the time (little D's near bottom of satellite bars visible on satellite page). Track distances obtained from bringing the .gpx track files into TopoFusion. Overlaying the tracks from the two days finds that the maximum distance between them over the ~8.4 miles is 18' (measured in MapSource).
Results: Surprising to me as I had expected the wheel distance to be a little longer than GPS track distance, but the opposite was true:
Day 1: Wheel: measured 43,800' = 8.296 miles, GPS track = 8.35 miles, wheel low by 0.65% (GPS odometer read 8.39 miles, and is updated every 1 second)
Day 2: Wheel: measured 43,785' = 8.293 miles, GPS track = 8.36 miles, wheel low by 0.99% (GPS odometer read 8.39 miles, and is updated every 1 second)
Question: Why is the GPS track distance slightly longer? One possibility is that track points bounce back and forth (right and left) as I walked. For example by bouncing 1' left of true track and then 1' right (total of 2') over a distance of 12' would cause an increase in track length of 1.4%. Another possibility is that the wheel "slid" rather than rolled some of the time.
-Java Joe
Share:
.jpg)
0 Comments to "Mileage Measurements: GPS vs. Wheel"