GPS Trail Distance Measuring Tests

The problem with trying to measure trail distance in the field with civilian GPS (handheld GPS not corrected by differential GPS signals) has been that GPS has not been accurate enough to measure distance at walking speeds.

The reason for this is that GPS receivers calculate distance by measuring your position, saving it as a track point, and adding up the distance between track points. The problem comes from the fact that these track points are not perfectly accurate. For example, Garmin specifies that their receivers are accurate to 15 meters. This means that as you walk, the GPS receiver is fixing your position with up to a 15 meter error to either side of your track (off track error), then measuring the distance between these scattered points. At higher speeds, such as riding a bicycle or paddling a boat, the track points are further apart and the off track error is smaller. You can also set most GPS receivers to record track points at larger intervals of time or distance, and this decreases the errors.

My earlier tests with the older Garmin GPSMAP 76Csx showed odometer errors of up to 5% at walking speeds. Others have reported errors of 10% or more. I found errors of up to 50% while mapping trails in Nevada for a hiking guide. The errors were greater on short trails that are relatively straight.

Meanwhile, the GPS system has been becoming more accurate as newer satellites are launched. The Federal Aviation Administration's (FAA) tests have recently been showing accuracies of 3 meters or better. Also, most newer GPS receivers can receive up to 12 satellites simultaneously. While a minimum of 4 satellites is required for an accurate fix, more satellites improve accuracy and also decrease the chance of loss of satellite lock in heavy forest or areas such as canyon with a high horizon. Most newer GPS units can also use the FAA's Wide Area Augmentation System (WAAS), which improves accuracy to 1 meter or better. Finally, the latest trail receivers can receive the Russian GLONASS satellites. While this may not directly improve accuracy, it does mean the receiver has a better chance of keeping at least four satellites locked on in difficult conditions such has tree cover or high horizons.

In view of all this, I decided to repeat my tests on the Buffalo Park trail with three current model Garmin trail GPS receivers.

Test conditions:

Garmin Oregon 600 and eTrex 30 set to receive GPS and GLONASS satellites, WAAS on. Garmin GPSMap 62s set to WAAS on. (The 62s does not receive GLONASS satellites). Buffalo Park loop is 2.00 miles as measured by calibrated bike cyclometer. Low horizon (below 5 degrees) in open meadows for 90% of the loop, remaining 10% open ponderosa pine forest. All of the receivers maintained their lock while carried in belt cases.

GPS Odometer readings:

eTrex 30: 2.04 mi, error 2%
Oregon 600: 2.02 mi, error 1%
GPSMap 62s: 2.02 mi, error 1%

Track distance in Expert GPS:

eTrex 30: 2.01 mi, error 0.5%
Oregon 600: 2.01 mi, error 0.5%
GPSMap 62s: 1.99 mi, error 0.5%

Conclusions: Though it is still more accurate to measure trail distances at walking speeds by recording a track and downloading it to mapping software, the accuracy of the GPS odometer function has improved to the point where it is certainly accurate enough for measuring trail distances for casual use. As a hiking guide book author, I'll still record and download tracks as a backup and also for mapping purposes.