I spend a lot of time out on the trails in Jasper National Park with a GPS these days, so I want to know how accurate my unit is. My Garmin Etrex 30x displays an accuracy number (EPE, or Estimated Position Error), typically ranging from +-3m to +-10m. Phone-based GPS apps display a similar number.
One would think that this means that I am within 4-10m of my true location. However, this is not the case. An EPE of 4m actually means that there is a 50% chance that you are within 4m, but also a 50% chance that you are further off! However, there is a 95% chance that you are within 10m, and a 98.9% chance that you are within 10.2m.
Worse, one cannot easily compare the EPE among different GPSs, even units made by Garmin. This is because the EPE is a number that Garmin arrives at through a proprietary formula. Different manufacturers arrive at their EPE differently, and even among models made by the same company, the calculations may vary. So EPE is merely a qualitative measure of accuracy.
Interestingly, older Garmin units, like the 12 series, used to display DOP (Dilution of Position). This measurement is calculated according to a standard formula, so it can be used to compare accuracy of a location among different units. Yet it was dropped from more recent models, presumably because it is just a number (a DOP of 3 is better than a 4), so it doesn’t mean much to the average user.
Once you have recorded a trail, there is no way to compare its relative accuracy to other GPS recordings of the same trail because none of the accuracy information is stored with the waypoint or track data.
What does this mean for the average user?
1. After you turn on your GPS, allow it to record its position in one spot for a while before taking a waypoint. Accuracy tends to increase once a GPS has been on for a few minutes.
2. Keep an eye on the accuracy numbers. If they go up to for example 15m, then your track may be off by as much as 50 m. Normally this is not a big deal, but it could mean that a trail running close to a creek may appear to be on the other side of the creek. Keep a close eye on the map, and know how to read it. I always carry paper maps as backup.
3. If you use your GPS to record trails, get one of the more advanced units, like the GPSMAP 64s. They have more sensitive antennas, which work well even within buildings or under tree cover. My Etrex also performs well in the forest. Surprisingly, my Android phones, with small antennas, work nearly as well as my Garmin GPS!
But most importantly, get out there, and have fun!
9 Comments
BluePeak Blog: How accurate is a Garmin GPS? – I spend a lot of time out on the trails in Jasper National Park with … http://ow.ly/19yC0I
How accurate is a Garmin GPS? http://bit.ly/abm1sS
I spend a lot of time out on the trails in Jasper National Park with a GPS these days, so … http://reduce.li/dj62vo #accurate
BluePeak Blog: How accurate is a Garmin GPS? – I spend a lot of time out on the trails in Jasper National Park with … http://ow.ly/19yC0I
How accurate is a Garmin GPS?: Home · About · Bluepeak Home · Photo Gallery · How accurate is a Garmin GPS? I spen… http://bit.ly/aiul9Y
How accurate is a Garmin GPS?: Home · About · Bluepeak Home · Photo Gallery · How accurate is a Garmin GPS? I spen… http://bit.ly/aiul9Y
How accurate is a Garmin GPS?: Home · About · Bluepeak Home · Photo Gallery · How accurate is a Garmin GPS? I spen… http://bit.ly/aiul9Y
Hi Rogier, something I think might be true but what is your opinion? Comparing logged distances from the newer 62s and the older 60csx the newer models always give greater distance. One day O plotted out the profile from the 62s then to my best ability measured the distance covered including the slopes. Straight line distances agreed with older model whilst the measured profile distance agreed pretty closely with the 62s. So, does the newer model record the real distance travelled including the slopes? which of course is much further than “flat-map” distances when the trace is dumped on the map!
Interesting observation, Austin! I always assumed that all GPSs would measure flat distances, not incorporating elevation profiles. That would certainly make the 62s more accurate. For distanced in Jasper we use measurements taken by a wheel, as they tend to be more accurate than our GPS distances. Also because unless you set your GPS to record at a short interval (eg 0.3m or so), you get inaccuracies when trails zig-zag a lot. I will have to test out our 62s with a smaller GPS next summer on a steep trail where we know the wheeled distance. Thanks! Rogier