How accurate is the trip odometer

[+dwiener]

I went on a hike today. The odometer on my GPS showed a little about 8.6 miles. The saved track measured 11.3 miles. Mapping the course (not saved track) on a topo map tracing the trails as shown on the map measured only 7.48 miles. What would be the most accurate measurment of the distance covered on this hike?

[Bender]

I would trust the GPS unit. But that is just me. The speedometer function is dead on on mine, so why would the odometer be off??

 

Bender

 

Searching, for the lost Xanadu

[+dwiener]

But there was even the 2.5 mile discrepancy between the trip odometer reading and the mileage shown by the track log saved.

[STREETSEAL & G.I. Jane]

My Odometer on my vista is witin a tenth of a mile compared to my car's odometer. The track milage will vary depending on the number of "points" that were marked, The more points or closer they are together the more accurate, it's like connect the dots, try drawing a circle with dots, the tighter they are the more like a circle it looks. Same goes with your gps track log. They can usually be set up to record in time, or distance. Driving I set mine for distance, say every 1/4 mile it marks a "point".

While hiking I set it for Time, usually about 30 seconds, it'll stamp a point. Most GPSr's

will have a max amount of points they will record then start recording over the first set of points. Hope this sheds some light.

[+embra]

I heard it explained once thus: the GPSr odometer is pretty accurate if you are moving fast enough. But at slow speeds (like walking), the error of the position is great enough that it is difficult for the unit to accurately calculate velocity...or perhaps more to the point, the distance actually traversed since the last position was taken.

 

You would think that things get averaged out over the long haul, but apparently the upshot is that the odometer underestimates the distance traveled.

 

I find this to be so on my hikes. The tracks, on the other hand, are accurate.

 

Hmmm. As I review it, that explanation doesn't strike me as compelling. I'll see if I can dig up the original.

 

Max

Often wrong but seldom in doubt

[Kerry.]

"Discrepancys", 2.5 mile would/could/might be considered not normal? Depending.

 

I suppose some other questions might be, Which GPS?, What user configurations are allowed (and what is it set to) for things like distance, speed, trip meters, track points etc etc? How long did this hike take? How was the GPS carried on this hike? In pack, in pocket, in hand (swinging or still), The certainty tat Sat lock can be guaranteed for the "entire" hike?

 

Another thing to think about is a stationary "fixed" GPS over a period of time will in fact physically not move but the track will show distance "travelled". These days with SA discontinued that "distance" can be around the 3-5 miles per 24 hours (depending).

 

Cheers, Kerry.

 

I never get lost everybody keeps telling me where to go

[+embra]

Found it...this was a reply to someone asking the same kind of question on the Yahoo Meridian group:

 

*****

 

The problem at low velocity is that the velocity is calculated by inversing the coordinate value between epochs. What this means is that your unit takes a GPS reading, then one second later it takes another reading; the distance between the two can then be calculated by the old 3D Pythag theorem of SQRT(x^2 + y^2 + z^2). Your particular problem with the Magellan is that the accuracy is say 5-10m +/- , and you are trying to calculate two positions taken one second apart (not so far apart at 2mph) and then you throw in the positional error and you are getting bad velocities.

 

The Leica system 500 receiver has DGPS position quality around 1/2 meter and an update rate of 10hz or 10 epochs per second (as opposed to the consumer grade 1 epoch per second receiver). This provides the ability to get very accurate velocities and because of our application software is made for data collection, we can record the individual points and line segments so exact distances can be determined.

 

The error you are seeing is fairly typical of a consumer grade unit, so you may want to look into something a bit more robust for your application.

 

Max

Often wrong but seldom in doubt

[+dwiener]

Some further details on my situation. I'm using a Garmin E-Trex Vista. I have currently have saved tracks set to auto, and track log set to auto. I noticed when looking at the tracks in Mapsource, that they often hop around, especially during stops. During the course of a hike satellite reception may come and go, and I'm sure accuracy varies quite a bit. What would be the best settings for best logging accuracy under those conditions?