Why big and sudden GPS track shifts?

[Amicus]

After hiking a few miles this morning around Mass. Audubon's Moose Hill Sanctuary in Sharon, I took a ten-minute break in the Visitor Center. I looked at the hike stats page on my Garmin GPSmap 60CSx at the start of that break. Glancing at that page again as I left the building, I was startled to see that I had gained about 1.8 phantom miles on my Trip Odom. Moreover, my Moving Avg had zoomed from 2.6 to 3.8 mph and I had now attained a Max Speed of 22.1 mph (very fast, even for me).

Scrolling to my map page, I found the explanation - two straight lines radiating from the Center at about 20 degrees (.25 mile) and 55 degrees (.4 mile), and a tangle of shorter lines right around the building. I have heard about GPS drift, but I have recorded 20 tracks like this for long hikes since I got my 60CSx and not experienced anything remotely like this (nor were there further gyrations over the long balance of today's hike).

What might have caused this? Something electronic in the building? The DoD fiddling with the satellite signals? Did this happen to you? (Time was a few minutes after 10 am, Eastern.)

 

[DougPaul]

Scrolling to my map page, I found the explanation - two straight lines radiating from the Center at about 20 degrees (.25 mile) and 55 degrees (.4 mile), and a tangle of shorter lines right around the building.

This is a typical recorded track when the signals are marginal. In this case, it was poor reception inside the building.

What might have caused this?

See above.

Something electronic in the building?

Possible, but not likely.

The DoD fiddling with the satellite signals?

No need for conspiracy theories... Too much depends on GPS for the DoD to risk fiddling with the signals except in a war zone.

Did this happen to you? (Time was a few minutes after 10 am, Eastern.)

I've seen it before. My GPS was off ~10am today.


Occam's Razor: when there are multiple possible explanations, the simplest one is most likely to be correct.

Doug

 

[Amicus]

Occam's Razor: when there are multiple possible explanations, the simplest one is most likely to be correct.

Doug

Thank you for pointing out what, in hindsight, seems obvious (no razor I, on this one anyway). I have long subscribed to Occam's parsimonious principle, but in this case am sorry to have to give up my theory of abduction by indecisive aliens, with powers to hypnotize and turn back the clock. (One other VFTT'r PM'd me the same possible solution, so maybe this one time, Occam's razor is dull?)

 

[NeoAkela]

I have the very same unit, and like you have never had much "drift" troubles for the three years I have hiked with it, from anywhere from the Great Gulf to Mahoosuc Notch. The 60csx has exceptional reception... However, I once left it on as I put my pack away in the basement of my home and when I checked the track, it was all over the place! Like DougPaul said, the poor signal inside a building (man made metals, concrete, etc) can cause the track to leap all around as it tries to figure out where you are.

I've found this to a much lesser extent when taking an extended break on the trail in a sheltered area... the track tends to create a star-shaped pincushion where I am sitting, but the spikes are short on distance. I make sure to filter the track in MapSource first and then manually remove any remaining wayward track points before saving at the end of a trip.

The only problem I have had with my Garmin is with the trip odometer - I've learned to ignore it and go with the track distance instead, as the odometer will occasionally over- or under-report the miles traveled. The track distance always seems to match the book, so I stick with that and turn the odometer off. I've had a few times the odometer telling me I had gone only a mile, when I turn the corner on a 3-mile trail and find the end. The track shows the 3 miles correctly, the odometer shows 1.2?

 

[Amicus]

The only problem I have had with my Garmin is with the trip odometer -

I too have been very happy with the reception of my 60CSx and haven't lost a track, although I learned today that it will make it up if it has to. In my experience, the Trip Odom. distance always differs from what Nat Geo gives me for the track, almost always exceeding the Nat Geo distance, but never by much. Today, typically, my Trip Odom (phantom miles included) said 16.54 m., while the Nat Geo track read 16.42 m.

 

[DougPaul]

The only problem I have had with my Garmin is with the trip odometer - I've learned to ignore it and go with the track distance instead, as the odometer will occasionally over- or under-report the miles traveled.

The GPS has only its recent track to decide whether you are moving or stationary (with random location drift).

Somewhere there is a speed threshold--if you are moving slower than this threshold, it assumes you are not moving (for the odometer). This is an easy decision in a car, but can be hard if your are hiking/walking slowly. IIRC, I reduced it on my 60CSx from something like 2 mph to 1 mph. However, I just wasted some time looking for where to set it and I couldn't find it... (It is also possible that I am remembering from a different model.)

Doug

 

[DougPaul]

I too have been very happy with the reception of my 60CSx and haven't lost a track, although I learned today that it will make it up if it has to.

It isn't making the track up. It is just trying to work under very poor signal conditions. Figuring out what to do under such conditions (declare a loss of signal, extrapolate for a short distance, or just present the potentially bad data point) has no unique solution. There is also a significant amount of smoothing used in the position and velocity calculations. Different manufacturers, for instance, extrapolate for different periods after a loss-of-signal leading to the illusion that some GPSes are more sensitive than others when they are just outputting less reliable data.

This is more obvious with less sensitive GPSes.

BTW, a low EPE (estimated position error or accuacy) does not guarantee that the error is less than or equal to the given number.

Doug

 

[RickB.]

(declare a loss of signal, extrapolate for a short distance, or just present the potentially bad data point)
Doug

So it may do all these things over the course of a route with an intermittent signal?
When I'm riding with my 76CSx on my handlebars, I can see real time when it loses the signal because the speedometer reading drops below my true speed. That is never evident in reviewing the recorded track points afterward.

 

[DougPaul]

So it may do all these things over the course of a route with an intermittent signal?

Yes.

When I'm riding with my 76CSx on my handlebars, I can see real time when it loses the signal because the speedometer reading drops below my true speed. That is never evident in reviewing the recorded track points afterward.

No.

Let's assume you are riding in an otherwise open field and pass by a tree which momentarily blocks the signal from one of the 4 satellites that you are using. The GPS will use the past data plus the current data from the 3 satellites to make an educated guess at your location and velocity. It will not indicate that anything is amiss unless such a condition persists for some period of time. When you are walking along in a forest (or traveling past multiple obstacles on your bike or car) individual satellites will disappear for a moment and reappear. Any practical GPS has to be designed so that it can continue to work accurately under such conditions.

The best indication of what is happening to the individual satellite signals is shown on the satellite page. (Changes may be too fast to be shown accurately.)

Doug