Went for a ride that included a ten mile stretch along the canal...

[twentysix by twentyfive]

[QUOTE 3261176, member: 259"]MapMyRide on a smartphone isn't terribly accurate on distance either. The routes are jagged to say the least. Run it against a Garmin on the same route and there are often big differences. Although you often have to wait for a few minutes for the Garmin to "acquire" the satellites before you start.[/QUOTE]
Yep - all these wot nots are subject to inaccuracies. From one set of data collected on my ride with a Garmin I got one result. Same track data in Mapmyride a second answer and again the same track data in bikehike a third. So I took my ride distance from my cateye as I calibrated it and I understand its limitations (me physicist [not software engineer] )

 

[HovR]

I find altitude on GPS is even more inaccurate than location. It's a good indication, but can't be relied on for dead accuracy. Recently I climbed Ben Nevis, and whilst at the summit my GPS claimed I was ~20ft higher than the known summit of the mountain. It was still a useful tool to know roughly how much vertical distance I had to climb, however.

 

[Markymark]

Gps has has a much greater level of accuracy for the US military than it does for the publically available signal.

 

[MontyVeda]

Which canal?

I've cycled Leeds to Liverpool and there is a massive difference in height.

The Lancaster Canal... it's famed for having no locks for 40 odd miles between Preston & Tewitfield.

 

[Globalti]

Riding along a canal your receiver's view of the required minimum of four satellites will have been inconsistent and partially blocked by trees, bridges, embankments and buildings so it's no surprise that the altitude reading is inconsistent, especially as altitude is less accurate on GPS than position.

I believe the military/civilian difference thing no longer applies. It's just that the military have much better quality receivers mounted in better locations.

 

[Cubist]

Of course many GPS units have barometric altimeters built in, which can give different results to a device plotting and recording mapping data. Some sites correct the data created by the unit to level out inaccuracies.

 

[ufkacbln]

Curious indeed; the tow path wouldn't vary by more than a couple of feet above the water. Could it just be down to aberrations in the mapping?

 

[swansonj]

I believe the military/civilian difference thing no longer applies. It's just that the military have much better quality receivers mounted in better locations.

Correct. The artificial degradation of the civilian signal was called "selective availability" and was ordered to be turned off by Bill Clinton. Better quality receivers improve accuracy in situations of poor reception. But where there is basically decent reception from enough satellites, there is no inherent difference between military and civilian. Accuracy depends then on which differential correction service or which version of something called RTK you implement - plenty of farmers subscribe to services that allow centimetre accuracy.

Sorry for displaying nerdish tendencies, I've had to learn a lot about this for work recently. Did you know farmers can use GPS to return to individual specific potatoes in a field?

 

[MontyVeda]

It's already been stated that there was nothing of this sort along my flat canal route... nice picture though.

 

[Soltydog]

I often get discrepancies with my Garmin. Just looked at my last commute, when I arrived at work, altitude was 94ft above sea level, then leaving work from the same location it was 160ft ??? So i'm guessing its not 100% accurate Gives you a rough idea i suppose

 

[MontyVeda]

I often get discrepancies with my Garmin. Just looked at my last commute, when I arrived at work, altitude was 94ft above sea level, then leaving work from the same location it was 160ft ??? So i'm guessing its not 100% accurate Gives you a rough idea i suppose

I guess I'll have to do the same stretch again and see if it records a different set of undulations.

 

[ufkacbln]

I guess I'll have to do the same stretch again and see if it records a different set of undulations.

Or try the paper map and graph paper?

Trick we used to use with cubs, piece of A4 paper and mark the contour lines along the adge, then transfer this information to graph paper and the profile will give the "accurate" heights

 

[MontyVeda]

I was thinking about that last night... in a fashion. Since the contours on my map are at 10 metre intervals, the canal could undulate by 36 foot yet remain within one pair of contour lines. Could it be the case that the altitude data from MapMyRide/smartphone is relatively accurate; the canal does indeed undulate and is fed from streams at the high points and feeds into streams at the low points?

 

[PapaZita]

Altitude from GPS is usually rather inaccurate. As mentioned above, barometric data can help. Services like MapMyRide may correct your altitude data by taking the relatively accurate latitude and longitude and looking up your altitude from a database. Essentially, they consult a contour map and replace the altitude data collected by the GPS. Canals may hit some of the interesting cases where this approach doesn't work very well.

For example, a canal achieves it's flatness using cuttings and embankments. These are normally relatively small features in the larger landscape. Depending on the quality of the altitude database, they may not feature in it at all. You may have ended up measuring the altitude of where the ground would have been had the canal not been there, e.g. where the track shows a peak, you were in a cutting.

Alternatively, assume the altitude database is perfect and captures every detail of the height of the tow path, but there is a small error of just a metre or two in your horizontal position. You may be riding along the top of an embankment, with a ditch to the side, or in a cutting, at the bottom of a steep slope. If the error in your position places you in the ditch, for example, then you'll see that reflected in your track. Similar errors can occur when riding in hills, e.g. when riding along one of a closely spaced set of contour lines.

 

[MontyVeda]

Holy thread resurrection!

got a new GPS device and did a 5 mile ride along the canal; no locks, no bridges, just a totally flat canal that can't undulate more than 12" along the stretch I cycled.

Garmin connect screenshot:

The canal stretch is between the purple markers on the elevation graph...
Average elevation is around 23m. highest is 27m and lowest is 5m (crossing the aqueduct over the river).

I suppose we can safely assume that @PapaZita above is correct and the GPS data is a few meters off... so it thinks I'm not on the towpath, but atop a cutting or fording the river.

I guess I'll have to (one day) get a device with an inbuilt altimeter, to see if i can get a flatline when riding along the Lancaster Canal.