How does the Garmin GPS62s calculate distance?

[LondonGPS62s]

I own both a Garmin GPS62s and a MemoryMap TX4. I regularly use both on my Sunday Ramblers' walks which are typically about 12 to 14 miles. The Garmin consistently reports a longer walk, probably about 5%? It is the norm on these walks that we have a couple of "comfort stops", a coffee stop and a lunch stop near a Pub where we use the Loo . . . and perhaps have a drink!

I also recently left both devices on a table in the garden for about eight hours. The Garmin created a large .GPX file detailing movement, the MemoryMap created a very small file. The MemoryMap is based on an industrial Smartphone (Defender from Toughphone) and may include an accelerometer? On the other hand, it may be slightly less sensitive and to ignore minor "movements"?

I also once used both devices on a 300 mile car journey from the Lake District to London and the reported distance covered was almost identical on both devices and pretty much agreed with the car odometer and TomTom SatNav.

Incidentally, the Ascent/Descent figures reported by the MemoryMap are quite ridiculous - the Garmin is probably not far off given changes in barometric pressure as one goes up a mountain and during the day - I often check and recalibrate at Trig Points.

My belief is that the MemoryMap is more accurate on distance and the Garmin is reasonably accurate on Ascent/Descent.

On a recent local walk,
the Garmin reported:

Points = 1,178
Distance = 21.3 km
Ascent = 239m
Descent = 236m​

the MemoryMap reported:

Points = 2,459
Distance = 20.0 km
Ascent = 582m
Descent = 595m​

What puzzles me is why the Garmin is less accurate than the MemoryMap when recording distance covered?

 

[Nuvi-Nebie]

When the units are stationary they will both 'dither', i.e. calculate that they are moving about, this is due to positioning errors, if the units have a poor 'view' of the sky this dither will be bigger, this can show up as travelled distance if the units are stationary for long periods of time, because they calculate they have moved say 100ft north then 200ft south and so on.

Some smartphones, PDAs and SatNavs get around this problem by only logging movement when it detects that the unit is travelling above a cetain speed, this works well when in a car, but can prevent any movement being logged when walking (at slow speed), so it is not used on hand held (walking / hiking) GPSs

All GPSs are less accurate at calculating vertical location rather then horizontal, this us why more expensive units have a seperate altimeter to 'help-out' the GPS measurement, my guess is that the GPS62s is more sensitive and therefore more accurate in general, but it won't ignore the 'dither' when stationary unless you turn it off during this period