My customers with smartphones sometimes dispute my height differences at the end of a hike. So I answer that if smartphones are precise in positioning, the altitude measurement is less precise and the measurement noise makes the cumulative height difference measurements sometimes very random.
I take advantage of this period of confinement to show on a simple case why smartphones are imprecise. This is a digestive walk in the center of Toulouse where my estimate of the cumulative difference in height is of the order of +/- 20 to +/- 30 meters maximum. But the measurement of my smartphone gives: +146 meters / -143 meters.
In fact, if we look at the figure below, we see that the altitude measurement oscillates by several meters whereas being on the path I can can see that it is flat and this measurement noise does not correspond at all to reality. But the smartphone or the software (BaseCamp) used to analyse the GPS track does not understand it and mathematically accumulates (let’s say the word stupidly …) the variations in height differences until making an error of a factor of 5 on such a small distance.
Fortunately, it is possible to correct the measurement noise and better assess the height difference. By uploading your track to the IBP Index of the FFRP, a very good estimate of the difference in height is made thanks to an algorithm whose secret is well kept … For the moment, I have not yet found fault!