Every so often I see people discussing the accuracy of altimeters, so as I was waiting for my mate Yair to send his "yes we are still alive" Spot message (for goodness sake , the worst thing it can happen in the Aussie bush is that you die or set fire to it...) I took the opportunity to take a pic of our instruments.
(there is a Kestrel 3500 missing from the pic, the one with the due point thingo)
Anyway, as it happens during the 4 days walk my Casio, his Suunto and the two Kestrels were usually within 10 m of each other, the two GPSs were 20-50 m off.
In this pic , my GPS was still adjusting and oddly my Casio let me down. The Kestrel has the closest reading .



Franco

Franco

To sort out your varying height measurements you need to consider the following:

When were the watches calibrated to a known elevation? They do not have GPS so rely on barometric data. If calibrated at the same time and location they should read the same. If not, a difference should be expected. I have experienced a change in altitude of almost 100 metres as shown by my Suunto Vector for my camp site between evening and the following morning due to a major low pressure system rolling in.

On the GPS units what datums were set? My Garmin Foretrex does not have GDA94 (current Australian datum) so I use WGS84 (Worldwide standard for GPS satellites). This is fine for horizontal location (no conversion necessary) but creates an difference of up to 20 metres in reported elevation due to differences between GDA94 and WGS84. This is due to the use of a worldwide sea level average for WGS84 and a more localised one for the area covered by GDA94. If your GPS units include a barometric sensor then you need to understand how they integrate the barometric and GPS measurements which are highly likely to be different.

Hope this helps

An additional error factor for GPS receivers has to do with Vertical Dilution Of Precision (VDOP). Because of the way that GPS position data is calculated, the best case vertical error is 1.5 times the horizontal error. If the satellite constellation happens to be lined up poorly that minute, the vertical error might be much worse than the 1.5 times. So, if you are used to horizontal accuracy of 10 or 20 meters, then the vertical accuracy might be 15 to 50 meters. Vertical accuracy can get very bad if you are in a box canyon. You can eliminate a lot of that error by averaging over a long period. However, very few backpackers are willing to sit in one place for a long time and let their receiver average.

--B.G.--

The various instruments had not been calibrated , so yes they could all do better.
it was a bit of a "a man who has a watch knows the time, a man with two is never sure..." as in more about having some fun than anything else.
Still I am impressed with my Casio , over several trips when I compared that with the various Suunto it does very well.
And no , I am not expecting 100% accuracy , what I get is good enough for me.
BTW, the pressure had dropped considerably just a few minutes before , causing the Casio "failure".
Franco