Temperature vs. depth

[Linus Torvalds]

On Wed, Nov 21, 2018 at 6:43 AM Lutz Vieweg <lvml at 5t9.de> wrote:
>
> They are slow responding in air, but under water, it usually takes just a few seconds
> for them to change.

That's definitely not true.

I looked at my dive profiles, and I see things like "it took five
minutes for the temperature to stabilize".

This obviously depends on the dive computer and some will be much
better than others, but I really don't think the "temperature by
depth" can work in general. You're much better off looking at the time
progression, and just eye-balling how it matches with depth.

For example, looking at my dives in Bonaire, there was a clear
thermocline at around 105 ft on one dive, but almost all the other
dives were pretty much "constant water temperature", so I can clearly
look at just that dive.

And I had four dive computers with me on that dive:

 - the Suunto EON Core tracks depth pretty closely, and with a delay
of only about one minute. The profile looks sane and the temperature
profile matches is well with a 0.1°C resolution, just delayed a bit.
For nice slow descent/ascent, the one-minute delay is probably not
even all that noticeable.

 - the Garmin Descent Mk1 only gives you 1°C resolution, but like the
EON Core, it seems to be fairly quick to react (again, looking at the
profile, I'd say "roughly one minute" delay). But the 1C resolution
means that it's fairly noisy.

 - The Aqualung i770R  gives 1°F resolution,m but the delay looks like
it's about three minutes.

 - The Shearwater Perdix AI similarly gives 1°F resolution with a
three-minute delay.

The reason I say that the delay can be up to five minutes is that I
look at the Shearwater profiles, and on the dives where water
termperature was pretty much constant, it quite commonly takes about
five minutes for the sensor to stabilize from the air temperature.

Looking around at other dive trips:

 - Maui has almost no temperature gradients by recreational depths

 - Okinawa had temperature gradients, but the biggest ones seem to be
only indirectly about depth. The shore dives show a big temperature
gradient in shallow waters, but it looks like it's less about
"shallow" and more about "close to shore".

 - some of my dives show temperature fluctuations that seem to have
nothing to do with either surface or depth, and seem to be more about
currents etc

Basically, on the data I have, I don't really see anything that
indicates that doing some depth-to-temperature mapping makes much
sense. The thermoclines obviously *do* exist, but when they are clear
they are quite visible from the existing time-based model, and most of
the time they aren't clear at all and certainly not easy to associate
with depth because of delays and other noise.

                  Linus