Calibrations at very low temperatures

[isgiddy]

For very low expected temperatures (-1.5-2), do we need to run this in a freezer room, or would it be acceptable to calibrate at higher temperatures? The SOP only recommend using an AC to change the temperature for example. What about sub-zero temperatures?

@patricialg @tomhull @hahn-johannes

[patricialg]

Hi Isabelle, Good question! I haven't worked in such low temperatures but, for other experiments, I normally use a thermostatic bath which can keep the sample at those low temp. The only problem I see here is that you normally do this 2-points calibration with distilled water and it will freeze below 0C, not like SW, but then if you use SW you will need to set up the sensor to the salinity of your sample, so:

• should we use SW and correct salinity? We will need to know the salinity of our sample.
• should we prepare a known salinity solution with NaCl to do this experiment?
• should we use distilled water and temperatures higher that 0C? Don't know what people think? Cheers, Patry

[soerenthomsen]

I am not an expert for very low temperatures but have carried out the two point 0/100% calibration many times before and after deployments.

Knowing that your calibration has to occur in the next days, I suggest to definitly try to get one pair (0/100) at normal room temperature and two more pairs (0/100) at very low temperature (i.e. something near 2 degrees) close to your range but above 0 with zero salinity to be able to later follow the calibration procedure developed by @hahn-johannes.

You always need at least two temperatures to span the temperature / O2 space.

You can read about it in his thesis here https://oceanrep.geomar.de/22943/1/Dissertation_JHahn.pdf Section 3.4 and for example Fig. 3.3

[tomhull]

Echoing Patry and Soeren, my knowledge of low temperature work is zero so I can't help here.

[soerenthomsen]

Hi @isgiddy did you already recover your glider, which you deployed at very low temperatures? Any experience to share?

[hahn-johannes]

should we use SW and correct salinity? We will need to know the salinity of our sample.:

• We never did this and I have no experience. I expect, when measuring the salinity (e.g. with a Multimeter) with a fair accuracy, this could be an option. Just to give some numbers: if you have e.g. S=35 PSU, T=0°C and you measure salinity with an uncertainty of 1 PSU (you are probably much better), you arrive with an uncertainty for the saturation concentration (100%) of about 3 µmol / L

should we prepare a known salinity solution with NaCl to do this experiment?:

• I have no idea, how accurate this is, but would not recommend it (however, I am not chemist). The salinity correction after Garcia and Gordon (1992) considers seawater. Probably, NaCl behaves different as you have different ions.

  should we use distilled water and temperatures higher that 0C?:

• I agree to that and what @soerenthomsen has recommended.

As a side note. The calibration at 0% is unaffected by salinity (application of salinity correction does not change an oxygen concentration of 0 µmol/L), i.e. you could use seawater to get temperatures lower than 0°C without freezing.

[soerenthomsen]

Thanks @hahn-johannes especially the last point is interesting in suggesting to use seawater to get temperatures lower than 0ºC without freezing

@isgiddy did you give this a try? Any chance to test this during/after recovery?

Would be great to prepare a small paragraph on very low temperatures during 2022 for version 2.0 i.e. after your @isgiddy glider is recovered from the Southern Ocean after this very long deployment.

If there are others who have experience at very low temperatures this would be great! @bastienqueste ?