support d170 data processing

[twollnik]

Ich als Gastwissenschaftlerin (Sonja) möchte das Interface nutzen, um d17O Daten zu prozessieren und den bestmöglichen Weg dafür erarbeiten. (Anmerkung: d17O-Daten werden nicht von allen Picarro-Geräten gemessen.)

[twollnik]

Open question: What is d170 data? How should d170 data be processed?

[thomas-muench]

Preliminary remark:

There are different isotopic compositions in water which you can measure:

1. the ratio of oxygen-18 to oxygen-16 isotopes (denoted as d18O)
2. the ratio of hydrogen-2 (deuterium) to hydrogen-1 isotopes (denoted as d2H or dD).

These two are the most common ones, measured by all devices, and are most import for paleoclimatic research. There is

3. the ratio of oxygen-17 to oxygen-16 isotopes - the mentioned d17O value.

This is a relatively new proxy and can only be measured with newer devices, such as the Picarro L2140i.

[thomas-muench]

Regarding the processing of the data:

For d18O and dD, methods are available to correct (process) the data, which are basically implemented in piccr and will be used here. It is yet unclear if they can be applied in the same way to the d17O data, which is what Sonja wants to investigate.

Since we have L2140i devices at AWI which can measure d17O, we have to take this into account when reading in data from these devices. However, the development and implementation of the d17O processing should take place in piccr. But I would not see this as a critical part for the project in the first place and would therefore push this issue to later.

[twollnik]

@thomas-muench todo

[thomas-muench]

I now received feedback from Alexandra and Maria on the d17O issue.

The main point: Measurement files from a Picarro device run in "17O mode" will not cause any problems in our current implementation, since these files will simply contain four additional columns:

d(17_16)Mean, d(17_16)_SD, E17_Mean and E17_SD;

all other columns are the same as for standard Picarro files.

This means:

1. Normal processing of d18O and dD is straighforwardly possible, we simply ignore the 17O columns for this.
2. Processing of d17O in the same way as d18O and dD would be possible simply by doing this in piccr in exactly the same way as for the standard isotopes. However, this probably will result in bad 17O data, since the optimal processing of these data is still a matter of debate. But it would be a good starting point to study the optimal way.
3. Including d17O in Cpt.-Picarr would need to
   • include a true value column in the processing table,
   • maybe include new processing options,
   • include plots for analysing the d17O (and E17) output.

Point 2. would be doable right away, but is of minor importance for the time being. Point 3. is needed when we do 2., or latest when we have an improved d17O version of 2.

So in summary, this entire issue seems not feasible to do in the limited time we currently have, but should be done, when we have time left.

Therefore, I leave the information here for later reference, but close this issue for now.

[twollnik]

@thomas-muench Thank you for the detailed follow-up.