missing documentation

[eclare108213]

The variable index is empty, there is a problem with the forum link, and also need to clarify where to find the forcing variable information.

Simon Driscoll writes:

I am a PhD student in the Scale Aware Sea Ice Project ("SASIP" - https://sasip-climate.github.io). My research is focusing on machine learning and data assimilation methods for emulating sea ice thermodynamics. We are looking at the Icepack model, and I can see there is a new version: https://zenodo.org/record/5423061#.YV8Hqy8w2Jk (1.3.0) (I tried the listed CICE consortium link is https://xenforo.cgd.ucar.edu/cesm/forums/cice-consortium.146/ which does not work).

Icepack 1.3.0 is said to be a "major release update" (including "advanced snow physics") from Icepack 1.2.5. I am looking for a full list of variables that icepack calculates in its model, as well as all fields needed to force the model (i.e. fields required for the forcing data).

In the documentation for the both the latest model, and the 'master' and 'main' versions, e.g.https://cice-consortium-icepack.readthedocs.io/en/master/icepack_index.htmlthe index of primary variables and parameters is empty. The documentation for the previous version includes in section 6 includes a table of "many" of the symbols/variables used in the model:https://buildmedia.readthedocs.org/media/pdf/cice-consortium-icepack/icepack1.2.5/cice-consortium-icepack.pdf

I was wondering if there was a complete list of variables in the 1.3.0 model (i.e. that are computed), and a list of variables required for forcing the Icepack model available?

[apcraig]

I found the bug in the variable index and will create a PR to fix that. The NCAR forum seems to be down at the moment, I've sent an email to @dabail10. What else do we want to do about the Icepack forcing documentation? We have a whole section already (https://cice-consortium-icepack.readthedocs.io/en/master/science_guide/sg_boundary_forcing.html), I'm not sure we need to do more?

[eclare108213]

I doubt that any list of variables computed in the code will be "complete" because many intermediate things are computed along the way to the final solution. Icepack's diagnostic and history output is also basic/incomplete, but looking at the history output for CICE would provide a fairly comprehensive list of what's available (ignore the dynamics variables, which aren't available in Icepack). I'm not sure whether these are documented anywhere except the code itself and in the CICE diagnostic output (for history variables that are turned on), unfortunately.

[eclare108213]

In addition to https://cice-consortium-icepack.readthedocs.io/en/master/science_guide/sg_boundary_forcing.html, further information about the forcing can be gleaned from the forcing data itself, https://github.com/CICE-Consortium/Icepack/wiki/Icepack-Input-Data

[apcraig]

Another thing I notice, some of the Zenodo links point back to Icepack1.0.0. If you go here,

https://zenodo.org/record/5423061#.YV8-8qBlAqT

You can see that "supplement to" points to Icepack1.0.0. At the same time, the "GITHUB" box also points to 1.0.0. Neither of those links are quite right. They should probably point to Icepack main branch, but not totally sure. I guess formally Icepack 1.3.0 is a supplement to Icepack1.0.0 which is probably why Zenodo is setup that way. But the links are not ideal. Maybe that's OK for now.

[dabail10]

Yes, we had a power outage at the Mesa yesterday and they are still trying to recover. In fact, I was at the office for in person meetings this morning but then came home to do Zoom meetings. I'll try to take a look at this tomorrow.

[dabail10]

So, what is left to do here?

[apcraig]

I think we could close this. Did anyone get back to Simon? I see he/she posted something on the forum.

[dabail10]

Yes, we have replied on the forum now that it is working again.

[eclare108213]

The zenodo weirdness remains, but we could make that a separate issue and include some other issues with zenodo like how datasets are documented there.