@tomnarock1 adding "& Planetary" seems completely reasonable to me.

Astrobiologist (part-time) here. Some (most) of our astrobiological work involves looking at terrestrial analogues for Mars and determining whether/how life could be observed there. It frequently involves a lot of geology. Exoplanets is more open -- but I think a lot of the exoplanetary modelling papers, especially for hot Jupiter big gassy planets already goes to arxiv and will continue to self-select to there.

I do not agree with the addition of planetary sciences to EarthArxiv, This would rather generalize the server. If that is the case, then we might well have to consider other areas of research for addition in future. The point is it would reduce the significance of the server title 'EarthArxiv', and unless we are willing to change the title, we should abstain from this addition.

It is true that planetary science is broad and it includes a very wide range of topics and disciplines (atmospheric, plasma, magnetospheric physics, etc.) . But papers dealing with "planetary geology", or - perhaps better - "planetary geoscience" would fit very well. Planetary geology papers sitting on ArXiv would be in fact better placed on EarthArXiv, instead...

There was a plaint a couple of days ago on Twitter (sorry, didn't keep the link) of a Planetary Scientist somewhat teed-off at always being a bridesmaid at meetings, but never getting the attention they consider their work deserves. AGU - subject distributed over multiple sessions ; Geochemists (I think - that's how I parsed the acronym) there are relevant topics, but it's splashed around over several subdivisions.
I can understand where they're coming from.
How much of an issue would it be to support such a section? And as a subsidiary, how to frame the logical boundaries for a paper to be considered relevant to both "Earth Science" and "Planetary Science"?

@aidankarley, there's not much involved in supporting "Planetary Science". The Advisory Committee decided to pursue supporting it based on the results of the poll. The only issue in doing so is framing the boundaries. To keep things consistent, we are proposing that the sub-terms of "Planetary Science" exactly match the existing sub-terms of "Earth Science", which are
Biogeochemistry,
Cosmochemistry,
Geochemistry,
Geology,
Geomorphology,
Geophysics and Seismology,
Glaciology,
Hydrology,
Mineral Physics,
Paleobiology,
Paleontology,
Sedimentology,
Soil Science,
Stratigraphy,
Tectonics and Structure,
Volcanology.
We also have in place a quick and informal manual review of all preprint submissions. Any preprints submitted outside of this scope, e.g. planetary astrophysics, would be rejected. That's the current plan we'd like to pursue with the Center for Open Science, which hosts EarthArXiv.

@All, apologies for the late reply.

From the public poll (the one I posted: https://www.loomio.org/p/e9ZV7cRq/add-planetary-science-to-the-taxonomy) it was clear that nearly everyone is in favor of adding Planetary Science, however, the poll specifically stated changing "Earth Science" to "Earth and Planetary Science" at the top level (well, top level of our domain).

I had a chat with @mattspitzer before the holiday break about this. We can request this edit, but this will not be backward compatible with any paper currently residing on the system, and it will cause some mismatch issues when searching across all *Xiv preprint servers on the main search page.

I, personally, think this is still the best option. However, as I don't really want to be the cause of a plethora of headaches for COS, my thinking is simply adding "Planetary Science" as a sub-term of "Earth Science". To me, this keeps with the spirit of the domain. I.e. a paper on Martian Tectonics is appropriate for EarthArXiv as it is applying Earth Science concepts to a planetary body. It's still a hack, but it makes more sense to me than having "Geology", "Geochemistry", etc. under "Planetary Science" as well as "Earth Science".

Futher, perhaps grouping terms like Geomorphology, Geophysics, etc. under (as sub-terms) Geology provides additional clarity.

@tomnarock1 are you able to share further insight into the splitting option you have described?

Would it be possible to keep on using the old taxonomy, but add tags for “Planetary Science” or “Earth Science”?

Having tags may anyway be nice, the current categories are very broad.

@brandon my proposed splitting option was to avoid the compatibility issues you mentioned. If we have "Earth and Planetary Sciences" while other preprint archives have just "Earth Sciences" then we will impact cross-archive searches. My personal preference is to have them distinct so that our "Earth Science" aligns with any other COS preprint system's "Earth Science". I'd rather add completely new categories than modify existing ones, which may impact future cross-archive compatibility. It also seems a bit cleaner to me to not put Planetary Science under Earth Science. For example, given a Martian geology paper, I prefer Planetary Science -> Geology over Earth Science -> Planetary Science -> Geology. Just my 2 cents. I'm happy to hear counter arguments. I don't think anything formal has been proposed to COS other than letting them know we want to add Planetary Science in some fashion.

Agree with Victor on this. Converting the current taxonomy to tags (with separate ones for Earth Science and Planetary Science) would solve many of these hierarchical issues, such as duplication of terms, and consistency between new and already submitted preprints. It would allow tagging a paper on Martian river structures with both planetary science and hydrology. Those following a hydrology tag mainly for Earth based topics would also then be able to find the paper (perhaps otherwise missed).

Ok. Apologies in advance for the length of this reply. :)

@tomnarock1, thanks for the explanation. Am I correct in thinking/remembering that the incompatibility issues (in this case it's a mismatch, or exposing terms more than once depending on hierarchy) are only evident when using the tick boxes to select domains of interest---i.e. it has no affect on free text searches?

As a point of clarification, all of the options involve "tags". The BePress taxonomy that COS use for all ArXiv variants is simply a way to standardize the terms (as tags) added to the metadata record generated when an article is uploaded to the system. So, if I add a paper describing a geostatistical model showing groundwater fluctuations in a specific aquifer, I might check the hydrology, spatial science, and statistical model terms (or tags) when uploading the paper, as they are all listed in the BePress taxonomy. These terms get added to the metadata record as tagged entities (think XML "tags" if that helps); for example:

... initial metadata bits (author name, ORCID, institution, etc.) ...

keyword: hydrology

keyword: statistical model

keyword: spatial science

... potentially more metadata ....

I interpreted @victorvenema 's suggestion as having an option to tag a paper with terms that are currently not part of the BePress taxonomy (i.e. a Folksonomy). Following my example, I might add my own tags referring to location, aquifer name, type of kriging, etc. in addition to the general terms I listed previously. Is this correct, @victorvenema, or have I completely misunderstood?

Also, for clarification, I have created a gdoc spreadsheeet with each of the proposed changes as individual tabs for comparison. Following the convention from the COS customization instructions, any added or edited cells are signified by a yellow background/fill color. The first tab, All, is the entire unedited BePress Taxonomy used by COS. The second tab, Physical Sciences and Mathematics, is only the Physical Sciences and Mathematics section of the taxonomy, also unedited (this tab is a subset for convenience only). The third tab, EarthArXiv opt1, illustrates the edits @tomnarock1 has proposed (more than once---sorry, Tom). The fourth tab, EarthArXiv opt2, is my understanding of what the taxonomy would look like via a strict interpretation of the previous poll results.

The fifth tab, EarthArXiv opt3_working, is also unedited---as of this posting. I added this tab as a workspace in case this discussion needs to continue.

@tomnarock1, your proposal seems reasonable. There are a few other bits that I think would be helpful, but it's probably better to set up a third party service to handle them for now.

Because @tomnarock1's proposal differs from the original poll, how does the community proceed? Should there be a new, separate, poll for how to proceed (it appears most think Planetary Science should be included, but there may be differences in opinion as to how)? Is this simply an issue for the advisory committee?

Just to clarify (and I think @brandon's post above is correct), COS can use a custom taxonomy for EarthArXiv. We would simply need the custom terms "mapped" into the existing one so that aggregate search continues to work as expected. The customization can be as simple as adding a single term, or using a completely unique taxonomy as the EarthArXiv community sees fit. My only suggestion is to not expect to add terms incrementally, but rather in bulk. This is easier for the code updates needed, but also for uptake and understanding by authors.
There is also the option to use author-generated "keywords" in addition to the subject terms. See http://help.osf.io/m/preprints/l/627729-share-a-preprint#A4.-Add-basic-information for more on this. These are not structured, but do show up when searching for preprints on OSF. So in Brandons example, there is where location or other metadata could be entered.

I was thinking of tags for two purposes, to distinguish between earth science and planetary science without changing the hierarchical taxonomy and to have a place where we can add more detail to make manuscripts more discoverable. The taxonomy is used by academic libraries all over the world, so there is value in not messing with it.

Tags would be similar to keywords. We could make a distinction if we wanted to where keywords would come from a list that we curate as EarthArXiv, while tags are any term the authors find useful to make it easier to find their articles.

Option 1 sounds better than option 2 to me because it provides more detail. The current version of Option 3 (no change), but keywords would also be a good option in my view. But I would say the planetary scientists are the ones who should make the choice. They know best what would work for them. I am just throwing some ideas into the air.

Sorry for my delay in responding. I lost track of this over the holidays. It sounds like we have consensus on moving ahead with planetary science terms; although there are varying implementation strategies. To me, the next step is forwarding this on to the EarthArXiv Advisory Board for approval. I will consolidate the discussions as well as the current options on the table and pass on to the Board. If anyone has additional comments you'd like included, please let me know in the next few days. My hope is to have the Advisory Board weigh in over the next few weeks and then present the new terms to COS. We have an EarthArXiv Town Hall at the upcoming EGU, which I think would be a good place to announce the new planetary science indexing terms.

Tom, I agree with the general drift of this discussion. Is it a good a idea to have a sort of connection between planetary and earth studies. For instance, (possible), studying vulcanoes on mars connected with studying vulcanoes on earth.

@hangeurdes yes, that is the intention. The recommendation we are putting forward to the Advisory Board is that the planetary terms exactly match the terms we currently have for earth studies. For example, you could tag a paper as mars volcanoes because we have volcanoes as a term for earth science papers. There would be no "mars astrophysics", for example, because we don't have astrophysics as part of the earth terms. Any term that we have for earth studies could be used in conjunction with the planets, but no other terms would be allowed. We also have human moderation of all papers submitted to EarthArXiv. The moderator performs a high-level survey of the paper to make sure it fits within the EarthArXiv scope. We're suggesting that moderators also check to make sure no planetary papers get through that don't have a connection to Earth studies.

I agree with Tom. Using tags to distinguish the subject area is a good idea for adding areas of planetary sciences to earth sciences.