CoDICE: Remove code for not valid products

[tech3371]

Emal from CoDICE team,

Hi All,

We have a new request from the CoDICE team regarding the production of some of the Lo products. 

Please do not produce the following products at all:
Lo_sw-angular
Lo_nsw-angular
Lo_nsw-species

We have determined that the use of the position variable (determined from a delay line circuit onboard) to bin counts into the sw or nsw products results in a large portion of sw counts ending up in the nsw products. Thus, the nsw products (and sw angular) are not representative of the true ion population. So, we have decided not to release these as is to the public to avoid misinterpretation of the data. 

We are working to develop a method to correct this issue on ground, with the possibility of producing a new L3 product of corrected intensities. For now, the best thing to is to simply not release those specific products. 

SDC Notes from discussion

Don't produce these products in prod:
Lo_sw-angular
Lo_nsw-angular
Lo_nsw-species

Questions:
When you say don’t produce, do you mean don’t produce at all at SDC from now?

yes

How long will that pause last?

May be indefinitely. Onboard processing to determine position is locked onboard. It’s hard coded. Due to that, data is binned incorrectly forever. Can correct on ground by using DE in L3.
Will it last until you have new algorithms to correct them?
If correction is found and finalized, it will happen in L3.
For your public release, we can definitely look into solution to hide it from public release.
Stop producing and then clean up and not release at all.

[jtniehof]

Per discussion with Margaret, we'll move IMAP-Science-Operations-Center/imap_L3_processing#53 up in priority to help compensate. It won't help fix these products, but will provide some spatial information.

[tech3371]

@jtniehof Thank you for quick comment on this!

Note to us, @lacoak21. Looks like they will be using some CoDICE L2 code to new L3 development. Let's check and tag Jon before we remove code.

[jtniehof]

Thanks for noticing the code reuse! How confident do we feel in the validation of the L2 code given that the inputs are bad? The two paths that seem to make sense to me:

1. Menlo copies the relevant code out of L2, incorporates it into the L3 repository, and maintains it there. Code is then deleted from L2. This would cause problems if the FSW is ever fixed and we wanted to restore the L2 products.
2. The code remains in L2 and is called from L3. In validating L3, we make any fixes as best we can to the L2 code, but if we're not routinely running that, it might break things if, again, we ever want to restore the L2 products.

We can also defer the decision until Menlo starts on IMAP-Science-Operations-Center/imap_L3_processing#53 and can report out how difficult it would be to reuse the code in-place.

[tech3371]

Both makes sense to me. I am curious to hear what Menlo think.