remove stems we know are missing

[teixeirak]

@jess-shue , @ValentineHerr ,

It would help the census team if we remove stems that we know are missing: status WR (wire removed), DT (dead tag found), for DN for two (5-yr) censuses in a row. We might also consider removing stems found as DN in mortality census for a few years in a row.

We had some discussion as to whether this should be applied on a stem or tree basis. @Iriskennedy , @rmh24 , @CarolineTroy , @krystalbagnaschi , @emacmonigle , please weigh in.

[jess-shue]

@teixeirak, @ValentineHerr

As I am not working from Python code, this would be a heavy lift, and in my opinion not a good idea at this point in the census. The current labels already show the previous codes (DN, DS, etc.). I may be able to change the symbology (color/shape etc.) of the points, but otherwise I would basically have to re-build the app to make this happen.

I understand that working in the field with 2018 data when there is newer information from the mortality censuses (and without dead stems removed) can be frustrating - I have done it myself - but initially we had kept dead stems in the data in case of previous errors.

This is something we should have discussed further during setup. Please let me know how seriously this is needed, but I would basically end up building a new app.

[teixeirak]

Sorry @jess-shue , I didn't realize what a heavy lift this would be. Please hold off.

@ValentineHerr , @Iriskennedy , @rmh24 , @CarolineTroy , @krystalbagnaschi , @emacmonigle , would there be a way to make this easier via GitHub, either by automatically marking those stems in the QA/QC phase or coming up with a list of stems meeting those criteria? Both would be clunky in their own way, but perhaps still helpful?

[ValentineHerr]

how would the product of this be used?

[teixeirak]

I'd envision that perhaps a simple list of stems meeting these criteria (separate from app) might be useful, but I'm not sure. @Iriskennedy , @rmh24 , @CarolineTroy , @krystalbagnaschi , @emacmonigle , please comment.

[teixeirak]

@Iriskennedy , @rmh24 , @CarolineTroy , @krystalbagnaschi , @emacmonigle , would it be helpful to get a list of these stems and then mark as not found ahead of censusing each quadrat? This could be done from the lab and save time in the field.

[jess-shue]

@teixeirak As I mentioned above, I may be able to do something with the symbology, or how the points appear as well if the list of stems were created. If Valentine could make the list, I'd be willing to attempt something on the appearance side, but I can't promise it will work. Completely removing them from the data would require the full re-build.

[teixeirak]

Thanks, @jess-shue! I think that altering the symbology could be very helpful., and a better solution than removing them altogether. @Iriskennedy , @rmh24 , @CarolineTroy , @krystalbagnaschi , @emacmonigle , please let us know ASAP if this sounds like a good solution to you.

[teixeirak]

@ValentineHerr , since you'll be out soon, let's go ahead and make the lists (assuming you can do this quickly, which I think you can). Let's make lists of:

• trees marked DN for two main (5-yr) censuses in a row
• trees marked DN for two mortality censuses in a row, and marked DS or DC at some point before that.

The first category is higher priority if you only have time for that.

[ValentineHerr]

@jess-shue Here is the list of 838 tags and stemtags that were marked DN at census 2 and 3: DNfor2mainCensus.csv

@teixeirak and @jess-shue, a quick look at mortality censuses showed no cases of trees marked "DN" two censuses in a row, so the second category would be empty.

[jess-shue]

@ValentineHerr Thank you for putting that together!

@Iriskennedy @emacmonigle @rmh24 @CarolineTroy @krystalbagnaschi Realizing it is only 838 stems, is it worth putting another layer together for you? I will gladly do it, but I want to make sure this is truly helpful before working through things. Thank you!

[krystalbagnaschi]

Thank you very much @ValentineHerr! Is there any way we could get a list of those marked with the code TR (tag removed) as well? Only if it's easy and you have time of course.

@jess-shue the list should be plenty - we're thinking that this is work that could be done remotely on rainy days etc so the color of the points changing to finished should be enough of a distinction on the map when we get to the field. Thank you!

[jess-shue]

@krystalbagnaschi @Iriskennedy @emacmonigle @rmh24 @CarolineTroy @krystalbagnaschi and @ValentineHerr

I added a file of stems with code 'TR' as TRstems.csv.

I wasn't sure how Valentine linked to the DN file above, so sorry if adding it to the raw_data folder isn't the best option.