rethinking inclusion of CP

[teixeirak]

@mcgregorian1 (and @ValentineHerr , would be great to get your thoughts too),

I came across a comment by R3 that I previously hadn't seen as a big deal: "L165-167: I think this needs to be explained more clearly. Why are type I errors unlikely and type II error possible? I also think there are real issue with the analysis of crown position. Presumably on average the crown illumination index of trees would have increased over the past 50 or so years (from suppressed to dominant) as trees grew and neighbouring trees died. So I’m not convinced you can include this in your models going back half a century. It’s just not sensible to assume that crown position would not have changed over such a long time period."

This makes me wonder if we should drop CP from the statistical models. I've been reluctant to do so because I really like the concept of testing it, but in the end we're not able to disentangle height and canopy position, which means it ends up just complicating the paper without making any progress.

@mcgregorian1 , would you be okay with dropping it from the models? Of course we still have it in fig. 3, so we can still discuss.

[teixeirak]

We could reformulate hypotheses as follows:

Then, move Fig. 3 up to being Fig. 1.

[teixeirak]

THere's also this comment from R1: "Line 266-267: why only in 1966? I think this shows that this variable is not properly assessed. The older the date, the more accurate the estimation of crown position and height were."

[mcgregorian1]

Yeah to be frank I was really confused by this last comment by R1 - did they mean that the older the date, the less accurate the estimation is?

Otherwise Krista, I think it's ok to drop it. We don't really have a good answer to the R3 comment.

On Thu, Jul 16, 2020 at 2:36 PM Kristina Anderson-Teixeira < notifications@github.com> wrote:

THere's also this comment from R1: "Line 266-267: why only in 1966? I think this shows that this variable is not properly assessed. The older the date, the more accurate the estimation of crown position and height were."

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/SCBI-ForestGEO/McGregor_climate-sensitivity-variation/issues/112#issuecomment-659595066, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJNRBEJHVEHOQ4HOALIDTWLR35CBXANCNFSM4O4WMDZQ .

--

Ian McGregor

Ph.D. Student | Center for Geospatial Analytics

He/Him/His

College of Natural Resources

Jordan Hall 4120 | Campus Box 7106

North Carolina State University

2800 Faucette Dr.

Raleigh, NC 27695 USA imcgreg@ncsu.edu | 714-864-1005 | geospatial.ncsu.edu

[teixeirak]

Okay, consider it dropped!

[mcgregorian1]

@teixeirak as per our answer to #106, we have the tables already without crown position as the CP_out suffixes. I will for now address #71 - I don't want to double on something you're checking at the moment.

[teixeirak]

Correct. I reverted the commit where I removed them

[teixeirak]

Then, move Fig. 3 up to being Fig. 1.

@mcgregorian1 , do you like this idea? Or do you think we should keep the drought years figure as 1 and this as 2? (I'm maybe leaning toward the latter, but not sure.)

[mcgregorian1]

I like the drought years figure as number 1, as it gives a visual first and then goes into more numbers in a table

[mcgregorian1]

To me, Figure 3 (NEON) should be later on (remain where it is) because it's linked more to our methods.