search

forc-db / IPCC-EFDB-integration

Creative Commons Attribution 4.0 International
4 stars 3 forks source link

clarify how to categorize <20 yr old forests following severe disturbance of forest or unknown land use #7

Closed teixeirak closed 1 year ago

teixeirak commented 3 years ago

my message to Valentyna;

Following up on your answer to question 1, my understanding is that if a forest undergoes severe disturbance (e.g., clear-cut logging, severe fire), the regrowth stands would be considered FF. Is that correct? In the case I described (stand age <20 and the former land use is not recorded in ForC), these could have been severe disturbance to forest land, so it seems LF might not be appropriate? Should these just be classified as Forest Land? Of should we set them aside to look up?

teixeirak commented 3 years ago

(These cases can be avoided in the first data delivery.)

teixeirak commented 3 years ago

from V:

You are right, in case previous land use for young forest (<20 yr.) is not known, this area could be affected by severe disturbances. Then, please classify it as FF. If the info on the disturbances is available to be input into the EFDB form, it is good to include.

teixeirak commented 3 years ago

There are no cases of this in the first data delivery, but we'll eventually need to deal with this.

teixeirak commented 2 years ago

@ValentineHerr , I'm revisiting old issues and noticing that we never resolved this. This may affect some studies in the 2nd batch, or the one we're preparing now. I'm trying to sort through this.

teixeirak commented 2 years ago

Recapping from these instructions:

if distmrs.type= "Burned", "StandClearing", or there's a missing value code, we don't know PAST LAND-USE. (from looking at distmrs.type).

Our default seems to be to just call this forest land.

This affects some studies in a funny way:

Here (Tepley_2017), all sites are forest land remaining forest land (having burned in severe fires), yet sites <20 are just classified as forest land because our system doesn't know the previous land use.

image

This relates to issue #29.

@ValentineHerr , any ideas on how we might deal with this? It seems like adding a past land use field to ForC would be useful (not just for this), but I'm not sure where we'd put it...

teixeirak commented 2 years ago

This also affects:

teixeirak commented 2 years ago

This isn't technically wrong, just loss of potentially useful info.

ValentineHerr commented 2 years ago

@ValentineHerr , any ideas on how we might deal with this? It seems like adding a past land use field to ForC would be useful (not just for this), but I'm not sure where we'd put it...

I need to look at this more carfully. Right now I am concerned about columns "dist1.ID", "dist1.type", "dist1.mort", "dist1.year", "dist2.ID", "dist2.type", "dist2.mort", "dist2.year".... I need to remind myself about what is going on with those in PLOTS

teixeirak commented 2 years ago

if distmrs.type= "Burned", "StandClearing", or there's a missing value code, we don't know PAST LAND-USE. (from looking at distmrs.type).

Actually, on second thought, "StandClearing" should always refer to forest remaining forest, so let's classify that as such.

For "Burned", let's apply the following rules:

teixeirak commented 2 years ago

Rothstein_2004 is also affected by this issue.

One easy option would be to just fix these by hand at the time of approval, but if this is going to be a relatively quick change to the code, we can fix it properly that way.

ValentineHerr commented 2 years ago

@teixeirak, I am concerned that the issue is more complicated as I am realizing now that I have not touched the script creating PLOTS in a very long time and it maybe obsolete.... I think it was made to deal with the previous PLOTS version, manually made, so I don't think it is actually filling some of the columns based on HISTORY but based on what is already in PLOTS..... That is not good.

So for now, I think it is best to manually edit at time of approval.