search

EDmodel / ED2

Ecosystem Demography Model
78 stars 112 forks source link

Did somebody change NBG init? #143

Open DanielNScott opened 8 years ago

DanielNScott commented 8 years ago

It used to be the case that restricting INCLUDE_THESE_PFTS to one type and running an NBG simulation would give you only one cohort... it seems like somebody changed this?

Or am I just mis-remembering?

mpaiao commented 8 years ago

It should start with a single cohort, but then the vegetation dynamics should create more cohorts as they cohorts grow and reproduce. If you want to keep this single cohort forever and you don't mind it will remain small, you can also turn of vegetation dynamics (IVEGT_DYNAMICS=0).

DanielNScott commented 8 years ago

Yeah that was what I thought (and was used to) but don't believe it does anymore.

DanielNScott commented 8 years ago

Ok, I just double checked, it does in the mainline but for some reason in one of my branches it does not. Unsure why, since I don't recall ever having made any intentional changes to that.

DanielNScott commented 8 years ago

Aha! Ok, I ran into this again and it seems like INCLUDE_THESE_PFTS get's ignored (in my very slightly modified copy of the mainline) when there's a config.xml present. Is this supposed to be the case?

mdietze commented 8 years ago

That sounds vaguely familiar (though it would have been written ~8yr ago). I may have wanted to avoid inconsistencies between INCLUDE_THESE_PFTS and the PFTs in the xml, in which case adding the PFT to the xml with no parameters (just the ID number) should work to include them. If not, feel free to modify.

fabeit commented 7 years ago

For me it doesn't work like that, it seems to work in an additive manner. If you include pft 2,3 and then your xml has 5,6 (even just the numbers with no parameters), then your run will have 2,3,5,6. Personally I would want that if an xml is present, ONLY the pfts in the xml are included, and the ones in ed2in ignored. I think it's a quick fix.

mdietze commented 7 years ago

That's a bit counter to the original design, which was that the XML is optional and only changed existing PFTs. If you went this route the XML would become required to do a default run and you could end up with a lot of PFTs that are just the number.

I didn't remember that the inclusion of PFTs in an XML makes them active -- that probably should be disabled.