incorrectly assuming at least canopy layer during promotion and demotion

[rgknox]

The algorithm in the canopy promotion/demotion code relies on the assumption that there is at least one canopy layer, particularly by setting this value to 1 (instead of 0) as the lower bound, see here:

https://github.com/NGEET/fates/blob/sci.1.80.3_api.37.0.0/biogeochem/EDCanopyStructureMod.F90#L2303

I don't know if this is creating any spurious results, but I suspect it is not. But, nonetheless, this is not correct. If there are no cohorts, this routine should report 0 canopy layers.

When I tried initializing z to 0 (on that line), I generated errors here: https://github.com/NGEET/fates/blob/sci.1.80.3_api.37.0.0/biogeochem/EDCanopyStructureMod.F90#L189-L191

I think this error triggered, because newly recruited cohorts were being added into the patches lowest canopy layer, (which is defined as patch%NCL_p), but that value was zero if the recruitment was applied to a patch with no existing cohorts.

Again, I don't think this is a big issue, but I wanted to document my thoughts in case it comes up. Here is a set of changes I would apply if I wanted to have the number of potential canopy layers (rightfully) return a 0. Note that it also initializes the NCL_p of a patch to -1, which is more appropriate, because at the time of creation, the number of canopy layers should not be assumed, it is unknown.

diff --git a/biogeochem/EDPhysiologyMod.F90 b/biogeochem/EDPhysiologyMod.F90
index 9f0db5ff..25e2e519 100644
--- a/biogeochem/EDPhysiologyMod.F90
+++ b/biogeochem/EDPhysiologyMod.F90
@@ -2542,7 +2542,7 @@ contains
             height             = EDPftvarcon_inst%hgt_min(ft)
             stem_drop_fraction = prt_params%phen_stem_drop_fraction(ft)
             fnrt_drop_fraction = prt_params%phen_fnrt_drop_fraction(ft)
-            l2fr               = currentSite%rec_l2fr(ft, currentPatch%NCL_p)
+            l2fr               = currentSite%rec_l2fr(ft, max(1,currentPatch%NCL_p))
             crowndamage        = 1 ! new recruits are undamaged
 
             ! calculate DBH from initial height 
@@ -2774,7 +2774,7 @@ contains
                call create_cohort(currentSite, currentPatch, ft, cohort_n,     &
                   height, 0.0_r8, dbh, prt, efleaf_coh, effnrt_coh, efstem_coh,  &
                   leaf_status, recruitstatus, init_recruit_trim, 0.0_r8,       &
-                  currentPatch%NCL_p, crowndamage, currentSite%spread, bc_in)
+                  max(1,currentPatch%NCL_p), crowndamage, currentSite%spread, bc_in)
 
                ! Note that if hydraulics is on, the number of cohorts may have
                ! changed due to hydraulic constraints.
diff --git a/biogeochem/FatesPatchMod.F90 b/biogeochem/FatesPatchMod.F90
index f8afc711..24298f81 100644
--- a/biogeochem/FatesPatchMod.F90
+++ b/biogeochem/FatesPatchMod.F90
@@ -737,7 +737,7 @@ module FatesPatchMod
 
       this%tr_soil_dir(:) = 1.0_r8
       this%tr_soil_dif(:) = 1.0_r8
-      this%NCL_p          = 1
+      this%NCL_p          = -1
 
       this%changed_landuse_this_ts = .false.