Merge pull request #1208 from rgknox/history-zerofixes

Updates and fixes to history zeroing and flushing

main/EDMainMod.F90

@@ -166,12 +166,6 @@ subroutine ed_ecosystem_dynamics(currentSite, bc_in, bc_out)
        call currentSite%flux_diags(el)%ZeroFluxDiags()
     end do
 
-    ! zero dynamics (upfreq_in = 1) output history variables
-    call fates_hist%zero_site_hvars(currentSite,upfreq_in=1)
-
-    ! zero nutrient fluxes (upfreq_in=5) output hist variables
-    call fates_hist%zero_site_hvars(currentSite,upfreq_in=5)
-
     ! Call a routine that simply identifies if logging should occur
     ! This is limited to a global event until more structured event handling is enabled
     call IsItLoggingTime(hlm_masterproc,currentSite)

main/EDTypesMod.F90

@@ -150,8 +150,13 @@ module EDTypesMod
   type, public :: site_fluxdiags_type
 
      ! ----------------------------------------------------------------------------------
-     ! Diagnostics for fluxes into the litter pool from plants
-     ! these fluxes are the total from 
+     ! Diagnostics of fluxes
+     ! These act as an intermediary to write fluxes to the history
+     ! file after number densities of plants have changed. They also
+     ! allow the history flux diagnostics to be rebuilt during restart
+     !
+     !
+     ! Litter fluxes are the total from 
      ! (1) turnover from living plants
      ! (2) mass transfer from non-disturbance inducing mortality events
      ! (3) mass transfer from disturbance inducing mortality events
@@ -162,6 +167,13 @@ module EDTypesMod
      real(r8) :: cwd_bg_input(1:ncwd)               
      real(r8),allocatable :: leaf_litter_input(:)
      real(r8),allocatable :: root_litter_input(:)
+
+     ! This variable is slated as to-do, but the fluxdiags type needs
+     ! to be refactored first. Currently this type is allocated
+     ! by chemical species (ie C, N or P). GPP is C, but not N or P (RGK 0524)
+     ! Previous day GPP [kgC/m2/year], partitioned by size x pft
+     !real(r8),allocatable :: gpp_prev_scpf(:)
+
 
    contains
 
@@ -464,6 +476,11 @@ subroutine ZeroFluxDiags(this)
       this%cwd_bg_input(:)      = 0._r8
       this%leaf_litter_input(:) = 0._r8
       this%root_litter_input(:) = 0._r8
+
+      ! We don't zero gpp_prev_scpf because this is not
+      ! incremented like others, it is assigned at the end
+      ! of the daily history write process
+
 
       return
     end subroutine ZeroFluxDiags