Merge branch 'extra_precision_control' into release-clm5.0

Add CN prec. control call to fix problems related to small neg. values

Small negative values (roughly roundoff-level different from zero) in
frootc (and possibly other quantities) were occasionally creating
problems with carbon isotope fluxes and FPI in the first time step of
the year, at the time of transient landcover change. This tag fixes the
problem by introducing an extra call to SoilBiogeochemPrecisionControl
in between computing the patch-level transient landcover fluxes and
moving these to column-level. In particular, this truncates small
negative values of decomp_cpools_vr_col to zero, which prevents the
previous blow-ups.

For most of the problematic fields, the explanation seems to be: frootc
can sometimes be negative; this is intentional. Negative frootc causes
negative dwt_frootc_to_litter if the patch in question is shrinking. The
resulting negative fluxes cause problems in the ciso calculation. This
can be worked around by inserting an extra precision control call
between the calculation of the dwt fluxes and the ciso fluxes, so that
small negative dwt fluxes are set to 0.

This does not necessarily fully explain the issue with FPI, but the
insertion of the extra precision control call fixes that issue, too.

For more details, see the discussion in
#741

Resolves #741

doc/ChangeSum

@@ -1,5 +1,6 @@
 Tag                   Who      Date  Summary
 ============================================================================================================================
+release-clm5.0.26    sacks 07/29/2019 Add a CN precision control call to fix problems related to small negative values
 release-clm5.0.25     erik 05/29/2019 Change two files from NetCDF-4 format to NetCDF-3 (because some machines have trouble with NetCDF-4 in pnetcdf)
 release-clm5.0.24     erik 05/22/2019 Correct ndep end year for SSPs, 2-degree CMIP6WACCMDECK with C-isotopes off, fixes mksurfdata for high resolution
 release-clm5.0.23     erik 05/15/2019 Update cime to bring in CO2 transient files for the CMIP6 SSP's as well as presaero for three of them

doc/release-clm5.0.ChangeLog

@@ -1,4 +1,167 @@
 ===============================================================
+Tag name:  release-clm5.0.26
+Originator(s):  sacks (Bill Sacks)
+Date: Mon Jul 29 18:11:26 MDT 2019
+One-line Summary: Add a CN precision control call to fix problems related to small negative values
+
+Purpose of this version:
+------------------------
+
+Small negative values (roughly roundoff-level different from zero) in
+frootc (and possibly other quantities) were occasionally creating
+problems with carbon isotope fluxes and FPI in the first time step of
+the year, at the time of transient landcover change. This tag fixes the
+problem by introducing an extra call to SoilBiogeochemPrecisionControl
+in between computing the patch-level transient landcover fluxes and
+moving these to column-level. In particular, this truncates small
+negative values of decomp_cpools_vr_col to zero, which prevents the
+previous blow-ups.
+
+For most of the problematic fields, the explanation seems to be: frootc
+can sometimes be negative; this is intentional. Negative frootc causes
+negative dwt_frootc_to_litter if the patch in question is shrinking. The
+resulting negative fluxes cause problems in the ciso calculation. This
+can be worked around by inserting an extra precision control call
+between the calculation of the dwt fluxes and the ciso fluxes, so that
+small negative dwt fluxes are set to 0.
+
+This does not necessarily fully explain the issue with FPI, but the
+insertion of the extra precision control call fixes that issue, too.
+
+For more details, see the discussion in
+https://github.com/ESCOMP/ctsm/issues/741
+
+
+CTSM Master Tag This Corresponds To: N/A
+
+Summary of changes:
+-------------------
+
+Issues fixed (include CTSM Issue #):
+- Resolves ESCOMP/ctsm#741
+
+Science changes since: release-clm5.0.25
+    None
+
+Software changes since: release-clm5.0.25
+    Extra call to SoilBiogeochemPrecisionControl
+
+Changes to User Interface since: release-clm5.0.25
+    None
+
+Testing:
+--------
+
+  build-namelist tests:
+
+    cheyenne - not run
+
+  unit-tests (components/clm/src):
+
+    cheyenne - not run
+    hobart --- not run
+
+  tools-tests (components/clm/test/tools):
+
+    cheyenne - not run
+    hobart --- not run
+
+  PTCLM testing (components/clm/tools/shared/PTCLM/test):
+
+     cheyenne - not run
+     hobart --- not run
+
+  regular tests (aux_clm):
+
+    cheyenne_intel ---- ok
+    cheyenne_gnu ------ ok
+    hobart_nag -------- ok
+    hobart_pgi -------- ok
+    hobart_intel ------ ok
+
+    ok means tests pass (other thana expected failures), answers change
+    as expected for some tests
+
+  regular tests (prealpha):
+
+    cheyenne_intel - not run
+    cheyenne_gnu --- not run
+    hobart_nag ----- not run
+
+  regular tests (prebeta):
+
+    cheyenne_intel - not run
+    cheyenne_gnu --- not run
+    hobart_nag ----- not run
+
+Summary of Answer changes:
+-------------------------
+
+If the tag used for baseline comparisons was NOT the previous tag, note that here: previous
+
+Changes answers relative to baseline: YES
+
+  If a tag changes answers relative to baseline comparison the
+  following should be filled in (otherwise remove this section):
+
+  Summarize any changes to answers, i.e.,
+    - what code configurations: Transient cases and a few other
+      configurations as noted below
+    - what platforms/compilers: all
+    - nature of change (roundoff; larger than roundoff/same climate; new climate): 
+      Larger than roundoff / same climate
+
+      According to Keith Oleson's analysis of the LMWG diagnostics:
+      Changes are possibly greater than roundoff-level, but are not
+      climate changing or scientifically meaningful.
+
+      Other than transient cases (including cases with transient glacier
+      area as well as those with transient natural vegetation / crop
+      area), the test suite showed differences in the following cases:
+
+      - Two present-day tests that specify changes in soil layer
+        structure
+        (ERI_D_Ld9.ne30_g16.I2000Clm50BgcCruGs.cheyenne_intel.clm-vrtlay
+        and
+        ERS_D_Ld3.f10_f10_musgs.I2000Clm50BgcCruGs.cheyenne_intel.clm-deepsoil_bedrock),
+        have changes in many fields (for the latter: mostly looks
+        roundoff-level, but a few greater than roundoff-level diffs)
+
+      - The present-day ciso test,
+        ERP_D_Ld5.f10_f10_musgs.I2000Clm50BgcCruGs.cheyenne_gnu.clm-ciso_flexCN_FUN,
+        has changes in a few c13 and c14 fields
+
+      - There are changes in many fields in SSP tests
+
+      - ERP_D_P36x2_Ld3.f10_f10_musgs.I2000Clm50BgcCruGs.cheyenne_intel.clm-noFUN_flexCN
+        (and a similar test on hobart) have roundoff-level changes in
+        SMINN_vr, SMIN_NH4_vr and SMIN_NO3_vr
+
+  URL for LMWG diagnostics output for new climate:
+  http://webext.cgd.ucar.edu/I20TR/clm50_cesm20R_2deg_GSWP3V1_issue741_hist/lnd/clm50_cesm20R_2deg_GSWP3V1_issue741_hist.1995_2014-clm50_cesm20R_2deg_GSWP3V1_hist.1995_2014/setsIndex.html
+
+  Will new REFCASES need to be made for cesm and/or CAM?: No
+  (This will likely be true if the LII tests failed)
+
+Detailed list of changes:
+------------------------
+
+Externals being used:  No updates
+
+   cism:   release-cesm2.0.04
+   rtm:    release-cesm2.0.02
+   mosart: release-cesm2.0.03
+   cime:   cim5.6.16
+   FATES:  fates_s1.21.0_a7.0.0_br_rev2
+   PTCLM:  PTCLM2_180611
+
+CTSM Tag versions pulled over from master development branch: None
+
+Pull Requests that document the changes (include PR ids): None
+(https://github.com/ESCOMP/ctsm/pull)
+
+===============================================================
+===============================================================
 Tag name:  release-clm5.0.25
 Originator(s):  erik (Erik Kluzek)
 Date: Wed May 29 11:16:13 MDT 2019

src/biogeochem/CNVegetationFacade.F90

@@ -721,6 +721,13 @@ subroutine DynamicAreaConservation(this, bounds, clump_index, &
          soilbiogeochem_nitrogenstate_inst)
     call t_stopf('CNUpdateDynPatch')
 
+    ! This call fixes issue #741 by performing precision control on decomp_cpools_vr_col
+    call t_startf('SoilBiogeochemPrecisionControl')
+    call SoilBiogeochemPrecisionControl(num_soilc_with_inactive, filter_soilc_with_inactive,  &
+         soilbiogeochem_carbonstate_inst, c13_soilbiogeochem_carbonstate_inst, &
+         c14_soilbiogeochem_carbonstate_inst,soilbiogeochem_nitrogenstate_inst)
+    call t_stopf('SoilBiogeochemPrecisionControl')
+
     call t_startf('dyn_cnbal_col')
     call dyn_cnbal_col(bounds, clump_index, column_state_updater, &
          soilbiogeochem_carbonstate_inst, c13_soilbiogeochem_carbonstate_inst, &