FATES restart fix for long runs

Externals_CLM.cfg

@@ -2,7 +2,7 @@
 local_path = src/fates
 protocol = git
 repo_url = https://github.com/NGEET/fates
-tag = sci.1.68.1_api.29.0.0
+tag = sci.1.68.2_api.30.0.0
 required = True
 
 [externals_description]

cime_config/testdefs/ExpectedTestFails.xml

@@ -57,6 +57,13 @@
     </phase>
   </test>
 
+  <test name="ERS_Lm20_Mmpi-serial.1x1_smallvilleIA.I2000Clm50BgcCropQianRs.izumi_gnu.clm-cropMonthlyNoinitial">
+    <phase name="COMPARE_base_rest">
+      <status>FAIL</status>
+      <issue>#2236</issue>
+    </phase>
+  </test>
+
   <!-- fates test suite failures -->
 
   <test name="ERS_Lm12.1x1_brazil.I2000Clm50FatesCruRsGs.cheyenne_intel.clm-FatesFireLightningPopDens">

cime_config/testdefs/testlist_clm.xml

@@ -1562,6 +1562,19 @@
       <option name="comment"  >Add at least one test of a NEON site with PRISM precipitation</option>
     </options>
   </test>
+
+  <test name="ERS_P144x1_Lm25" grid="f10_f10_mg37" compset="I2000Clm51Fates" testmods="clm/FatesColdNoComp">
+       <machines>
+	 <!-- no izumi tests since this has a pe layout specific for cheyenne -->
+	 <machine name="cheyenne" compiler="intel"   category="aux_clm"/>
+	 <machine name="cheyenne" compiler="intel"   category="fates"/>
+       </machines>
+       <options>
+	 <option name="wallclock">00:60:00</option>
+	 <option name="comment"  >This is a long fates test on full dynamics. The 144 count is designed to minimize wall time.</option>
+       </options>
+  </test>
+
   <test name="SMS_Ld10_D_Mmpi-serial" grid="CLM_USRDAT" compset="I1PtClm51Fates" testmods="clm/FatesFireLightningPopDens--clm/NEON/FATES/NIWO">
     <machines>
       <machine name="izumi"    compiler="nag"   category="aux_clm"/>

doc/ChangeLog

@@ -1,4 +1,78 @@
 ===============================================================
+Tag name: ctsm5.1.dev151
+Originator(s): rgknox (Ryan Knox,LAWRENCE BERKELEY NATIONAL LABORATORY,510-495-2153)
+Date: Sat Nov 11 16:53:01 MST 2023
+One-line Summary: Fixes to FATES long run restarts
+
+Purpose and description of changes
+----------------------------------
+
+This is a set of changes that enables exact restart tests to pass with FATES, for 
+	longer periods, particularly those that have elapsed over a year.
+	We removed calls that were incrementing, uncecessary calls, and added key new
+	variables to the restart file (such as the leaf layer carbon balance vector).
+
+Collaborators: @mvdebolskiy, @mvertens, @glemieux, @ekluzek, @ckoven, @rosiealice
+
+Significant changes to scientifically-supported configurations
+--------------------------------------------------------------
+
+No changes to scientifically-supported configurations.
+
+Bugs fixed or introduced
+------------------------
+
+Fixes: FATES#1051 https://github.com/NGEET/fates/issues/1051
+
+Notes of particular relevance for users
+---------------------------------------
+
+This set of changes is introduced, while there is a known test failure in:
+ERS_Lm20_Mmpi-serial.1x1_smallvilleIA.I2000Clm50BgcCropQianRs.izumi_gnu.clm-cropMonthlyNoinitial
+
+This is documented in issue: 2236 and a fix is slated for integration. The nature of the changes 
+	in this PR where obviously orthogonal to the issue.
+
+Substantial timing or memory changes: None
+
+
+Notes of particular relevance for developers:
+---------------------------------------------
+
+A new test was added to aux_clm and fates test suites. The walltime for this test 
+	was 29 minutes. This is an important new test for FATES because it is the 
+	first gridded test that spans significantly over a year (and passes). Various test
+	configurations were explored to find a gridded test that completed 25 months
+	with some attempt at expedience in walltime. A walltime
+	allowance of 60 minutes is set in the test.  Here is the test:
+	ERS_P144x1_Lm25.f10_f10_mg37.I2000Clm51Fates   clm-FatesColdNoComp
+
+
+Testing summary:
+----------------
+  regular tests:
+
+    cheyenne ---- ok
+    izumi ------- ok
+
+Answer changes
+--------------
+
+No answer changes
+
+Other details
+-------------
+
+List any externals directories updated (cime, rtm, mosart, cism, fates, etc.): fates API30
+
+Pull Requests that document the changes (include PR ids):
+
+https://github.com/ESCOMP/CTSM/pull/2199
+https://github.com/NGEET/fates/pull/1098
+
+
+===============================================================
+===============================================================
 Tag name: ctsm5.1.dev150
 Originator(s): rgknox (Ryan Knox,LAWRENCE BERKELEY NATIONAL LABORATORY,510-495-2153)
 Date: Mon Nov  6 14:12:37 MST 2023

doc/ChangeSum

@@ -1,5 +1,6 @@
 Tag                   Who      Date  Summary
 ============================================================================================================================
+    ctsm5.1.dev151   rgknox 11/11/2023 Fixes to FATES long run restarts
     ctsm5.1.dev150   rgknox 11/06/2023 FATES API fix to support future fates npp-fixation coupling, and urgent coupling fixes with E3SM.
     ctsm5.1.dev149 samrabin 11/03/2023 Rearrange leaf/stem "harvest" and fix soil gas diffusivity
     ctsm5.1.dev148 samrabin 11/03/2023 Add GRAINN outputs

src/utils/clmfates_interfaceMod.F90

@@ -1087,7 +1087,8 @@ subroutine dynamics_driv(this, nc, bounds_clump,      &
 
             call ed_update_site(this%fates(nc)%sites(s), &
                   this%fates(nc)%bc_in(s), &
-                  this%fates(nc)%bc_out(s))
+                  this%fates(nc)%bc_out(s), &
+                  is_restarting = .false.)
       enddo
 
 
@@ -1764,9 +1765,17 @@ subroutine restart( this, bounds_proc, ncid, flag, waterdiagnosticbulk_inst, &
                   this%fates(nc)%bc_in(s)%max_rooting_depth_index_col = &
                        min(this%fates(nc)%bc_in(s)%nlevsoil, active_layer_inst%altmax_lastyear_indx_col(c))
 
+                  ! When restarting the model, this subroutine has several
+                  ! procedures that are incremental or don't need to be performed for 
+                  ! during the restart sequence. For the prior, we don't want the restarted
+                  ! run to call these routines more than would had been called during
+                  ! a continuous simulation period, as it would change results. So 
+                  ! we pass in the "is_restarting=.true." flag so we can bypass those procedures
+
                   call ed_update_site( this%fates(nc)%sites(s), &
                         this%fates(nc)%bc_in(s), &
-                        this%fates(nc)%bc_out(s) )
+                        this%fates(nc)%bc_out(s), &
+                        is_restarting = .true. )
 
                end do
 
@@ -1995,7 +2004,8 @@ subroutine init_coldstart(this, waterstatebulk_inst, waterdiagnosticbulk_inst, &
 
               call ed_update_site(this%fates(nc)%sites(s), &
                     this%fates(nc)%bc_in(s), &
-                    this%fates(nc)%bc_out(s))
+                    this%fates(nc)%bc_out(s), &
+                    is_restarting = .false.)
 
            end do