=================== General Information =================== Purpose of Run: 1850 control run (initially a tuning experiment for 1-degree) Scientific Lead: Jim Hurrell Software Engineering Lead: Mariana Vertenstein, Jon Wolfe Assigned to: Andy Mai Date: 2009-09-03 Run Length: 1300 years (initially 30 years) ========================= Case Creation Information (all fields are required) ========================= CCSM tag: ccsm4_0_beta29 (ccsm4_0_beta23 for first 300 years) Case Name: b40.1850.track1.1deg.006 Machine: bluefire Compset: B_1850_TRACK1_CN Resolution: 0.9x1.25_gx1v6 ============================= Pre-Configuration Information ============================= Runtype: Initially hybrid - after 300 years a branch from the initial run to allow this control to continue with a new tag containing enhanced IPCC output and diagnostics. Please note that the two tags are bfb. RUN_STARTDATE = 0001-01-01 RUN_REFCASE = 1850_f09_g16_c090901 RUN_REFDATE = 0001-01-01 Note: the necessary restart files can be found in: /fis01/cgd/cseg/csm/inputdata/ccsm4_init/1850_f09_g16_c090901 env_conf.xml mods _________________ none env_mach_pes.xml mods _____________________ component comp_pes root_pe tasks x threads (stride) --------- ------ ------- ------ ------ ------ cpl = cpl 320 0 320 x 1 (1 ) glc = sglc 1 0 1 x 1 (1 ) lnd = clm 128 320 128 x 1 (1 ) ice = cice 320 0 320 x 1 (1 ) atm = cam 448 0 448 x 1 (1 ) ocn = pop2 64 448 64 x 1 (1 ) ============================== Post-Configuration Information ============================== Buildconf _________ * make sure ice albedos get set to the following in the namelist: dT_mlt = 1.5_dbl_kind rsnw_melt = 1500._dbl_kind ====================== SourceMods Information ====================== 1. Starting with year 0851, a change in the lnd model to prevent slow water leakage to the ocean was introduced. This change is not bfb. See the beta35 tag. Also, the restart files were massaged by Keith Oleson before this restart. From an email from Keith Oleson: The code mods are in: /fis01/cgd/tss/oleson/clm3_6_48/scripts/clm4_cn_ss128n/SourceMods/src.clm There are three routines there: BalanceCheckMod.F90 SoilHydrologyMod.F90 driver.F90 These mods are to fix the trend in aquifer water that results in a trend in ocean salinity. 2. At year 1001, a code mode to the cice routine ice_restart.F90 was incorporated into the build. From an email from David Bailey: I have a SourceMod in: /blhome/dbailey/ice_restart.F90 to fix the FY area restart problem. This is purely diagnostic (output) and so does not the calculation in any way (it is still b4b). ========================== Performance/Cost Estimates ========================== * 13.4 years/day across 8 nodes for beta23 * 13.0 years/day across 8 nodes for beta29 ==================== Special Instructions ==================== For the branch at 300 years, this will require some slight kludges. We want the branch to have the exact same case name as for the first 300 years, so it will mean: * renaming the original case directory to /fis01/cgd/cseg/csm/runs/ccsm4_0/b40.1850.track1.1deg.006.beta23 * renaming the original run directory in a similar way * making sure all data from the original 300 year run has been archived * running create_newcase with the same case name, b40.1850.track1.1deg.006 and with the "-skip_rundb" option -- we'll have to modify the run database to handle this situation after the branch run has started * please make sure brnch_retain_casename is set to true in the new drv_in file * please make sure this branch is well-documented in the case directory On 12 November 2009 and starting at 0715-01-01 we are using the new nitrogen deposition file (the one with corrected units dated 091106). Also, see "New run instructions 12 November 2009" below concerning an ocn source mod put in place at this time. On 24 November 2009 and starting at 0851-01-01 we are using new SourceMods for clm that fix a problem with its aquifer water treatment. this issue had been causing long-term salinity issues in the ocean. Also, see "New run instructions 24 November 2009" below for more details. ==================== Pre-Run Instructions ==================== * Run create_production_test (on the side) -- not necessary for the branch at 300 years * Run debug smoke test (on the side) -- not necessary for the branch at 300 years ================ Run Instructions ================ Run Length: 1300 years (initially 30 years) Account key: 97780014 Priority/Targeted queue: regular/premium Other: * New run instructions 12 November 2009 and at 0715-01-01 in the run: Introduce the bug-fixed version of overflows.F90 from a directory that is compatible with the ccsm4_0_beta29 tag: cp /fis/cgd/cseg/people/njn01/test/overflows/c40.32.C.gx1v6.ovf.test/SourceMods/src.pop2/overflows.F90 $CASE/SourceMods/src.pop2 then rebuild the ocean model before running * New run instructions 24 November 2009: On this date, and at 0851-01-01 in the run we will introduce the bug-fixed versions of the following CLM source files: * BalanceCheckMod.F90 * SoilHydrologyMod.F90 * driver.F90 These files were copied from: * /fis01/cgd/tss/oleson/clm3_6_48/scripts/clm4_cn_ss128n/SourceMods/src.clm At this point, we also replaced the clm restart file with one that had been modifed by Dave Lawrence, correcting 11 columns to go along with the above code mods. The following file was copied into this case's run directory: * /ptmp/dlawren/corrected_restart/modified.b40.1850.track1.1deg.006.clm2.r.0851-01-01-00000.nc For doumentation sake, the original clm restart file was renamed: * orig.b40.1850.track1.1deg.006.clm2.r.0851-01-01-00000.nc and kept in the run directory. Also please note that this case did not correctly pick up the clm SourceMods during rebuild, due to an error in clm.buildexe.csh which included a relative path for SourceMods insted of an explicit one. ================ Diagnostics Plan ================ ====================== Additional Information ====================== Please do the RESTOM after 5 and 10 years to compare with the 005 run. Please note that CSEG did additional testing to make sure the two tags used for this run are bfb for this compset, resolution and machine.Return to Top
Complete the following checklist prior to beginning the production run: 1. Update status file: /web/web-data/cseg/ccsm4_0_runs/b40.1850.track1.1deg.006/status.txt [assigned, setup in progress, running] 2. Complete the following table, as necessary, showing the component liaison's name and the date the setup was approved. Component Liaison/ Date Approved Reviewer ================+==========================+================== atm rneale Wed. Sep 2 16:36 cpl [kauff,mvertens,tcraig,other] ---- ice dbailey ---- lnd slevis Sep 2 16:45 ocn njn01 Wed Sep 2 16:07:12 MDT 2009 njn01 Thu Nov 12 17:59:18 MST 2009 (recompiled with overflows.F90 mod) erik Thu Nov 12 15:26:17 MST 2009 env_ file mvertens Wed. Sep 2 16:35 settings 3. Create_production_test completed mai, 2009-09-03 4. Debug smoke test completed mai, 2009-02-03 5. Performance review completed [who,when]Return to Top
1. It appears in the Run Specifications above: ============================== Post-Configuration Information ============================== Buildconf _________ * make sure ice albedos get set to the following in the namelist: dT_mlt = 1.5_dbl_kind rsnw_melt = 1500._dbl_kind By default, these are not in the namelist. Dave Bailey assures me that these are the default values for these two variables. This was verified by examination of the first ice.log file where they are printed out directly from the model. 2. This run was selected as the baseline to be extended to 1300 years. It was branched at year 300 using the beta29 tag to allow for more output variables and extended diagnostic information. Beta29 is bfb with beta23. The cice variables shear and divu in the model history files were also changed from averaged to instantaneous to comply with IPCC requirements. The pop output now includes (for out-of-the-box for beta29) an "h2" set of daily-averaged history files. From Nancy Norton: The "h2" files are daily averages of surface potential temperature, surface potential temperature squared, mixed-layer depth and maximum mixed-layer depth, all bundled into monthly timeseries files. The naming convention for the daily averaged fields file in this tag used the string "h2" to identify the daily averaged files, but the convention changed to "nday1" in later pop2 tags. The h2/nday1 files are standard output from all pop2 tags, beginning from the time at which the "tavg streams" capability was added to pop2. 3. Starting on 12 Nov 2009 at model year 0715 a small change was made to the overflows.F90 module. From the CCSM Gatekeeper ccsm4_0_beta33 email: in overflows.F90 module, replace constants with CCSM shared-constants values This change results in differences larger than roundoff. Nancy Norton notes that first month differences between the 006 and 011 runs were two orders of magnitude greater than this difference, so Gokhan deemed this a good time to make this small change. Also on 12 November 2009 and starting at 0715 we are using the new nitrogen deposition file (the one with corrected units dated 091106). 4. Years 0851 --> 0900 had some daily cice output added to the history files. Mods to Buildconf/cice.buildnml.csh: setup_nml: histfreq = 'm','d','x','x','x' icefields_nml: f_FY = 'mdxxx' f_Tsfc = 'mdxxx' f_aero = 'mxxxx' f_aice = 'mdxxx' f_aicen = 'mdxxx' f_aisnap = 'mdxxx' f_albice = 'mdxxx' f_albpnd = 'mdxxx' f_albsno = 'mdxxx' f_apondn = 'mdxxx' f_congel = 'mdxxx' f_daidtd = 'mdxxx' f_daidtt = 'mdxxx' f_divu = 'mxxxx' f_dvidtd = 'mdxxx' f_dvidtt = 'mdxxx' f_evap_ai = 'mdxxx' f_faero_atm = 'mxxxx' f_faero_ocn = 'mxxxx' f_fcondtop = 'mdxxx' f_fhocn = 'mxxxx' f_fhocn_ai = 'mdxxx' f_flat_ai = 'mdxxx' f_flwdn = 'mdxxx' f_flwup_ai = 'mdxxx' f_frazil = 'mxxxx' f_fresh = 'mxxxx' f_fresh_ai = 'mxxxx' f_frz_onset = 'xxxxx' f_frzmlt = 'xxxxx' f_fs = 'mdxxx' f_fsalt = 'mxxxx' f_fsalt_ai = 'mxxxx' f_fsen_ai = 'mdxxx' f_fswabs_ai = 'mdxxx' f_fswdn = 'mdxxx' f_fswthru_ai = 'mdxxx' f_fy = 'mdxxx' f_hi = 'mdxxx' f_hisnap = 'mdxxx' f_hs = 'mdxxx' f_icepresent = 'mxxxx' f_meltb = 'mdxxx' f_meltl = 'mdxxx' f_melts = 'mdxxx' f_meltt = 'mdxxx' f_mlt_onset = 'xxxxx' f_opening = 'mxxxx' f_rain_ai = 'mdxxx' f_shear = 'mxxxx' f_sig1 = 'mxxxx' f_sig2 = 'mxxxx' f_snoice = 'mxxxx' f_snow_ai = 'mdxxx' f_sss = 'xxxxx' f_sst = 'xxxxx' f_strairx = 'mxxxx' f_strairy = 'mxxxx' f_strcorx = 'mxxxx' f_strcory = 'mxxxx' f_strength = 'mxxxx' f_strintx = 'mxxxx' f_strinty = 'mxxxx' f_strocnx = 'mxxxx' f_strocny = 'mxxxx' f_strtltx = 'xxxxx' f_strtlty = 'xxxxx' f_uocn = 'xxxxx' f_uvel = 'mdxxx' f_vicen = 'mdxxx' f_vocn = 'xxxxx' f_vvel = 'mdxxx' 5. At year 0851, a change in the lnd model to prevent slow water leakage to the ocean was introduced. This change is not bfb. See the beta35 tag. Also, the restart files were massaged by Keith Oleson before this restart. From an email from Keith Oleson: The code mods are in: /fis01/cgd/tss/oleson/clm3_6_48/scripts/clm4_cn_ss128n/SourceMods/src.clm There are three routines there: BalanceCheckMod.F90 SoilHydrologyMod.F90 driver.F90 These mods are to fix the trend in aquifer water that results in a trend in ocean salinity. 6. At year 1001, a code mode to the cice routine ice_restart.F90 was incorporated into the build. From an email from David Bailey: I have a SourceMod in: /blhome/dbailey/ice_restart.F90 to fix the FY area restart problem. This is purely diagnostic (output) and so does not the calculation in any way (it is still b4b).