========== 20190102: 1. Checked how vcmax is computed in CLM4.5. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98481909 2018-08-27 10:30 ../build/Linux/bin/LDASsaCN_mpi.x* > ls -l ../../clm4.5_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70928420 2018-09-18 13:53 ../../clm4.5_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4/clm4_point/run > ll ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 94470615 2017-11-03 15:27 ../build/Linux/bin/LDASsaCN_mpi.x* > ll ../../clm4_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70033364 2018-07-13 16:45 ../../clm4_DE720/build/Linux/bin/LDASsaCN_mpi.x* For both CLM4 and CLM4.5, re-compile the drivers for the debugging runs without any modifications to the drivers. CLM4.5: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/ > source $ESMADIR/src/g5_modules > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake clean > gmake install BOPT=g > cd ../../.. > /bin/cp -pr Linux/ exec/clm4.5_debug/. CLM4: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/ > source $ESMADIR/src/g5_modules > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake clean > gmake install BOPT=g > cd ../../.. > /bin/cp -pr Linux/ exec/clm4_debug/. Do the point runs on DDT. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4/clm4_point/run > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i Are parsun and parsha in Photosynthesis per LAI? They are per VAI. However, after calling compute_rc, they are multiplied by LAI instead of VAI. Why?! Are vcmaxcintsun and vcmaxcintsha correct? ========== 20190103: 1. Checked the original CLM4.5 code about the photosynthesis calculation. (1) Is dividing par by vai (=elai+esai) in compute_rc.F90 consistent with the original CLM4.5? Yes! See below. Subroutine Photosynthesis in /discover/nobackup/fzeng/clm_orig/cesm1_2_2/models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90: real(r8), pointer :: par_z(:,:) ! par absorbed per unit lai for canopy layer (w/m**2) real(r8) :: fnps ! fraction of light absorbed by non-photosynthetic pigments if (phase == 'sun') then lai_z => pps%laisun_z par_z => pef%parsun_z psn_z => pcf%psnsun_z ci_z => pcf%cisun_z lmr_z => pcf%lmrsun_z rs_z => pps%rssun_z psn => pcf%psnsun psn_wc => pcf%psnsun_wc psn_wj => pcf%psnsun_wj psn_wp => pcf%psnsun_wp lmr => pcf%lmrsun rs => pps%rssun vcmaxcint => pps%vcmaxcintsun if ( use_c13 ) then alphapsn => pps%alphapsnsun end if else if (phase == 'sha') then lai_z => pps%laisha_z par_z => pef%parsha_z psn_z => pcf%psnsha_z ci_z => pcf%cisha_z lmr_z => pcf%lmrsha_z rs_z => pps%rssha_z psn => pcf%psnsha psn_wc => pcf%psnsha_wc psn_wj => pcf%psnsha_wj psn_wp => pcf%psnsha_wp lmr => pcf%lmrsha rs => pps%rssha vcmaxcint => pps%vcmaxcintsha if ( use_c13 ) then alphapsn => pps%alphapsnsha end if end if Subroutine SurfaceRadiation in /discover/nobackup/fzeng/clm_orig/cesm1_2_2/models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90: parsun_z(p,iv) = forc_solad(g,ib)*fabd_sun_z(p,iv) + forc_solai(g,ib)*fabi_sun_z(p,iv) parsha_z(p,iv) = forc_solad(g,ib)*fabd_sha_z(p,iv) + forc_solai(g,ib)*fabi_sha_z(p,iv) Subroutine TwoStream /discover/nobackup/fzeng/clm_orig/cesm1_2_2/models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90: ! absorbed PAR (per unit sun/shade lai+sai) laisum = elai(p)+esai(p) fabd_sun_z(p,1) = fabd_sun(p,ib) / (fsun_z(p,1)*laisum) fabi_sun_z(p,1) = fabi_sun(p,ib) / (fsun_z(p,1)*laisum) fabd_sha_z(p,1) = fabd_sha(p,ib) / ((1._r8 - fsun_z(p,1))*laisum) fabi_sha_z(p,1) = fabi_sha(p,ib) / ((1._r8 - fsun_z(p,1))*laisum) So dividing par by (elai+esai) in compute_rc.F90 is consistent with the original CLM4.5. (2) laisun and laisha are two components of elai (lai not buried by snow) instead of tlai (lai not buried and buried by snow) in compute_rc.F90. Is this consistent with the original CLM4.5? Yes. See below. Subroutine SurfaceRadiation in /discover/nobackup/fzeng/clm_orig/cesm1_2_2/models/lnd/clm/src/clm4_5/biogeophys/SurfaceRadiationMod.F90: real(r8), pointer :: tlai_z(:,:) ! tlai increment for canopy layer tlai_z => pps%tlai_z laisun(p) = 0._r8 laisha(p) = 0._r8 do iv = 1, nrad(p) laisun_z(p,iv) = tlai_z(p,iv) * fsun_z(p,iv) laisha_z(p,iv) = tlai_z(p,iv) * (1._r8 - fsun_z(p,iv)) laisun(p) = laisun(p) + laisun_z(p,iv) laisha(p) = laisha(p) + laisha_z(p,iv) end do Subroutine SurfaceAlbedo /discover/nobackup/fzeng/clm_orig/cesm1_2_2/models/lnd/clm/src/clm4_5/biogeophys/SurfaceAlbedoMod.F90: ! Diagnose number of canopy layers for radiative transfer, in increments of dincmax. ! Add to number of layers so long as cumulative leaf+stem area does not exceed total ! leaf+stem area. Then add any remaining leaf+stem area to next layer and exit the loop. ! Do this first for elai and esai (not buried by snow) and then for the part of the ! canopy that is buried by snow. ! ------------------ ! tlai_z = leaf area increment for a layer ! tsai_z = stem area increment for a layer ! nrad = number of canopy layers above snow ! ncan = total number of canopy layers ! ! tlai_z summed from 1 to nrad = elai ! tlai_z summed from 1 to ncan = tlai ! tsai_z summed from 1 to nrad = esai ! tsai_z summed from 1 to ncan = tsai ! ------------------ ! Sun/shade big leaf code uses only one layer (nrad = ncan = 1), triggered by ! nlevcan = 1 if (nlevcan == 1) then nrad(p) = 1 ncan(p) = 1 tlai_z(p,1) = elai(p) tsai_z(p,1) = esai(p) else if (nlevcan > 1) then ... end if So for nrad = ncan = 1 in current Catchment-CN, using elai in compute_rc.F90 is correct. (3) In compute_rc.F90, parsun and parsha, and therefore psnsun and psnsha, are per vai. In process_cn.F90, after calling compute_rc: ! save fluorescence and photosynthesis as diagnostic ! -------------------------------------------------- cn_diagn(:)%sif = 0. cn_diagn(:)%psn = 0. do nz = 1,num_zon do nv = 1,num_veg cn_diagn(:)%sif = cn_diagn(:)%sif + fveg(:,nv,nz)*wtzone(:,nz)* & (sifsun(:,nv,nz)*laisun(:,nv,nz) + sifsha(:,nv,nz)*laisha(:,nv,nz)) cn_diagn(:)%psn = cn_diagn(:)%psn + fveg(:,nv,nz)*wtzone(:,nz)* & (psnsun(:,nv,nz)*laisun(:,nv,nz) + psnsha(:,nv,nz)*laisha(:,nv,nz)) end do end do psnsun and psnsha are multiplied by laisun and laisha instead of vaisun and vaisha. Is this consistent with the original CLM4.5? Yes. See below. It is done in this way probably to avoid the par absorbed by stem to be used for photosynthesis. Subroutine CanopyFluxes in /discover/nobackup/fzeng/clm_orig/cesm1_2_2/models/lnd/clm/src/clm4_5/biogeophys/CanopyFluxesMod.F90: real(r8), pointer :: fpsn(:) ! photosynthesis (umol CO2 /m**2 /s) ! total photosynthesis fpsn(p) = psnsun(p) *laisun(p) + psnsha(p) *laisha(p) And laisun and laisha are from elai (see above). ========== 20190104: 1. Double checked subroutine compute_rc and subroutine Photosynthesis in compute_rc.F90. (1) How is tlai and t10 used in subroutine Photosynthesis? tlai is elai, consistent with the original CLM4.5. See notes on 20190103. lai is laisun or laisha, consistent with the original CLM4.5. jmax25top = (2.59_r8 - 0.035_r8*min(max((t10(p)-tfrz),11._r8),35._r8)) * vcmax25top if (iv == 1) then laican = 0.5_r8 * tlai_z(p,iv) else laican = laican + 0.5_r8 * (tlai_z(p,iv-1)+tlai_z(p,iv)) end if vcmaxse = 668.39_r8 - 1.07_r8 * min(max((t10(p)-tfrz),11._r8),35._r8) jmaxse = 659.70_r8 - 0.75_r8 * min(max((t10(p)-tfrz),11._r8),35._r8) 2. Double checked the subroutines called in subroutine Photosynthesis, such as hybrid, ci_func, quadratic, and brent, in compute_rc.F90. 3. Found a difference in compute_rc.F90.clm4VCMAX.tileid of CLM4.5 and CLM4 compute_rc.F90: compute_rc.F90.clm4VCMAX.tileid of CLM4.5: lnc(p) = 1._r8 / (slatop(ivt) * leafcn(ivt)) CLM4 compute_rc.F90: lnc = 1. / (sla(n,nv) * leafcn(ivt)) lnc is divided by different sla. Try to also use sla in compute_rc.F90.clm4VCMAX for CLM4.5 next. Next: Subroutine Photosynthesis: use lnccan to distinguish from lnc (which is for canopy top) Subroutine compute_rc: compute slasun and slasha ========== 20190107: 1. Read the "Stomatal Resistance and Photosynthesis" chapter of CLM4.5 Tech Note and checked if the Catchment-CN with CLM4.5 model code is consistent with the Tech Note. Some differences are found: (1) jmax25: The difference can not explain the low GPP though. Tech Note: jmax25 = 1.97*vcmax25 Code: jmax25top = (2.59_r8 - 0.035_r8*min(max((t10(p)-tfrz),11._r8),35._r8)) * vcmax25top (2) Temperature dependence parameters for C3 photosynthesis: Ha, Hd and se are all larger in the code than in the Tech Note for vcmax, jmax and tpu. What's the effect? Tech Note: see Table 8.2 Code: kcha = 79430._r8 koha = 36380._r8 cpha = 37830._r8 !KO vcmaxha = 65330._r8 !KO jmaxha = 43540._r8 !KO tpuha = 53100._r8 !KO vcmaxha = 72000._r8 jmaxha = 50000._r8 tpuha = 72000._r8 !KO lmrha = 46390._r8 ! High temperature deactivation, from: ! Leuning (2002) Plant, Cell and Environment 25:1205-1210 ! The factor "c" scales the deactivation to a value of 1.0 at 25C !KO vcmaxhd = 149250._r8 !KO jmaxhd = 152040._r8 !KO tpuhd = 150650._r8 !KO vcmaxhd = 200000._r8 jmaxhd = 200000._r8 tpuhd = 200000._r8 !KO lmrhd = 150650._r8 !KO vcmaxse = 485._r8 !KO jmaxse = 495._r8 !KO tpuse = 490._r8 lmrse = 490._r8 !KO vcmaxc = fth25 (vcmaxhd, vcmaxse) !KO jmaxc = fth25 (jmaxhd, jmaxse) !KO tpuc = fth25 (tpuhd, tpuse) lmrc = fth25 (lmrhd, lmrse) !KO vcmaxse = 668.39_r8 - 1.07_r8 * min(max((t10(p)-tfrz),11._r8),35._r8) jmaxse = 659.70_r8 - 0.75_r8 * min(max((t10(p)-tfrz),11._r8),35._r8) tpuse = vcmaxse vcmaxc = fth25 (vcmaxhd, vcmaxse) jmaxc = fth25 (jmaxhd, jmaxse) tpuc = fth25 (tpuhd, tpuse) !KO 2. Changed the code in subroutine Photosynthesis so that the parameter values are consistent with the Tech Note. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake install BOPT=g > cd ../../.. > /bin/cp -pr Linux/ exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98468942 2019-01-07 15:03 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i Both psnsun and psnsha increased for the point in SE China after using the temperature dependence parameters from the Tech Note. Ran clm4.5_point for a few days. Also ran clm4_point for a few days. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4/clm4_point/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 94094447 2019-01-02 14:02 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i Compared the GPP. GPP is higher in the new CLM4.5 compared to the old CLM4.5 for the point in SE China. 3. Following 2 above, do CLM4 and CLM4.5 point runs for a point in the Amazon. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > nedit driver_inputs_0.5D_point.nml & minlon = -75.5 ! min longitude maxlon = -75 ! max longitude minlat = -5.5 ! min latitude maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4/clm4_point/run > nedit driver_inputs_0.5D_point.nml & minlon = -75.5 ! min longitude maxlon = -75 ! max longitude minlat = -5.5 ! min latitude maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i GPP decreased at this point though. Actually, it says in the Tech Note below Table 8.2 that "The parameters in Table 8.2 do not allow for temperature acclimation of photosynthesis. In the model, acclimation is implemented as in Kattge and Knorr (2007). In this parameterization, maxcV and maxJ vary with the plant growth temperature. This is achieved by allowing SΔto vary with growth temperature according to ...". The above test is not needed at all. Next: Continue to compare Tech Note and code. Think about tuning the parameters. ========== 20190108: 1. Checked the t10 and t_veg of the Amazon point. The domain has been changed to the Amazon point yesterday. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4/clm4_point/run > /bin/rm ../output/point/rc_out/* > ./lenkf.0.i t10 and t_veg look correct in this Amazon point. 2. Compared global patterns and values of several variables between CLM4 and CLM4.5: sfmc, rzmc, tsurf, tc1, tc2, tc4, tg1, tg2, tg4 All look similar except sfmc and rzmc. sfmc and rzmc are much lower in CLM4.5. Why? 3. Continued to compare Tech Note and code. (1) On page 191 of the Tech Note, it says "Tropical broadleaf evergreen trees are an exception, and a higher max25cV is used to alleviate model biases (Bonan et al. 2012).". Is the code consistent with it? Do point runs to find out. Since changing the temperature dependence parameters of photosynthesis to match Table 8.2 in the Tech Note is inappropriate, restored compute_rc.F90 and recompiled with BOPT=g. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > /bin/cp -pr Linux/ exec/clm4.5_debug/. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98469182 2019-01-08 12:03 ../build/Linux/bin/LDASsaCN_mpi.x* Yes. The code is consistent with the Tech Note. Next: Continue to compare Tech Note and code. Think about tuning the parameters. ========== 20190110: 1. P191 in CLM4.5 Tech Note says "FLNR was chosen to give Vcmax25 consistent with Kattge et al. (2009), as discussed by Bonan et al. (2011, 2012). Table 8.1 lists derived values for Vcmax25 at the top of the canopy using SLA0. Tropical broadleaf evergreen trees are an exception, and a higher Vcmax25 is used to alleviate model biases (Bonan et al. 2012)." flnr: fraction of leaf N in the Rubisco enzyme (gN Rubisco / gN leaf) ! pftname / 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 / CLM4: data flnx / 0., 0.050, 0.040, 0.080, 0.060, 0.060, 0.090, 0.090, 0.090, 0.060, 0.090, 0.090, 0.090, 0.090, 0.090, 0.090, 0.090, 0.090, 0.100, 0.100/ CLM4.5: data flnx / 0.,0.0509,0.0466,0.0546,0.0461,0.0515,0.0716,0.1007,0.1007,0.0517,0.0943,0.0943,0.0943,0.1365,0.1365,0.1365,0.0900,0.0900,0.1758,0.1758/ Can I tune flnx to improve GPP in CLM4.5? 2. P192 in CLM4.5 Tech Note says "The value Kn=0.11 chosen by Bonan et al. (2011) is consistent with observationally-derived estimates for forests, mostly tropical, and provides a gradient in Vcmax similar to the original CLM4 specific leaf area scaling. However, Bonan et al. (2012) showed that the sunlit/shaded canopy parameterization does not match an explicit multi-layer canopy parameterization. The discrepancy arises from absorption of scattered radiation by shaded leaves and can be tuned out with higher Kn. The model uses Kn=0.30 to match an explicit multi-layer canopy.". It seems that Kn=0.30 works better for a multi-layer canopy. What about two-leaf canopy? In the abstract of Bonan et al. (2012), it says "Larger Kn decreases shaded leaf photosynthesis". Should we use lower Kn for our two-leaf canopy? ========== 20190111: 1. p37 in CLM4.5 Tech Note, K = G(u)/u is the optical depth of direct beam per unit leaf and stem area. It's extkb in subroutine compute_rc. extkn = 0.30 (nitrogen allocation coefficient) in subroutin compute_rc is the Kn (decay coefficient for nitrogen) on p192 of CLM4.5 Tech Note, and is the kn for multi-layer canopy in subroutine photosynthesis (see Bonan et al. 2012). ========== 20190114: 1. Continued reading Bonan et al., 2012 to figure out why our GPP is so low. ========== 20190115: 1. Continued reading Bonan et al., 2012 to figure out why our GPP is so low. 2. Wrote a script and used it to compare the spatial patterns and zonal mean of GPP from Catchment-CN with CLM4.5 to those of the reference GPP dataset. 3. Tried changing extkn from 0.30 back to 0.11 in compute_rc.F90 to see if this would improve GPP in Catchment-CN with CLM4.5. ! real, parameter :: extkn = 0.30 ! nitrogen allocation coefficient real, parameter :: extkn = 0.11 ! nitrogen allocation coefficient, fzeng changed it back to 0.11 (see Bonan et al., 2012) on 1/15/2019 Compiled the code with BOPT=g. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > /bin/cp -pr Linux/ exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98469182 2019-01-15 17:16 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/point/rc_out/* > ./lenkf.0.j ========== 20190116: 1. Analyzed the point run output using ~/Catchment/CLM4.5/debug/plot_ts.gs. Using Kn=0.11 instead of Kn=0.30 increased GPP in the Amazon point. 2. With Kn=0.11 in compute_rc.F90, re-compiled the code without BOPT=g: "gmake clean" and "gmake install". > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec > mv clm4.5 clm4.5_Kn30 > mkdir clm4.5_Kn11 > cd .. > /bin/cp -pr Linux/ exec/clm4.5_Kn11/. > ls -l exec/clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-01-16 10:20 exec/clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x* > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-01-16 10:20 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_11/output > cd ../../run/ > ./ldas_batchrun.j ========== 20190118: 1. Analyzed the clm4.5_DE720_12 output: > cd ~/matlab/clm4-to-clm4.5 > matlab -nodisplay < compare_ts.m ========== 20190122: 1. Read Kattge and Knorr 2007 about temperature acclimation of photosynthesis and compared the parameterization to Catchment-CN with CLM4.5. 2. Checked the temperature acclimation of photosynthesis in Bonan et al., 2012. 3. Analyzed the clm4.5_DE720_13 output: > cd ~/matlab/clm4-to-clm4.5 > matlab -nodisplay < compare_ts.m 4. Analyzed the output of the 27th, 28th and 29th cycles of clm4.5_Princeton_DE720: > cd ~/matlab/clm4-to-clm4.5 > matlab -nodisplay < compare_ts.m 5. Also continued the clm4.5_Princeton_DE720 spinup run with extkn = 0.11. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-01-16 10:20 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_29/output > cd ../../run/ > ./ldas_batchrun.j ========== 20190123: 1. Analyzed the clm4.5_DE720_14 output: > cd ~/matlab/clm4-to-clm4.5 > matlab -nodisplay < compare_ts.m ========== 20190124: 1. Analyzed the clm4.5_Princeton_DE720_30 output: > cd ~/matlab/clm4-to-clm4.5 > matlab -nodisplay < compare_ts.m ========== 20190125: 1. Analyzed the clm4.5_DE720_15 output: > cd ~/matlab/clm4-to-clm4.5 > matlab -nodisplay < compare_ts.m ========== 20190212: 1. I have been using /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point for the experiment, which is confusing. Now make it the control of CLM4.5. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/input/restart > ls -l total 0 lrwxrwxrwx 1 fzeng g0620 90 2017-11-02 14:28 catchcn_internal_rst -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/OutData2/catchcn_internal_clm45 So the restart file is the one regridded from CLM4 M09 restart. This is not so important. It just needs to match what's used for the experiment. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn30 build Re-run clm4.5_point: > cd run/ > qsub lenkf.0.j Updated the control file: > cd ~/Catchment/CLM4.5/debug > nedit clm4.5_point.ctl & TITLE Catchment-CN offline control run, CN from CLM4.5, dtsetp=450, dtcn=5400, Kn=0.3 (default) 2. For the experiment of removing qa from denominator in ea calculation: Modified compute_rc.F90: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp compute_rc.F90.restore compute_rc.F90 > nedit compute_rc.F90 & ! ea(:) = pbot(:) * qa(:) / (0.622 + qa(:)) ! canopy air vapor pressure (Pa) ea(:) = pbot(:) * qa(:) / 0.622 ! canopy air vapor pressure (Pa), Zeng and Koster removed qa(:) from denominator > cp -p compute_rc.F90 compute_rc.F90.ea Compile: > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn30_ea > /bin/cp -pr Linux exec/clm4.5_Kn30_ea/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > cp -p ../clm4.5_point/run/D0.5_clm4.5_point.bat . > nedit D0.5_clm4.5_point.bat & job-name = clm4.5_ea_point > nedit D0.5_clm4.5_point.exe & exp_id = clm4.5_ea_point start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_ea_point > cp -p ../clm4.5_point/run/driver_inputs_0.5D_point.nml . > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn30_ea/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_clm4.5_point.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_clm4.5_point.bat --runmodel --monthsperjob 12 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ea_point/input/restart > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/OutData2/catchcn_internal_clm45 catchcn_internal_rst > /bin/rm rst_clm4.5 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ea_point/run > nedit lenkf.0.j & change "-restart_path ../input/restart/rst_clm4.5 \" to "-restart_path ../input/restart/ \" Make sure it runs (no need to do "interactive.py -A sp3 -n 16 -a g0620 -X --debug" because it only needs one node). > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_060730z. > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j 3. For the experiment of adding 0.5 to Tveg in ft equation in subroutine Photosynthesis: Modified compute_rc.F90: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp compute_rc.F90.restore compute_rc.F90 > nedit compute_rc.F90 & L747-748: ! ft(tl,ha) = exp( ha / (rgas*1.e-3_r8*(tfrz+25._r8)) * (1._r8 - (tfrz+25._r8)/tl) ) ft(tl,ha) = exp( ha / (rgas*1.e-3_r8*(tfrz+25._r8)) * (1._r8 - (tfrz+25._r8)/(tl+0.5)) ) ! fzeng and rkoster, 12 Feb 2019 > cp -p compute_rc.F90 compute_rc.F90.ft Compile: > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn30_ft > /bin/cp -pr Linux exec/clm4.5_Kn30_ft/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_clm4.5_point.bat & job-name = clm4.5_ft_point > nedit D0.5_clm4.5_point.exe & exp_id = clm4.5_ft_point restart_id = clm4.5_ft_point > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn30_ft/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_clm4.5_point.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_clm4.5_point.bat --runmodel --monthsperjob 12 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ft_point/input/restart > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/OutData2/catchcn_internal_clm45 catchcn_internal_rst > /bin/rm rst_clm4.5 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ft_point/run > nedit lenkf.0.j & change "-restart_path ../input/restart/rst_clm4.5 \" to "-restart_path ../input/restart/ \" Make sure it runs (no need to do "interactive.py -A sp3 -n 16 -a g0620 -X --debug" because it only needs one node). > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_040730z. > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j 4. Analyzed the output using ~/Catchment/CLM4.5/debug/plot_ts.gs 5. Did the point runs at another location, this time in China (was in Amazon). Amazon site: /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/output/point/rc_out > cat *domain* 1 11488 11488 Same for clm4.5_ea_point and clm4.5_ft_point. This is just to double check the simulations were done at the same site. Now for the site in China: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > nedit driver_inputs_0.5D_point.nml & minlon = 117 ! min longitude maxlon = 117.5 ! max longitude minlat = 27 ! min latitude maxlat = 27.5 ! max latitude !minlon = -75.5 ! min longitude !maxlon = -75 ! max longitude !minlat = -5.5 ! min latitude !maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ea_point/run > nedit driver_inputs_0.5D_point.nml & minlon = 117 ! min longitude maxlon = 117.5 ! max longitude minlat = 27 ! min latitude maxlat = 27.5 ! max latitude !minlon = -75.5 ! min longitude !maxlon = -75 ! max longitude !minlat = -5.5 ! min latitude !maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ft_point/run > nedit driver_inputs_0.5D_point.nml & minlon = 117 ! min longitude maxlon = 117.5 ! max longitude minlat = 27 ! min latitude maxlat = 27.5 ! max latitude !minlon = -75.5 ! min longitude !maxlon = -75 ! max longitude !minlat = -5.5 ! min latitude !maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point > cat output/point/rc_out/*domain* 1 26216 26216 Same for clm4.5_ea_point and clm4.5_ft_point. Good! ========== 20190213: 1. Analyzed the China point run output using ~/Catchment/CLM4.5/debug/plot_ts.gs ========== 20190214: 1. For the experiment of adding 1.0 to Tveg in whenever Tveg is used in compute_rc.F90: Checked compute_rc.F90 and it's the same as compute_rc.F90.restore (the default). Good! > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSlana_GridComp > cp -p process_cn.F90 process_cn.F90.restore > nedit process_cn.F90 & [Added this line below right before calling compute_rc] TCx = TCx + 1. ! fzeng and rkoster, assume that our TCx is 1 degree lower and correct this bias, 14 Feb 2019 > cp -p process_cn.F90 process_cn.F90.tc Compile: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install [This is needed because I restored computed_rc.F90 after the ft experiments.] > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn30_tc > /bin/cp -pr Linux exec/clm4.5_Kn30_tc/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_clm4.5_point.bat & job-name = clm4.5_tc_point > nedit D0.5_clm4.5_point.exe & exp_id = clm4.5_tc_point restart_id = clm4.5_tc_point > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn30_tc/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_clm4.5_point.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_clm4.5_point.bat --runmodel --monthsperjob 12 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_tc_point/input/restart > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/OutData2/catchcn_internal_clm45 catchcn_internal_rst > /bin/rm rst_clm4.5 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_tc_point/run > nedit lenkf.0.j & change "-restart_path ../input/restart/rst_clm4.5 \" to "-restart_path ../input/restart/ \" > cp ../../clm4.5_point/run/driver_inputs_0.5D_point.nml . Make sure it runs (no need to do "interactive.py -A sp3 -n 16 -a g0620 -X --debug" because it only needs one node). > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_040730z. > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_ft_point/output/point/rc_out > cat *domain* 1 26216 26216 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 > cat clm4.5_point/output/point/rc_out/*domain* 1 26216 26216 Good! ========== 20190219: 1. Analyzed the China point run output using ~/Catchment/CLM4.5/debug/plot_ts.gs 2. Conducted the tc point runs at the Amazon site. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_point/run > nedit driver_inputs_0.5D_point.nml & !minlon = 117 ! min longitude !maxlon = 117.5 ! max longitude !minlat = 27 ! min latitude !maxlat = 27.5 ! max latitude minlon = -75.5 ! min longitude maxlon = -75 ! max longitude minlat = -5.5 ! min latitude maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > qsub lenkf.0.j > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_tc_point/run > nedit driver_inputs_0.5D_point.nml & !minlon = 117 ! min longitude !maxlon = 117.5 ! max longitude !minlat = 27 ! min latitude !maxlat = 27.5 ! max latitude minlon = -75.5 ! min longitude maxlon = -75 ! max longitude minlat = -5.5 ! min latitude maxlat = -5 ! max latitude > /bin/rm ../output/point/rc_out/* > nedit lenkf.0.j & #SBATCH --time=02:00:00 > qsub lenkf.0.j > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_clm4.5_point.bat & walltime = 02:00:00 3. The clm4.5_DE720 stopped on Saturday 2019-02-16. Checked why and fixed the issue. It stopped in simulating date_time_new 20020405_230730z due to time out. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/output/global > ls -l rs/ens0000/Y2002/M01 total 296800 -rw-r--r-- 1 fzeng g0620 303909763 2019-02-16 06:49 clm4.5_DE720.ens0000.catchcn_ldas_rst.20020101_0000z This was the last restart file written out. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > cp lenkf.2.j lenkf.2b.j > nedit lenkf.2b.j & -start_year 2002 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ > cp ldas_batchrun.j ldas_batchrun_b.j > nedit ldas_batchrun_b.j & #!/bin/bash -f jobid0=$(echo $(sbatch lenkf.2b.j) | cut -d' ' -f 4) echo $jobid0 jobid1=$(echo $(sbatch --dependency=afterok:$jobid0 lenkf.3.j) | cut -d' ' -f 4) echo $jobid1 jobid2=$(echo $(sbatch --dependency=afterok:$jobid1 lenkf.4.j) | cut -d' ' -f 4) echo $jobid2 > ./ldas_batchrun_b.j 4. Processed spinup output: clm4.5_DE720: 25,26,27 clm4.5_Princeton_DE720: 47,48,49 Stopped the spinup when the current cycles are done. ========== 20190307: 1. Set down regulation to 0 to write out potential GPP and see if it matches the numbers in Bonan et al., 2012. Modified the model code: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNAllocationMod.F90 & [L1379-1380] downreg(p) = excess_cflux(p)/gpp(p) downreg(p) = 0. ! fzeng and rkoster: force downreg to zero to write out potential GPP, 7 Mar 2019 > cp compute_rc.F90.Kn11 compute_rc.F90 [use Kn=0.11] Compiled: > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > xdiff process_cn.F90 process_cn.F90.restore & [Identical. Good!] > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir exec/clm4.5_Kn11_NOdownreg > /bin/cp -pr Linux/ exec/clm4.5_Kn11_NOdownreg/. > ls -l exec/clm4.5_Kn11_NOdownreg/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-03-07 13:53 exec/clm4.5_Kn11_NOdownreg/Linux/bin/LDASsaCN_mpi.x* Spin up: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > ls -l build lrwxrwxrwx 1 fzeng g0620 104 2019-02-20 08:11 build -> /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11/ > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_NOdownreg build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-03-07 13:53 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_28/output > cd ../../run/ > ./ldas_batchrun.j 2. Restored CNAllocationMod.F90 and compute_rc.F90: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp CNAllocationMod.F90.restore CNAllocationMod.F90 > cp compute_rc.F90.restore compute_rc.F90 3. What I should have done in 1 above is for the spinup using Princeton forcing (i.e. clm4.5_Princeton_DE720), not the spinup using MERRA2 forcing (i.e. clm4.5_DE720), so cancelled all the clm4.5_DE720 jobs submitted. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-01-16 10:20 build/Linux/bin/LDASsaCN_mpi.x* Submitted the clm4.5_Princeton_DE720 jobs: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > ls -l build lrwxrwxrwx 1 fzeng g0620 104 2019-02-20 10:53 build -> /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11/ > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_NOdownreg build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-03-07 13:53 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_56/output > cd ../../run/ > ./ldas_batchrun.j ========== 20190313: 1. The clm4.5_Princeton_DE720 set up on 20190307 to write out potential GPP crashed over the weekend when it was simulating date_time_new 20050409_150730z. The error message is: column cbalance error = 0.1083146 1249 Latdeg,Londeg= -0.2500000 38.25000 begcb = 142639.2 endcb = 142640.1 delta store = 0.8906250 I noticed this during the weekend, but didn't have time to investigate the issue until today. This happens in CNBalanceCheckMod.F90 when (col_cinputs(c) - col_coutputs(c))*dt - (col_endcb(c) - col_begcb(c)) > 0.1 ========== 20190321: 1. Modified CN_DriverMod.F90 and process_cn.F90 to write out potential GPP. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > xdiff CNAllocationMod.F90 CNAllocationMod.F90.restore & [Nothing returned -> they are identical. Great!] > cp compute_rc.F90.Kn11 compute_rc.F90 [use Kn=0.11] > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_gpp_pot > /bin/cp -pr Linux/ exec/clm4.5_Kn11_gpp_pot/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > ls -l build lrwxrwxrwx 1 fzeng g0620 114 2019-03-07 15:56 build -> /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_NOdownreg/ > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_gpp_pot build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71558063 2019-03-21 13:37 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-03-07 15:57 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_56/output/ [Good!] > cd ../../run/ > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_133000z. > ./ldas_batchrun.j Restored process_cn.F90, CN_DriverMod.F90 and compute_rc.F90. 2. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec > rm -rf clm4.5_Kn11_NOdownreg ========== 20190325: 1. Calculated the global potential GPP using ~/python/clm4-to-clm4.5/compare_ts.py. What I got is 216 Pg C/yr which is much larger than the values in Bonan et al., 2012. Why? Emailed Justin asking if there is output of the original CLM4.5CN that I can compared with. He is on leave until Mar 28. ========== 20190326: 1. Read CLM5 Tech Note about photosynthesis and Vcmax. 2. Thought about if we can change Vcmax in Catchment-CN4.5 to improve GPP. ========== 20190327: 1. Checked the CNFireMod.F90 code, comparing to the original CLM4.5, and don't see anything wrong. ========== 20190328: 1. Read the CNFireMod.F90 code and Fang Li's papers about the fire model. ========== 20190401: 1. Drafted email to Dave and sent the draft to Randy for his comments: Corrected fire model code About my trip in May ========== 20190402: 1. Read Thornton and Zimmermann 2007 about CLM4 Vcmax. 2. Checked the GPP in the outputs of the original CLM4.5 (see Dave Lawrence's email on 20190401). ========== 20190403: 1. Continued checking the GPP in the outputs of the original CLM4.5. Emailed Danica 2. Checked the nitrogen deposition data used for Catchment-CN4.5. https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata/lnd/clm2/ndepdata/ndep_clm_simyr2000_0.23x0.31_c091106.nc This is the one obtained by Greg and has been used for Catchment-CN4.0. It has the finest spatial resolution and is the latest one that has this resolution. It's also used for Catchment-CN4.5. 3. Read Charlie Koven's presentation slides about CLM5 carbon cycle. 4. Contacted Ben Paulter. He suggested me to compare our N stocks, N mineralization and N fixation to the literature. Read Smith et al., 2014 and Zaehle and Friend, 2010. ========== 20190404: 1. Created ~/python/clm4-to-clm4.5/tile2grid_ndep.py and used it to check Catchment-CN NDEP for the CANIF sites in Zaehle and Friend, 2010. 2. Need to modify our model code to write out N stocks, N mineralization and N fixation and do a simulation. Checked the model code to see what's available about N stocks, N mineralization and N fixation. CNAllocationMod.F90: real(r8), pointer :: sminn_vr(:,:) ! (gN/m3) soil mineral N real(r8), pointer :: plant_nalloc(:) ! total allocated N flux (gN/m2/s) real(r8), pointer :: plant_calloc(:) ! total allocated C flux (gC/m2/s) real(r8), pointer :: sminn_to_npool(:) ! deployment of soil mineral N uptake (gN/m2/s) real(r8), pointer :: plant_ndemand(:) ! N flux required to support initial GPP (gN/m2/s) CNNStateUpdate1Mod.F90: nfix_to_sminn(:) fert_to_sminn(:) NOTE: only for crop_prog = .true. soyfixn_to_sminn(:) NOTE: only for crop_prog = .true. sminn_to_npool(:) ! deployment of soil mineral N uptake (gN/m2/s) sminn_vr(:,:) ! (gN/m3) soil mineral N gross_nmin_vr(:,:) ! (gN/m3/s) CNSummaryMod.F90: This is the one I need! real(r8), pointer :: sminn_leached(:) ! total N deployment (from sminn and retranslocated N pool) (NDEPLOY) ndeploy(p) = & sminn_to_npool(p) + & retransn_to_npool(p) ! total vegetation nitrogen (TOTVEGN) totvegn(p) = dispvegn(p) + storvegn(p) ! vertically integrate bulk denitrification and leaching flux do j = 1, nlevdecomp do fc = 1,num_soilc c = filter_soilc(fc) sminn_to_denit_excess(c) = sminn_to_denit_excess(c) + sminn_to_denit_excess_vr(c,j) * dzsoi_decomp(j) sminn_leached(c) = sminn_leached(c) + sminn_leached_vr(c,j) * dzsoi_decomp(j) end do end do ! total N denitrification (DENIT) do l = 1, ndecomp_cascade_transitions do fc = 1,num_soilc c = filter_soilc(fc) denit(c) = denit(c) + sminn_to_denit_decomp_cascade(c,l) end do end do do fc = 1,num_soilc c = filter_soilc(fc) denit(c) = denit(c) + sminn_to_denit_excess(c) end do According to the diagnostics package on http://webext.cgd.ucar.edu/I20TR/clm50_r270_1deg_GSWP3V1_iso_newpopd_hist/lnd/clm50_r270_1deg_GSWP3V1_iso_newpopd_hist.1995_2014-clm45_r270_1deg_GSWP3V1_hist.1995_2014/setsIndex.html (see Dave Lawlence's email on Apr 1, 2019): NDEPLOY Total N Deployed in New Growth NDEP_TO_SMINN nitrogen deposition (ndep_to_sminn(c) = forc_ndep(g), constant over time, don't write out) GROSS_NMIN Gross N Mineralization NET_NMIN Net N Mineralization NFIX_TO_SMINN nitrogen fixation DENIT Total Denitrification SMINN_LEACHED Nitrogen Leached SMINN soil mineral N COL_FIRE_NLOSS total column-level fire N loss ACTUAL_IMMOB Actual Immobilization Look for these diagnostics in CNSummaryMod.F90: only found 5 of them below ndeploy => pnf%ndeploy ! total N deployed to growth and storage (gN/m2/s) denit => cnf%denit ! total rate of denitrification (gN/m2/s) sminn_leached => cnf%sminn_leached ! soil mineral N pool loss to leaching (gN/m2/s) sminn => cns%sminn ! (gN/m2) soil mineral N col_fire_nloss => cnf%col_fire_nloss ! (gN/m2/s) total column-level fire N loss leafn => pns%leafn ! (gN/m2) leaf N CNPhenologyMod.F90: leafc => pcs%leafc ! (gC/m2) leaf C CNDecompMod.F90: gross_nmin => cnf%gross_nmin ! gross rate of N mineralization (gN/m2/s) net_nmin => cnf%net_nmin ! vert-int (diagnostic) net rate of N mineralization (gN/m2/s) CNNDynamicsMod.F90: nfix_to_sminn => cnf%nfix_to_sminn ! symbiotic/asymbiotic N fixation to soil mineral N (gN/m2/s) CNAllocationMod.F90: actual_immob => cnf%actual_immob ! vert-int (diagnostic) actual N immobilization (gN/m2/s) 3. Modified model code to add these 11 diagnostics. Follow notes in the section of "Adding diagnostics to offline code, the LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 case" in /discover/nobackup/fzeng/notes/cn.txt. Modified CN_DriverMod.F90 to aggregate the fields from column- or pft- levels to grid-level. Next: modify process_cn.F90 and others. ========== 20190405: 1. Continued modifying model code to add the new diagnostics: process_cn.F90 [follow 'closs'] clsm_ensdrv_out_routines.F90 [follow 'closs'] catch_types.F90 [update 'N_cn_diagn' and follow 'closs'] clsm_ensdrv_glob_param.F90 [change 'N_CatchCN_OutFields = 37' to 'N_CatchCN_OutFields = 48'] Used kn = 0.11 in compute_rc.F90. After adding a new diagnostic, the model should be rebuilt from a clean state: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install |& tee make.install.log > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mkdir -p exec/clm4.5_Kn11_Ndiags > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags/. > ls -l exec/clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70993956 2019-04-05 14:19 exec/clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x* 2. Run one cycle of clm4.5_Princeton_DE720 and clm4.5_DE720: (1) clm4.5_Princeton_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70993956 2019-04-05 14:19 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-03-25 11:43 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_57/output/ Good. > cd ../../run/ > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800103_150730z. > ./ldas_batchrun.j (2) clm4.5_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70993956 2019-04-05 14:19 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 80 2019-03-07 14:00 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_28/output/ Good. > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_070730z. > ./ldas_batchrun.j 3. Created ~/python/clm4-to-clm4.5/extract_DE720_data_for_sites.py ========== 20190408: 1. Both clm4.5_Princeton_DE720 and clm4.5_DE720 crashed on Friday night when writing out the first monthly output file: forrtl: error (65): floating invalid Image PC Routine Line Source libirc.so 00002AAAAB590961 Unknown Unknown Unknown libirc.so 00002AAAAB58F0B7 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E3362 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E31B6 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2D1A0C Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2B2E82 Unknown Unknown Unknown libc.so.6 00002AAAABA16910 Unknown Unknown Unknown LDASsaCN_mpi.x 00000000006C7540 tile_coord_routin 2076 tile_coord.F90 LDASsaCN_mpi.x 000000000056D042 clsm_ensdrv_out_r 1629 clsm_ensdrv_out_routines.F90 LDASsaCN_mpi.x 00000000004B6676 MAIN__ 2655 cnlsm_ensdrv_main.F90 Found a bug in my modifications on Friday in CN_DriverMod.F90: denit = 0. sminn_leached = 0. sminn = 0. col_fire_nloss = 0. gross_nmin = 0. net_nmin = 0. nfix_to_sminn = 0. actual_immob = 0. should be denitg = 0. sminn_leachedg = 0. sminng = 0. col_fire_nlossg = 0. gross_nming = 0. net_nming = 0. nfix_to_sminng = 0. actual_immobg = 0. Corrected it. Since Danica suggested a few more diagnostics to look at, add them in and try again. 2. Five diagnostics suggested by Danica. fpg fpi sminn_to_plant sminn_to_npool ndep_to_sminn CNAllocationMod.F90: fpg => cps%fpg ! fraction of potential gpp (no units) fpi => cps%fpi ! fraction of potential immobilization (no units) sminn_to_plant => cnf%sminn_to_plant ! vert-int (diagnostic) plant uptake of soil mineral N (gN/m2/s) sminn_to_npool => pnf%sminn_to_npool ! deployment of soil mineral N uptake (gN/m2/s) CNNDynamicsMod.F90 ndep_to_sminn => cnf%ndep_to_sminn ! atmospheric N deposition to soil mineral N (gN/m2/s) 3. Modified these files to add these 5 diagnostics: CN_DriverMod.F90 [follow 'gross_nmin'] process_cn.F90 [follow 'gross_nmin'] clsm_ensdrv_out_routines.F90 [follow 'gross_nmin'] catch_types.F90 [update 'N_cn_diagn' and follow 'gross_nmin'] clsm_ensdrv_glob_param.F90 [change 'N_CatchCN_OutFields = 48' to 'N_CatchCN_OutFields = 53'] 4. Compiled. Used kn = 0.11 in compute_rc.F90. After adding a new diagnostic, the model should be rebuilt from a clean state: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install |& tee make.install.log > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mkdir -p exec/clm4.5_Kn11_Ndiags [Skipped this which was already done on Friday last week.] > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags/. > ls -l exec/clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70989860 2019-04-08 12:01 exec/clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x* 5. Run one cycle of clm4.5_Princeton_DE720 and clm4.5_DE720: (1) clm4.5_Princeton_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70989860 2019-04-08 12:01 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-03-25 11:43 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_57/output/ Good. > cd ../../run/ > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800205_210730z. > /bin/rm ../output/global/rc_out/Y1980/M01/* > ./ldas_batchrun.j (2) clm4.5_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 70989860 2019-04-08 12:01 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 80 2019-03-07 14:00 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_28/output/ Good. > cd ../../run/ > /bin/rm ../output/global/rc_out/Y1980/M01/* > ./ldas_batchrun.j 6. Modified the post-processing Fortran scripts and compiled them. Processed the 198001 clm4.5_Princeton_DE720 output. Modified the GrADS control scripts. Checked the 198001 clm4.5_Princeton_DE720 output fields on GrADS and they look correct. ========== 20190409: 1. Improved ~/python/clm4-to-clm4.5/tile2grid_ndep.py and used it to plot ndep comparison between Catchment-CN and CANIF sites. 2. Improved ~/python/clm4-to-clm4.5/extract_DE720_data_for_sites.py and renamed it to compare_CatCN_sites.py. The improved script can extract multiple fields for multiple sites simultaneousely. 3. Processed available output: ~/Catchment/CLM4.5 > grid_restore_hdeg_Princeton clm4.5_Princeton_DE720 ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_DE720 Checked the output on GrADS. Looks correct. ========== 20190410: 1. Extracted site-level NPP, Nup and Nmin from clm4.5_Princeton_DE720_58 and clm4.5_DE720_29 and compared to the CANIF site obs. 2. Created ~/python/clm4-to-clm4.5/plot_CatCN_fields.py and used it to plot global maps of FPG for clm4.5_Princeton_DE720_58 and clm4.5_DE720_29. 3. Improved ~/python/clm4-to-clm4.5/compare_ts.py ========== 20190412: 1. Read CNAllocationMod.F90 to understand down regulation of GPP due to N limitation. 2. Emailed comparisons of Catchment-CN4.5 to CANIF site observations and NCAR CLM4.5 outputs to Ben for his suggestions. ========== 20190415: 1. Made more plots comparing N fixation and FPI between Catchment-CN4.5 and NCAR CLM4.5. Sent the plots to Ben. Ben's reply: Hi Fanwei, It seems you are spot on for N fixation! Last suggestion before you dive into parameters would be to check Table 1 of Zaehle et al in the following: https://pdfs.semanticscholar.org/4750/b9db88de29cd78abb90f3b06e272ace26279.pdf It seems the N mineralization is a problem. It might be that your N soil stocks are ok, but the soil temperature or moisture is too low for mineralization to happen. Does CANIF provide total soil N stocks? Or, perhaps you could increase precipitation or air temperature by 50% for these sites artificially and see if mineralization picks up. Ben ========== 20190416: 1. Made plots of global N leaching loss, plant N uptake, gross N mineralization and net N mineralization and compared to Zaehle 2013 and the CLM4.5CN output. Emailed Ben the plots. ========== 20190418: 1. N leaching is very small. Checked my scripts about global N leaching loss calculation and checked the model code about the unit of N leaching loss. I did not see any mistake. 2. Plotted SMINN and found it also very small. 3. Did some literature search about N cycle and compared our results to the literature. 4. Met with Eunjee to discuss what I'd found about N cycle so far. ========== 20190419: 1. Plotted GPP/GPPpot ratio for Catchment-CN4.5 and Catchment-CN4.0, and for MERRA-2 forcingg and Princeton forcing, and compared to literature. 2. Added these 5 diagnostics to the output of Catchment-CN4.5. totvegn totlitn totsomn retransn retransn_to_npool CNSummaryMod.F90: totvegn => pns%totvegn ! (gN/m2) total vegetation nitrogen totlitn => cns%totlitn ! (gN/m2) total litter nitrogen totsomn => cns%totsomn ! (gN/m2) total soil organic matter nitrogen retransn => pns%retransn ! (gN/m2) plant pool of retranslocated N retransn_to_npool => pnf%retransn_to_npool ! deployment of retranslocated N (gN/m2/s) 3. Modified these files to add these 5 diagnostics: CN_DriverMod.F90 [follow 'gross_nmin'] process_cn.F90 [follow 'gross_nmin'] clsm_ensdrv_out_routines.F90 [follow 'gross_nmin'] catch_types.F90 [update 'N_cn_diagn' from 48 to 53 and follow 'gross_nmin'] clsm_ensdrv_glob_param.F90 [change 'N_CatchCN_OutFields = 53' to 'N_CatchCN_OutFields = 58'] 4. Compiled. Used kn = 0.11 in compute_rc.F90. After adding a new diagnostic, the model should be rebuilt from a clean state: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install |& tee make.install.log /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec > mv clm4.5_Kn11_Ndiags clm4.5_Kn11_16Ndiags > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mkdir -p exec/clm4.5_Kn11_Ndiags [Skipped this which was already done on Friday last week.] > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags/. > ls -l exec/clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71022628 2019-04-19 12:40 exec/clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x* 5. Run one cycle of clm4.5_Princeton_DE720 and clm4.5_DE720: (1) clm4.5_Princeton_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71022628 2019-04-19 12:40 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_58/output > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-04-19 13:31 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_58/output/ > cd ../../run/ > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800201_232230z after one month of simulation. > ./ldas_batchrun.j (2) clm4.5_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71022628 2019-04-19 12:40 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_29/output > ls -l lrwxrwxrwx 1 fzeng g0620 80 2019-04-19 13:38 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_29/output > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800201_060730z. > ./ldas_batchrun.j 5. Updated and re-compiled the post-processing scripts grid_restore_hdeg.f90 and grid_restore_hdeg_Princeton.f90 in ~/Catchment/CLM4.5. Processed the available output. 6. Updated clm4.5_DE720.ctl and clm4.5_Princeton_DE720.ctl in ~/Catchment/CLM4.5 and checked the output on GrADS. Note: variable name can't be too long in GrADS control files. ========== 20190422: 1. Modified CN_DriverMod.F90: use 10 KgN/day as N deposition so that N supply is unlimited. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mkdir -p exec/clm4.5_Kn11_Ndiags_unlimitedN > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags_unlimitedN/. > ls -l exec/clm4.5_Kn11_Ndiags_unlimitedN/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71022628 2019-04-22 15:34 exec/clm4.5_Kn11_Ndiags_unlimitedN/Linux/bin/LDASsaCN_mpi.x* 2. Run one cycle of clm4.5_Princeton_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_unlimitedN build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71022628 2019-04-22 15:34 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_59/output > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-04-22 15:38 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_59/output/ > cd ../../run/ > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800203_113730z after one month of simulation. > qsub lenkf.0.j 3. Restored CN_DriverMod.F90: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp -p CN_DriverMod.F90 CN_DriverMod.F90.Ndiags.unlimitedN > cp -p CN_DriverMod.F90.Ndiags CN_DriverMod.F90 > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install ========== 20190423: 1. The clm4.5_Princeton_DE720 run finished 8 years of simulation. Processed the output: > cd ~/Catchment/CLM4.5 > grid_restore_hdeg_Princeton clm4.5_Princeton_DE720 Plotted GPP/GPPpot ratio using plot_CatCN_GPP_to_GPPpot.py. This ratio is 1 everywhere when N supply is unlimited. Renamed the experiment dir for possible future analysis: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 > mv clm4.5_Princeton_DE720 clm4.5_Princeton_DE720_unlimitedN Created a new experiment dir following ~/Catchment/CLM4.5/submit_next_batch_Princeton_DE720: > set exp = clm4.5_Princeton_DE720 > set num = 59 > cd $NOBACKUP/clm4-to-clm4.5/LDAS/tests/clm4.5 > /bin/cp -r ${exp}_template ${exp} > cd ${exp}/input/restart > ln -s $NOBACKUP/clm4-to-clm4.5/LDAS/tests/clm4.5/${exp}_${num}/output > /bin/rm rst_clm4.5 2. For Ben's suggestion about increasing precipitation and see how N mineralization responds: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSlana_GridComp > cp -p clsm_ensdrv_force_routines.F90 clsm_ensdrv_force_routines.F90.restore > nedit clsm_ensdrv_force_routines.F90 & In subroutine get_forcing, added to L380-385: ! Increase precipitation by 50% to see the response of N mineralization ! fzeng, 23 Apr 2019 met_force_obs_tile_new(:)%Rainf = met_force_obs_tile_new(:)%Rainf * 1.5 met_force_obs_tile_new(:)%Rainf_C = met_force_obs_tile_new(:)%Rainf_C * 1.5 met_force_obs_tile_new(:)%Snowf = met_force_obs_tile_new(:)%Snowf * 1.5 if(master_proc) print *, 'Increased precipitation by 50%' > cp -p clsm_ensdrv_force_routines.F90 clsm_ensdrv_force_routines.F90.PPT50pctHigher > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mkdir -p exec/clm4.5_Kn11_Ndiags_PPT50pctHigher > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags_PPT50pctHigher/. > ls -l exec/clm4.5_Kn11_Ndiags_PPT50pctHigher/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71027376 2019-04-23 14:33 exec/clm4.5_Kn11_Ndiags_PPT50pctHigher/Linux/bin/LDASsaCN_mpi.x* 3. Run one cycle of clm4.5_Princeton_DE720: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_PPT50pctHigher build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71027376 2019-04-23 14:33 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-04-23 10:45 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_59/output/ > cd ../../run/ > cp lenkf.0.j lenkf.0a.j > nedit lenkf.0a.j & -end_year 1981 -end_month 1 -end_day 1 -end_hour 0 -end_min 0 -end_sec 0 \ [Just run for one year.] > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0a.j Processed the output and checked precipitation in GrADS. Looks correct. > cd ~/Catchment/CLM4.5 > grid_restore_hdeg_Princeton clm4.5_Princeton_DE720 4. For Ben's suggestion about increasing temperature and see how N mineralization responds: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSlana_GridComp > nedit clsm_ensdrv_force_routines.F90 & In subroutine get_forcing, added to L380-383: ! Increase Tair by 3 degrees to see the response of N mineralization ! fzeng, 23 Apr 2019 met_force_obs_tile_new(:)%Tair = met_force_obs_tile_new(:)%Tair + 3 if(master_proc) print *, 'Increased Tair by 3 degrees' > cp -p clsm_ensdrv_force_routines.F90 clsm_ensdrv_force_routines.F90.Tair3degHigher > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mkdir -p exec/clm4.5_Kn11_Ndiags_Tair3degHigher > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags_Tair3degHigher/. > ls -l exec/clm4.5_Kn11_Ndiags_Tair3degHigher/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71027376 2019-04-23 14:51 exec/clm4.5_Kn11_Ndiags_Tair3degHigher/Linux/bin/LDASsaCN_mpi.x* Renamed the PPT experiment dir for possible future analysis: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 > mv clm4.5_Princeton_DE720 clm4.5_Princeton_DE720_PPT50pctHigher 5. Run one cycle of clm4.5_Princeton_DE720: Created a new experiment dir following ~/Catchment/CLM4.5/submit_next_batch_Princeton_DE720: > set exp = clm4.5_Princeton_DE720 > set num = 59 > cd $NOBACKUP/clm4-to-clm4.5/LDAS/tests/clm4.5 > /bin/cp -r ${exp}_template ${exp} > cd ${exp}/input/restart > ln -s $NOBACKUP/clm4-to-clm4.5/LDAS/tests/clm4.5/${exp}_${num}/output > /bin/rm rst_clm4.5 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_Tair3degHigher build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71027376 2019-04-23 14:51 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 90 2019-04-23 15:45 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720_59/output/ > cd ../../run/ > cp lenkf.0.j lenkf.0a.j > nedit lenkf.0a.j & -end_year 1981 -end_month 1 -end_day 1 -end_hour 0 -end_min 0 -end_sec 0 \ [Just run for one year.] > interactive.py -A sp3 -n 140 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ./lenkf.0a.j Processed the output and checked precipitation in GrADS. Looks correct. > cd ~/Catchment/CLM4.5 > grid_restore_hdeg_Princeton clm4.5_Princeton_DE720 Renamed the PPT experiment dir for possible future analysis: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 > mv clm4.5_Princeton_DE720 clm4.5_Princeton_DE720_Tair3degHigher 6. Analyzed the output of the PPT and Tair experiments using ~/python/clm4-to-clm4.5/compare_CatCN_sites.py ========== 20190424: 1. Read the CN code and CLM4.5 Tech Note to understand N cycle. 2. Calculated denitrification flux and found it to be quite large. Updated Ben about this. ========== 20190425: 1. Compared Catchment-CN4.5 denitrification to the literature. 2. Restored clsm_ensdrv_force_routines.F90 from the Tair3degHigher experiment > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSlana_GridComp > cp clsm_ensdrv_force_routines.F90.restore clsm_ensdrv_force_routines.F90 > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install 3. Plotted sminn, fpg and fpi in /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/OutData2/catchcn_internal_clm45 sminn is very low, even lower than sminn written out from the Catchment-CN4.5 simulations. On the other hand, fpg and fpi are very close to 1. 4. Compiled thed code with BOPT=g > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake clean > gmake install BOPT=g > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > /bin/cp -pr Linux/ exec/clm4.5_debug/. Restored: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > gmake clean > gmake install > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install > cd ../../Applications/LDAS_App/ > gmake clean > gmake install 5. Did point runs on DDT (now ARM DDT and no longer Allinea DDT): > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98880188 2019-04-25 14:43 build/Linux/bin/LDASsaCN_mpi.x* > ls -l input/restart/ lrwxrwxrwx 1 fzeng g0620 81 2018-05-23 13:36 catchcn_internal_rst -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/catchcn_internal_clm45 > cd run/ > nedit driver_inputs_0.5D_GLOBAL.nml & minlon = -75.5 ! min longitude maxlon = -75 ! max longitude minlat = -5.5 ! min latitude maxlat = -5 ! max latitude > /bin/rm ../output/global/rc_out/* > nedit lenkf.0.i & #module load tool/allinea-tools-7.0.4 module load tool/arm-forge-18.2.1 > ./lenkf.0.i ========== 20190426: 1. Read CLM4.5 Tech Note about nitrogen fixation and denitrification. 2. Emailed Melanie my question about burned area aggregation to Fang Li. Melanie will email Fang using Fang's UCAR email. ========== 20190429: 1. Drafted an email to Ben Bond-Lamberty who is the 2nd author on Fang Li's 2014 Biogeosciences paper. Sent the draft to Melanie. Melanie suggested to email Sam Levis instead. Modified the email a little bit and email Sam. Sam replied: Hi Fanwei, One suggestion is to contact Keith Oleson (CLM liaison) at NCAR. I do not work at NCAR any more and have not kept up with model versions enough to be helpful. If Keith is unable to help, then I (or Keith) could email Fang and ask her to send the corrected version, as she offered. I do not know if it would be a violation of the law to then forward you that version... Sam So I may just have to wait until next week to ask Keith about this when I am visiting NCAR. 2. Continued investigate denitrification on DDT. The values of T2M10D are correct in the /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/catchcn_internal_clm45. However, after it's read during the model run, it became 0 at the beginning of the 1st time step. Try using the restart file from the existing run instead of this very first restart file and see if T2M10D is read correctly during run time. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/input/restart > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_30/output > cd ../../run > cp -p lenkf.0.i lenkf.0.i.old > nedit lenkf.0.i & changed "-restart_path ../input/restart/ \" to "-restart_path ../input/restart/output \" > /bin/rm ../output/global/rc_out/* > ./lenkf.0.i Yes. Using the restart file from clm4.5_DE720_30, T2M10D is correct in the restart file, and is also correct after it's read during model run. ========== 20190430: 1. Plotted global map of denitrification and compared to NCAR's to see where we are too high. 2. Continued doing the point run on DDT to figure out why our denitrification is so large. Found in CNMResp that frac_veg_nosno is a huge number 2147483647 (Weiyuan said it is the largest 32 bit integer the computer can recognize, and it is questionable to use the huge number in a meaningful way in the program). This is a bug. Actually in the original CLM4.5 code: main/initSurfAlbMod.F90 99: integer , pointer :: frac_veg_nosno_alb(:) ! fraction of vegetation not covered by snow (0 OR 1) [-] 110: integer , pointer :: frac_veg_nosno(:) ! fraction of vegetation not covered by snow (0 OR 1) [-] 161: frac_veg_nosno_alb => pps%frac_veg_nosno_alb 162: frac_veg_nosno => pps%frac_veg_nosno 184: ! albedo subroutines. Note: elai, esai, frac_veg_nosno_alb are computed in 186: ! frac_veg_nosno is needed by FracWet 223: frac_veg_nosno_alb(p) = 0._r8 224: frac_veg_nosno(p) = 0._r8 316: frac_veg_nosno(p) = frac_veg_nosno_alb(p) main/CNiniSpecial.F90 145: pps%frac_veg_nosno_alb(p) = 0._r8 146: pps%frac_veg_nosno(p) = 0._r8 main/clmtypeInitMod.F90 925: allocate(pps%frac_veg_nosno(beg:end)) 926: allocate(pps%frac_veg_nosno_alb(beg:end)) 1053: pps%frac_veg_nosno(beg:end) = huge(1) 1054: pps%frac_veg_nosno_alb(beg:end) = 0 biogeochem/STATICEcosysDynMod.F90 149: integer , pointer :: frac_veg_nosno_alb(:) ! frac of vegetation not covered by snow [-] 176: frac_veg_nosno_alb => pps%frac_veg_nosno_alb 230: frac_veg_nosno_alb(p) = 1 232: frac_veg_nosno_alb(p) = 0 biogeochem/CNVegStructUpdateMod.F90 88: integer , pointer :: frac_veg_nosno_alb(:) ! frac of vegetation not covered by snow [-] 157: frac_veg_nosno_alb => pps%frac_veg_nosno_alb 313: frac_veg_nosno_alb(p) = 1 315: frac_veg_nosno_alb(p) = 0 Can initialize frac_veg_nosno in CN_Driver following CNVegStructUpdate and CLM4.5's initSurfAlbMod.F90: CLM4.5's CNiniSpecial.F90: pps%frac_veg_nosno(p) = 0._r8 CNVegStructUpdate: ! Fraction of vegetation free of snow if ((elai(p) + esai(p)) > 0._r8) then frac_veg_nosno_alb(p) = 1 else frac_veg_nosno_alb(p) = 0 end if CLM4.5's initSurfAlbMod.F90: do p = begp, endp l = plandunit(p) if (.not. lakpoi(l)) then frac_veg_nosno(p) = frac_veg_nosno_alb(p) fwet(p) = 0._r8 end if end do ========== 20190501: 1. Modified CN_DriverMod.F90 following CNVegStructUpdate and CLM4.5's initSurfAlbMod.F90. Compiled the code with BOPT=g: > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake clean > gmake install BOPT=g > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > mv exec/clm4.5_debug exec/clm4.5_debug_old > mkdir -p exec/clm4.5_debug > /bin/cp -pr Linux/ exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98884764 2019-05-01 10:45 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/global/rc_out/* > ./lenkf.0.i 2. Need to output lmrsun and lmrsha from Photosynthesis for use in CNMRestMod.F90. Modified the code: compute_rc.F90: Added lmr to the output of Photosynthesis, and lmrsun and lmrsha to the output of compute_rc. Also made other modifications. process_cn.F90: Added lmrsun and lmrsha, and average them over the DTCN period following psnsun(m) before feeding into CN_Driver. CN_DriverMod.F90: Added lmrsunm and lmrsham to input argument list. Modified the code following psnsun. catch_types.F90: Added LMRSUNM and LMRSHAM to cn_progn type. Modified the N_cn_progn accordingly. clsm_ensdrv_init_routines.F90: Added LMRSUNM and LMRSHAM following PSNSUNM and PSNSHAM. NOTE: code not compiled after the modifications yet!! 3. Need to add LMRSUNM and LMRSHAM to the restart file: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts > rm -rf 0.5D_clm4.5 > /bin/cp -pr 0.5D 0.5D_20180917 > rm -rf OutData1/ OutData2 bin catchcn_internal_clm45 catchcn_internal_rst mkLDAS.e mkLDAS.o (1) First I need to re-order the clm4.5_DE720.ens0000.catchcn_ldas_rst.20170101_0000z so that the tiles are in ascending order of tile id. Followed notes on 20180907. (1.1) Create path: cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D mkdir reordered (1.2) Create shell script "reorder_restarts" for reordering: cd reordered nedit reorder_restarts & chmod 755 reorder_restarts reorder_restarts reads: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ #!/bin/csh set YYYY = 2017 set YYYYMMDD=`echo ${YYYY}0101` mpirun -np 1 bin/mk_LDASsaRestarts -b /discover/nobackup/fzeng/bcs/Icarus-NL/Icarus-NL_Reynolds/DE_00720x00360_PE_0720x0360/ -d ${YYYYMMDD} -e clm4.5_DE720 -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_30/output/global/ -m catchcn -s 50 -r Y -t DE_00720x00360_PE_0720x0360.til ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ (1.3) Get a copy of mk_LDASsaRestarts.F90, modify and compile it: cp -p ../mk_LDASsaRestarts.F90 . nedit mk_LDASsaRestarts.F90 & [Actually no change in this case.] In SUBROUTINE reorder_LDASsa_restarts: (a) L297: tile_coord = trim(EXPDIR)//'rc_out/'//trim(expname)//'.ldas_tilecoord.bin' (b) L540: rst_file = trim(EXPDIR)//'rs/ens0000/Y'//YYYYMMDD(1:4)//'/M'//YYYYMMDD(5:6)//'/'//trim(ExpName)//& '.ens0000.catchcn_ldas_rst.'//trim(YYYYMMDD)//'_0000z' cp mk_LDASsaRestarts.F90 /discover/nobackup/fzeng/clm4-to-clm4.5/GEOS5/Heracles-5_4_p3-M3_V24_C05/src/GEOSgcs_GridComp/GEOSgcm_GridComp/GEOSagcm_GridComp/GEOSphysics_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatch_GridComp/mk_restarts/. cd /discover/nobackup/fzeng/clm4-to-clm4.5/GEOS5/Heracles-5_4_p3-M3_V24_C05/src/GEOSgcs_GridComp/GEOSgcm_GridComp/GEOSagcm_GridComp/GEOSphysics_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatch_GridComp/mk_restarts setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/GEOS5/Heracles-5_4_p3-M3_V24_C05/ source $ESMADIR/src/g5_modules gmake install (1.4) Run "reorder_restarts" to reorder the offline restart file: cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts/0.5D/reordered ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/GEOS5/Heracles-5_4_p3-M3_V24_C05/src/GEOSgcs_GridComp/GEOSgcm_GridComp/GEOSagcm_GridComp/GEOSphysics_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatch_GridComp/mk_restarts bin reorder_restarts Failed. Need more modifications. Try some other way. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/restarts > rm -rf 0.5D 4. Tried not to add lmrsunm and lmrsham to the restart file, as done in LDASsa_m3-15_2-CN. Restored catch_types.F90: > mv catch_types.F90 catch_types.F90.lmr > mv catch_types.F90.bug catch_types.F90 Restored clsm_ensdrv_init_routines.F90: > mv clsm_ensdrv_init_routines.F90 clsm_ensdrv_init_routines.F90.lmr > mv clsm_ensdrv_init_routines.F90.bug clsm_ensdrv_init_routines.F90 Modified process_cn.F90 to add lmrsunm and lmrsham following psnsunm and psnsham in process_cat.F90 in LDASsa_m3-15_2-CN. 5. Compiled the code with BOPT=g: > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake clean > gmake install BOPT=g > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > /bin/cp -pr Linux/ exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98970061 2019-05-01 20:54 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/global/rc_out/* > ./lenkf.0.i It can run now, and the information printed out looks correct. date_time_new 19800103_000000z CNMResp: 4.3827366E-02 3.716973 1.000000 1.000000 4.5170975E-05 wood: 2.214248 0.6623655 6.5148824E-06 1.9488484E-06 CNMResp: 1.4868548E-02 2.492742 1.000000 1.000000 3.0118910E-05 CNMResp: 4.3827355E-02 3.704863 1.000000 1.000000 4.5025521E-05 wood: 2.204768 0.6608238 6.4869910E-06 1.9443125E-06 CNMResp: 1.4868545E-02 2.490885 1.000000 1.000000 3.0096600E-05 CNMResp: 4.3299660E-02 0.6122209 1.000000 1.000000 7.8734565E-06 wood: 8.9914106E-02 2.6965352E-02 2.6455027E-07 7.9338953E-08 CNMResp: 1.4847703E-02 1.130326 1.000000 1.000000 1.3754677E-05 froot: 12.06429 3.5994213E-05 froot: 30.49858 9.0993526E-05 froot: 12.02626 3.5880734E-05 froot: 30.42542 9.0775218E-05 froot: 1.984255 5.9200888E-06 froot: 7.420730 2.2139990E-05 CNAlloc: 4.5976108E-06 1.9308301E-05 2.3905912E-05 4.5170975E-05 3.5994213E-05 8.1165184E-05 6.5148824E-06 1.9488484E-06 2.3905912E-05 8.9628920E-05 -6.5723012E-05 CNAlloc: 4.6112182E-06 1.6376247E-05 2.0987465E-05 3.0118910E-05 9.0993526E-05 1.2111243E-04 0.0000000E+00 0.0000000E+00 2.0987465E-05 1.2111243E-04 -1.0012497E-04 CNAlloc: 4.5972420E-06 1.9257659E-05 2.3854900E-05 4.5025521E-05 3.5880734E-05 8.0906255E-05 6.4869910E-06 1.9443125E-06 2.3854900E-05 8.9337555E-05 -6.5482658E-05 CNAlloc: 4.6106161E-06 1.6370990E-05 2.0981606E-05 3.0096600E-05 9.0775218E-05 1.2087182E-04 0.0000000E+00 0.0000000E+00 2.0981606E-05 1.2087182E-04 -9.9890211E-05 CNAlloc: 3.6180206E-06 6.3424009E-06 9.9604213E-06 7.8734565E-06 5.9200888E-06 1.3793546E-05 2.6455027E-07 7.9338953E-08 9.9604213E-06 1.4137435E-05 -4.1770136E-06 CNAlloc: 4.5527049E-06 1.3516927E-05 1.8069632E-05 1.3754677E-05 2.2139990E-05 3.5894667E-05 0.0000000E+00 0.0000000E+00 1.8069632E-05 3.5894667E-05 -1.7825036E-05 5. Restore and recompiled the code without BOPT=g: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > mv CNMRespMod.F90.restore CNMRespMod.F90 > mv CNAllocationMod.F90.restore CNAllocationMod.F90 > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake clean > gmake install > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install > cd ../../Applications/LDAS_App/ > gmake clean > gmake install > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags/. 6. Continued the clm4.5_DE720 run using the bug fixed executable: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71047468 2019-05-01 21:26 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_30/output > cd ../../run > ./ldas_batchrun.j ========== 20190502: 1. The clm4.5_DE720 run stopped at date_time_new 19870721_040730z around 9:30am today due to time out. /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > ls -l output/global/rs/ens0000/Y1987/M01 total 296800 -rw-r--r-- 1 fzeng g0620 303909763 2019-05-02 08:39 clm4.5_DE720.ens0000.catchcn_ldas_rst.19870101_0000z > cd run > cp -p lenkf.0.j lenkf.0a.j > nedit lenkf.0.j & #SBATCH --time=02:00:00 #SBATCH --output=../output/global/rc_out/Y1980/M01/clm4.5_DE720.ldas_log.19870101_0000.txt #SBATCH --error=../output/global/rc_out/Y1980/M01/clm4.5_DE720.ldas_out.19870101_0000.txt -start_year 1987 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ -end_year 1988 -end_month 1 -end_day 1 -end_hour 0 -end_min 0 -end_sec 0 \ -restart_path ../output \ > ./ldas_batchrun.j 2. Continued the clm4.5_Princeton_DE720 run using the bug fixed executable: > set exp = clm4.5_Princeton_DE720 > set num = 59 > cd $NOBACKUP/clm4-to-clm4.5/LDAS/tests/clm4.5 > /bin/cp -r ${exp}_template ${exp} > cd ${exp}/input/restart > ln -s $NOBACKUP/clm4-to-clm4.5/LDAS/tests/clm4.5/${exp}_${num}/output > /bin/rm rst_clm4.5 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71047468 2019-05-01 21:26 build/Linux/bin/LDASsaCN_mpi.x* > cd run > ./ldas_batchrun.j 3. Processed and analyzed the available output from the current clm4.5_DE720 cycle. ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_DE720 Used ~/python/clm4-to-clm4.5/compare_ts.py ========== 20190503: 1. Processed output: clm4.5_DE720: 29, 30 clm4.5_Princeton_DE720: 58, 59 2. Need to update the plots in slides 14 and 15. Scripts needed: read_fields.m, compare_GPP_site_FLUXNET2015.m and zonal_mean.m in ~/matlab/clm4-to-clm4.5 ========== 20190504: 1. Analyzed the output of clm4.5_DE720_31. ========== 20190506: 1. Modified CNDecompMod.F90 to turned on nitrificaiton and denitrification. Also modified CNFireMod.F90 to fix one bug described in Fang Li's 2014 paper. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../../ > mkdir -p exec/clm4.5_Kn11_Ndiags_nitrif_denitrif > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags_nitrif_denitrif/. 2. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 > cp clm4.5_DE720/run/D0.5_cn.exe run/. > cp clm4.5_DE720/run/D0.5.bat run/. > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_nitrif_denitrif/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_nitrif_DE720/run > nedit lenkf.0.j -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_31/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_nitrif_DE720/run > ./lenkf.0.j Crashed. forrtl: severe (174): SIGSEGV, segmentation fault occurred Image PC Routine Line Source libirc.so 00002AAAAB590961 Unknown Unknown Unknown libirc.so 00002AAAAB58F0B7 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E3362 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E31B6 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2D1A0C Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2B2BA8 Unknown Unknown Unknown libc.so.6 00002AAAABA16910 Unknown Unknown Unknown LDASsaCN_mpi.x 00000000008F4A5B cndecompmod_mp_cn 295 CNDecompMod.F90 L295 CNDecompMod.F90: phr_vr(c,j) = 0._r8 3. "gmake clean" and "gmake install BOPT=g" to recompile the code for debugging. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug Changed build. Run and got: forrtl: severe (408): fort: (2): Subscript #2 of the array PHR_VR has value 1 which is greater than the upper bound of -1 Image PC Routine Line Source libmpi_usempif08. 00002AAAAC2D2093 Unknown Unknown Unknown LDASsaCN_mpi.x 0000000001A5DEDD cndecompmod_mp_cn 295 CNDecompMod.F90 Talked to Keith. He pointed out that there may be other places I need to modify as well. 4. Modify: CNNDynamicsMod.F90 (done) 27:#ifndef NITRIF_DENITRIF 240:#ifndef NITRIF_DENITRIF 254:#ifndef NITRIF_DENITRIF 277:#ifndef NITRIF_DENITRIF 290:#ifndef NITRIF_DENITRIF 335:#ifndef NITRIF_DENITRIF CNBalanceCheckMod.F90 (done) 316:#ifndef NITRIF_DENITRIF 354:#ifndef NITRIF_DENITRIF 395:#ifndef NITRIF_DENITRIF CNSummaryMod.F90 (done) 1416:#ifndef NITRIF_DENITRIF 1562:#ifndef NITRIF_DENITRIF 1779:#ifdef NITRIF_DENITRIF 1807:#ifndef NITRIF_DENITRIF clmtypeInitMod.F90 (done) <= forgot to save a copy before making the modifications 2556:#ifdef NITRIF_DENITRIF 2581:#ifdef NITRIF_DENITRIF 2695:!#ifdef NITRIF_DENITRIF 2776:#ifdef NITRIF_DENITRIF 3040:#ifndef NITRIF_DENITRIF 3153:#ifndef NITRIF_DENITRIF CNDecompMod.F90 (done) 23:!#ifdef NITRIF_DENITRIF 100:!#ifndef NITRIF_DENITRIF 139:!#ifndef NITRIF_DENITRIF 168:!#ifndef NITRIF_DENITRIF 309:!#ifdef NITRIF_DENITRIF 363:!#ifndef NITRIF_DENITRIF 384:!#ifndef NITRIF_DENITRIF clmtype.F90 (no modification needed) 1369: ! NITRIF_DENITRIF 1679: ! NITRIF_DENITRIF 1856: ! NITRIF_DENITRIF 1957: ! NITRIF_DENITRIF 1989: ! NITRIF_DENITRIF diagnostic quantities 2008: ! no NITRIF_DENITRIF clm_varcon.F90 (done) 64:#ifdef NITRIF_DENITRIF CNNStateUpdate1Mod.F90 (done) 46:#ifdef NITRIF_DENITRIF 122:#ifdef NITRIF_DENITRIF 199:#ifndef NITRIF_DENITRIF 322:#ifndef NITRIF_DENITRIF 348:#ifndef NITRIF_DENITRIF 394:#ifndef NITRIF_DENITRIF 434:#else !------------- NITRIF_DENITRIF -------------------- CNSetValueMod.F90 (done) 1083:#ifdef NITRIF_DENITRIF 1101:#ifdef NITRIF_DENITRIF 1311:#ifndef NITRIF_DENITRIF 1376:#ifdef NITRIF_DENITRIF 1419:#ifndef NITRIF_DENITRIF 1431:#ifndef NITRIF_DENITRIF CNNStateUpdate3Mod.F90 (done) 64:#ifndef NITRIF_DENITRIF 154:#ifndef NITRIF_DENITRIF 240:#ifndef NITRIF_DENITRIF CNPrecisionControlMod.F90 (done) 171:#ifdef NITRIF_DENITRIF 311:#ifdef NITRIF_DENITRIF 804:#ifdef NITRIF_DENITRIF CNAllocationMod.F90 (done) 143:#ifdef NITRIF_DENITRIF 318:#ifndef NITRIF_DENITRIF 377:#ifndef NITRIF_DENITRIF 532:#ifndef NITRIF_DENITRIF 850:#ifndef NITRIF_DENITRIF 1027:#else !----------NITRIF_DENITRIF-------------! CNiniTimeVar.F90 (done) 78:#ifdef NITRIF_DENITRIF 388:#ifdef NITRIF_DENITRIF 772:#ifdef NITRIF_DENITRIF CNNitrifDenitrifMod.F90 3:!#ifdef NITRIF_DENITRIF 278: ! NITRIF_DENITRIF requires Methane model to be active, 282: !call endrun( trim(subname)//' ERROR: NITRIF_DENITRIF requires Methane model to be active' ) Modified the code to turn on nitif_denitrif. ========== 20190507: 1. Fixed the bugs in the fire model according to the bug report (http://bugs.cgd.ucar.edu/show_bug.cgi?id=1805) and the tag (clm4_6_00) mentioned in the report after the bug fix, thanks to Keith Oleson's help! 2. Flied back to Maryland. ========== 20190508: 1. Updated max fire spread rate following Li et al., 2014 BG. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/data > mkdir paramdata4.5 > cd paramdata4.5 > svn co --username guestuser --password friendly https://svn-ccsm-inputdata.cgd.ucar.edu/trunk/inputdata/lnd/clm2/paramdata > cd paramdata > cp -p clm_params.c140423.nc ../. [This is the one we need to update max fire spread rate]. pftnum = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 ; pftname = fsr_pft "not_vegetated ", 0 "needleleaf_evergreen_temperate_tree ", 0.4 "needleleaf_evergreen_boreal_tree ", 0.43 "needleleaf_deciduous_boreal_tree ", 0.43 "broadleaf_evergreen_tropical_tree ", 0.4 "broadleaf_evergreen_temperate_tree ", 0.4 "broadleaf_deciduous_tropical_tree ", 0.4 "broadleaf_deciduous_temperate_tree ", 0.4 "broadleaf_deciduous_boreal_tree ", 0.4 "broadleaf_evergreen_shrub ", 0.46 "broadleaf_deciduous_temperate_shrub ", 0.46 "broadleaf_deciduous_boreal_shrub ", 0.46 "c3_arctic_grass ", 0.6 "c3_non-arctic_grass ", 0.6 "c4_grass ", 0.6 "c3_crop ", 0.6 "c3_irrigated ", 0.6 "corn ", 0 "irrigated_corn ", 0 "spring_temperate_cereal ", 0 "irrigated_spring_temperate_cereal ", 0 "winter_temperate_cereal ", 0 "irrigated_winter_temperate_cereal ", 0 "soybean ", 0 "irrigated_soybean " ; 0 } fsr_pft = 0, 0.4, 0.43, 0.43, 0.4, 0.4, 0.4, 0.4, 0.4, 0.46, 0.46, 0.46, 0.6, 0.6, 0.6, 0.6, 0.6, 0, 0, 0, 0, 0, 0, 0, 0 ; Updated fsr_pfx in CN_init accordingly. Also updated the calculation of grid level "burn" and "nfire" in CN_Driver, since the burned area and nfire from CNFireMod are now per second instead of per time step. 2. Compiled the code with "gmake clean" and "gmake install"!! > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake clean > gmake install > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install > cd ../../Applications/LDAS_App/ > gmake clean > gmake install > cd ../../../ > mkdir -p exec/clm4.5_Kn11_Ndiags_fire > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags_fire/. 3. Do a test run. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71047518 2019-05-08 20:39 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_31/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j Crashed at date_time_new 19800102_193000z. forrtl: error (72): floating overflow Image PC Routine Line Source libirc.so 00002AAAAB590961 Unknown Unknown Unknown libirc.so 00002AAAAB58F0B7 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E3362 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E31B6 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2D1A0C Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2B3392 Unknown Unknown Unknown libc.so.6 00002AAAABA16910 Unknown Unknown Unknown LDASsaCN_mpi.x 00000000004D62CB catch_types_mp_ad 1757 catch_types.F90 LDASsaCN_mpi.x 00000000005030F9 clsm_ensdrv_out_r 725 clsm_ensdrv_out_routines.F90 LDASsaCN_mpi.x 00000000004AEF09 MAIN__ 2335 cnlsm_ensdrv_main.F90 4. Re-compiled with "gmake clean" and "gmake install BOPT=g". > cd ../../../ > /bin/cp -pr Linux/ exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98960695 2019-05-08 21:08 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/global/rc_out/* > ./lenkf.0.i It was able to run until date_time_new 19800210_060730z and still working on the Amazon point. Ran on the entire globe using 24 cores on DDT. Crashed at date_time_new 19800102_193000z. It's because nfire is too large: 3.36e+38! ========== 20190509: 1. Initialized nfire to 0 in CNFireArea. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake clean > gmake install BOPT=g > cd ../../../ > /bin/cp -pr Linux/ exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 98960775 2019-05-09 13:28 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/global/rc_out/* > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > ./lenkf.0.j Tried twice on DDT and worked at the 2nd time. Re-compiled with "gmake clean" and "gmake install". > cd ../../../ > /bin/cp -pr Linux/ exec/clm4.5_Kn11_Ndiags_fire/. 2. Do a test run. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71846003 2019-05-09 14:43 ../build/Linux/bin/LDASsaCN_mpi.x* > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It stopped at date_time_new 19801025_090730z due to time out (1 hour). 2. clm4.5_DE720 and clm4.5_Princeton_DE720 just finished another cycle. Now it's time to update the executable and then resume the spinup. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec > mv clm4.5_Kn11_Ndiags clm4.5_Kn11_Ndiags_20190419 > mv clm4.5_Kn11_Ndiags_fire clm4.5_Kn11_Ndiags > rm -rf clm4.5_Kn11_16Ndiags > ls -l clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71846003 2019-05-09 14:43 clm4.5_Kn11_Ndiags/Linux/bin/LDASsaCN_mpi.x* > cd discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71846003 2019-05-09 14:43 build/Linux/bin/LDASsaCN_mpi.x* Checked restart file. Looks correct. > cd run > ./ldas_batchrun.j > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71846003 2019-05-09 14:43 ../build/Linux/bin/LDASsaCN_mpi.x* Checked restart file. Looks correct. > ./ldas_batchrun.j ========== 20190510: 1. Did some comparisons before and after fire bug fix. Emailed Melanie: (1) now there are 102 fields (2) now nfire does not have the huge number as missing values Changes in code: CNFireMod.F90 vs. CNFireMod.F90.bug or read Fang's 2014 BG paper. 2. Emailed Ben and Matt asking if they can help us to speed up the spinup process. 3. Met with Ben. He showed me how to time a particular block of code (e.g. call a subroutine) takes, using /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Applications/LDAS_App/cnlsm_ensdrv_main.F90 as an example. (1) Add "use m_zeit" to the module (2) Add "call zeit_ci("total")" before the block of code, and add "call zeit_co("total")" after the block of code. "total" can be something else but should be unique for each individual block of code. (3) Add "call zeit_allflush(comm=MPI_COMM_WORLD,root=0,lu=OUTPUT_UNIT)" at the end of the module. This is what it will write out as an example: [MWTIME]x308 NET avg %tt max imx x% r% i% SCP avg %tt max imx x% r% i% -------------------------------------------------------------------------------- .zeit. 0.2 1 0.2 0E5 0 1 18 14.0 100 14.1 019 0 0 1 308x total 13.9 99 14.0 00F 0 0 1 13.9 99 14.0 00F 0 0 1 308x -------------------------------------------------------------------------------- Note: The above test run was done in /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720. The executable is in /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/Linux/bin or Applications/LDAS_App. The clm4.5_fire_DE720 run I did on 20190509 was renamed to clm4.5_fire_DE720_keep. ========== 20190513: 1. Both clm4.5_DE720 and clm4.5_Princeton_DE720 stopped over the weekend. clm4.5_DE720: stopped at date_time_new 20161224_090730z due to time out. It was able to finish 8 years of simulation within 12 hours for 1980-2012. Why did it take more than 12 hours to finish 2012-2016? /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/output/global/rs/ens0000 > ls -l Y2016/M01 total 296800 -rw-r--r-- 1 fzeng g0620 303909763 2019-05-11 15:30 clm4.5_DE720.ens0000.catchcn_ldas_rst.20160101_0000z > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > cp -p lenkf.4.j lenkf.4b.j > nedit lenkf.4b.j & #SBATCH --time=01:00:00 #SBATCH --qos=debug #SBATCH --output=../output/global/rc_out/Y2012/M01/clm4.5_DE720.ldas_log.20160101_0000.txt #SBATCH --error=../output/global/rc_out/Y2012/M01/clm4.5_DE720.ldas_out.20160101_0000.txt -start_year 2016 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ > nedit lenkf.4b.j & clm4.5_Princeton_DE720: stopped at the beginning of 19960101. In the *out* file: process killed (SIGTERM) LDASsaCN_mpi.x 000000000049FB45 MAIN__ 722 cnlsm_ensdrv_main.F90 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/output/global/rs/ens0000/Y1996/M01 > ls -l total 296800 -rw-r--r-- 1 fzeng g0620 303909763 2019-05-11 14:41 clm4.5_Princeton_DE720.ens0000.catchcn_ldas_rst.19960101_0000z > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > cp ldas_batchrun.j ldas_batchrun_b.j > nedit ldas_batchrun_b.j & > ./ldas_batchrun_b.j & 2. Made plots using GrADS to compare fire counts before and after bug fixed. This wasn't done on Friday last week because the nfire missing value is a huge number in the cycle prior to fire bug fix, and it's 0 in the cycles afte the fire bug fix. Found a way to set the huge nfire missing values in the previous cycles to 0 on GrADS and made the comparison plots for nifre before and after bug fix. Sent the plots (path of the plot file) to Melanie. 3. The clm4.5_DE720 run stopped after 1 hour but it wasn't able to finish simulating 2016. Received email from NCCS saying "slower than normal I/O operations" and some other problem. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > nedit lenkf.4b.j & #SBATCH --time=02:00:00 #SBATCH --output=../output/global/rc_out/Y2012/M01/clm4.5_DE720.ldas_log.20160101b_0000.txt #SBATCH --error=../output/global/rc_out/Y2012/M01/clm4.5_DE720.ldas_out.20160101b_0000.txt > qsub lenkf.4b.j The clm4.5_Princeton_DE720 run also stopped. It finished simulating 1996. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > cp lenkf.2.j lenkf.2.j.orig > nedit lenkf.2.j & #SBATCH --output=../output/global/rc_out/Y1996/M01/clm4.5_Princeton_DE720.ldas_log.19970101_0000.txt #SBATCH --error=../output/global/rc_out/Y1996/M01/clm4.5_Princeton_DE720.ldas_out.19970101_0000.txt -start_year 1997 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ > ./ldas_batchrun_b.j 4. Following 3 on 20190510. Added some "call zeit_ci" and "call zeit_co" in /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Applications/LDAS_App/cnlsm_ensdrv_main.F90. Compiled. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j ========== 20190514: 1. The clm4.5_DE720 is running, which is good. But the clm4.5_Princeton_DE720 stopped when it's processing the output. 1996 6 1996 7 1996 8 1996 9 1996 10 1996 11 1996 12 3 67 Checked the size of the 199612 output file. It's the same as that of 199611. Tried to process the output manually but it just hanged there after printing out some information and did not even process 198001 file. Opened another terminal window and tried there. It also hang there. Submitted the job again to re-do the whole cycle. ========== 20190515: 1. The clm4.5_DE720 and clm4.5_Princeton_DE720 runs stopped again. Discover was still having some issue yesterday. clm4.5_DE720: stopped at date_time_new 19970925_230730z /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720> ls -l output/global/rs/ens0000/Y1997/M01 total 296800 -rw-r--r-- 1 fzeng g0620 303909763 2019-05-14 15:08 clm4.5_DE720.ens0000.catchcn_ldas_rst.19970101_0000z > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > cp -p lenkf.2.j lenkf.2.j.orig > nedit lenkf.2.j & #SBATCH --output=../output/global/rc_out/Y1996/M01/clm4.5_DE720.ldas_log.19970101_0000.txt #SBATCH --error=../output/global/rc_out/Y1996/M01/clm4.5_DE720.ldas_out.19970101_0000.txt -start_year 1997 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ > cp -p ldas_batchrun.j ldas_batchrun.j.orig > nedit ldas_batchrun.j & > ./ldas_batchrun.j clm4.5_Princeton_DE720: stopped at the very beginning of 20120101. > ls -l output/global/rs/ens0000/Y2012/M01 total 296800 -rw-r--r-- 1 fzeng g0620 303909763 2019-05-14 04:34 clm4.5_Princeton_DE720.ens0000.catchcn_ldas_rst.20120101_0000z > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > qsub lenkf.4.j It looks like there is something wrong with the 199612 output file. It stopped when the post processing script is processing the 199612 output. Just re-run starting from 199601. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_Princeton_DE720/run > cp -p lenkf.2.j.orig lenkf.2.j > ./ldas_batchrun_b.j 2. Following 3 on 20190510. Added some "call zeit_ci" and "call zeit_co" in /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Applications/LDAS_App/cnlsm_ensdrv_main.F90. Compiled. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > nedit lenkf.0.j & -start_year 1980 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ -end_year 1980 -end_month 7 -end_day 1 -end_hour 0 -end_min 0 -end_sec 0 \ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j ========== 20190516: 1. Followed 2 on 20190515 to add more timers. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j Stopped at date time 19800628 due to time out!! > nedit lenkf.0.j & -start_year 1980 -start_month 1 -start_day 1 -start_hour 0 -start_min 0 -start_sec 0 \ -end_year 1980 -end_month 6 -end_day 1 -end_hour 0 -end_min 0 -end_sec 0 \ ========== 20190517: 1. Emailed Ben what I found yesterday: reading and distributing forcing data takes 67% of the totol run time. Talked to Ben. He suggested me to add a couple of more timers to see how much time reading forcing takes and how much time distributing forcing takes. Check the time needed: LDASsa vs. GEOSldas Added the two timers Ben suggested: "MPI_SCATTERV" is the step that takes most of the time! According to this, the new GEOSldas which doesn't need to distribute the forcing data should be faster than the old LDASsa. However, this is not the case for me. I checked two of my previous simulations, both using MERRA-2, at CF90 resolution and with the CN model on. The LDASsa one takes 4 hours to simulate 1 year, while the GEOSldas one takes 24 hours to simulate 1 year. One difference is that the LDASsa one wrote restart file once a year, and the GEOSldas one wrote restart file twice a year. I will talk to Weiyuan and/or Justin if they also find that GEOSldas is slower than LDASsa. ========== 20190520: 1. Emailed Ben what I found on 20190517. 2. Why is my GEOSldas_CF90 run that uses the GEOSldas tag so slow? Did I forget to re-build with "gmake realclean" and "gmake install" after I built with "gmake install BOPT=g"? Re-built GEOSldas with "gmake realclean" and "gmake install" and did a test run to see how long it takes to finish 1 year of simulation. > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/src > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install > cd .. > mkdir -p exec/tranCO2_test > /bin/cp -pr Linux exec/tranCO2_test/. Set up the GEOSldas_CF90_test run again. > cd /discover/nobackup/fzeng/Catchment/M2n5P/run > cp ../GEOSldas_CF90/run/CF0090x6C.exec . > nedit CF0090x6C.exec & EXP_ID : GEOSldas_CF90_test BEG_DATE : 19810101 000000 END_DATE : 19820101 000000 JOB_SGMT : 00000600 000000 > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/exec/tranCO2_test/Linux/bin > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/ > source $ESMADIR/src/g5_modules > ./ldas_setup setup --runmodel /discover/nobackup/fzeng/Catchment/M2n5P/ /discover/nobackup/fzeng/Catchment/M2n5P/run/CF0090x6C.exec /discover/nobackup/fzeng/Catchment/M2n5P/run/CF0090x6C.bat > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test/run/ Check restart file: > cat cap_restart 19810101 000000 > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 151 2019-05-20 14:57 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test/output/CF0090x6C/rs/ens0000/Y1981/M01/GEOSldas_CF90_test.catchcn_internal_rst.19810101_0000 lrwxrwxrwx 1 fzeng g0620 126 2019-05-20 14:57 vegdyn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test/output/CF0090x6C/rs/ens0000/GEOSldas_CF90_test.vegdyn_internal_rst Check executable: > ls -l ../build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50249434 2019-05-20 14:42 ../build/Linux/bin/GEOSldas.x* CAP.rc: check the END_DATE, JOB_SGMT and NUM_SGMT. No change needed. HISTORY.rc: tavg1_2D_mf-st.resolution: 360 181, tavg1_2D_lnd_Nx.resolution: 360 181, Added to L236: 'CNCO2' , 'CATCHCN' , LDAS.rc: added "IMPOSE_CT_CO2: 1" to the end (L38). Did an interactive run: > interactive.py -A sp3 -n 168 -a g0620 -X --debug > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test/run > cp lenkf.j lenkf.j.orig > nedit lenkf.j & Add "exit" after L179 "mpirun -map-by core --mca btl ^vader -np $numprocs $GEOSBIN/GEOSldas.x" > ./lenkf.j It's running. Stopped it at 1981-01-01T04:52:30. > mv lenkf.j.orig lenkf.j > qsub lenkf.j 3. Processed output: clm4.5_DE720: 32 through 38 clm4.5_Princeton_DE720: 63 through 71 ========== 20190521: 1. The GEOSldas_CF90_test run took ~9 hours to finish 1 year of simulation. Talked to Weiyuan. We did a run to write daily output as well as monthly output. It took at lease 1 minute to finish 1 day of simulation, that is more than 6 hours for 1 year of simulation. Weiyuan will check how to enable timer so that we can see if writing output is the slowest step. 2. Made plots of fire burned area and C loss. Also made time series plots of a few variables for a couple of grids to see if we can find something wrong or some way to improve the fire model. Sent Melanie the time series plots. ========== 20190522: 1. Weiyuan told me that I can add "MAPL_ENABLE_TIMERS: YES" to CAP.rc to enable timer. CAP.rc: Added "MAPL_ENABLE_TIMERS: YES". Set "END_DATE: 19820601 000000". Submitted the job. It didn't print out the information about time consumed. Actually in CAP.rc there is a line "MAPL_ENABLE_TIMERS: NO". I should just change "NO" to "YES". Ran for 5 days. Got this below: Times for DATAATM TOTAL : 47.936 Initialize : 1.689 Run_GetForcing : 47.872 Run_RepairForcing : 0.000 GenInitTot : 0.038 --GenInitMine : 0.038 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.002 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for CATCHCN TOTAL : 360.372 RUN1 : 19.043 -RUN0 : 7.543 -SURF : 3.382 RUN2 : 271.994 -CATCH : 0.000 -ALBEDO : 1.997 GenInitTot : 64.109 --GenInitMine : 64.109 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.324 --GenFinalMine : 5.324 GenRecordTot : 0.001 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 363.459 INITIALIZE : 65.404 RUN1 : 20.785 RUN2 : 272.040 VEGDYN : 2.954 CATCHCN : 360.431 GenInitTot : 65.404 --GenInitMine : 0.078 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.324 --GenFinalMine : 0.000 GenRecordTot : 0.014 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 504.627 Initialize : 111.953 -LocStreamCreate : 13.211 Run : 346.425 DATAATM : 50.310 LANDPERT : 1.252 LAND : 363.484 ENSAVG : 0.040 GenInitTot : 89.182 --GenInitMine : 22.002 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 6.042 --GenFinalMine : 0.000 GenRecordTot : 0.041 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 10.735 Initialize : 1.344 Finalize : 0.002 Run : 9.400 --Couplers : 3.514 --I/O : 7.429 ----IO Create : 0.228 ----IO Write : 7.200 -----IO Post : 2.339 -----IO Wait : 3.322 -----IO Write : 4.780 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/src/GEOSldas_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit GEOS_CatchCNGridComp.F90 & call MAPL_TimerOn ( MAPL, "-CATCHCN" ) [it was "call MAPL_TimerOn ( MAPL, "-CATCHCN" )" which is giving some error message.] > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/ > source $ESMADIR/src/g5_modules "gmake install" in: GEOScatchCN_GridComp Applications/LDAS_App > cd ../../.. > /bin/cp -pr Linux exec/tranCO2_test/. /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test > ls -l build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50249434 2019-05-22 16:07 build/Linux/bin/GEOSldas.x* Run for another 3 months. Times for DATAATM TOTAL : 1013.261 Initialize : 1.781 Run_GetForcing : 1013.166 Run_RepairForcing : 0.000 GenInitTot : 0.041 --GenInitMine : 0.041 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.018 --GenRecordMine : 0.008 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for CATCHCN TOTAL : 6401.329 RUN1 : 409.566 -RUN0 : 166.848 -SURF : 74.449 RUN2 : 5928.810 -CATCHCN : 5783.581 -ALBEDO : 43.815 GenInitTot : 64.877 --GenInitMine : 64.878 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 9.299 --GenFinalMine : 9.299 GenRecordTot : 0.017 --GenRecordMine : 0.007 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 6424.634 INITIALIZE : 66.271 RUN1 : 430.428 RUN2 : 5929.838 VEGDYN : 21.959 CATCHCN : 6402.648 GenInitTot : 66.271 --GenInitMine : 0.117 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 9.299 --GenFinalMine : 0.000 GenRecordTot : 0.305 --GenRecordMine : 0.007 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 7652.441 Initialize : 105.566 -LocStreamCreate : 5.265 Run : 7487.577 DATAATM : 1016.152 LANDPERT : 27.009 LAND : 6425.171 ENSAVG : 0.162 GenInitTot : 90.067 --GenInitMine : 21.929 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 10.267 --GenFinalMine : 0.000 GenRecordTot : 0.879 --GenRecordMine : 0.008 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 129.823 Initialize : 1.545 Finalize : 0.002 Run : 128.471 --Couplers : 77.091 --I/O : 94.380 ----IO Create : 4.973 ----IO Write : 89.402 -----IO Post : 3.132 -----IO Wait : 11.947 -----IO Write : 80.818 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 2. Read the plot Melanie made and Fang Li's paper to understand the fire model and investigate why our burned area is so high in S America. ========== 20190523: 1. Talked to Sarith that CATCHCN is slower in GEOSldas than in LDASsa. Sarith suggested me to add timers for "compute_rc", "CN_Driver" and "CATCHCN". Modified GEOS_CatchCNGridComp.F90 to do so. > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/src/GEOSldas_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_6/ > source $ESMADIR/src/g5_modules "gmake install" in: GEOScatchCN_GridComp Applications/LDAS_App > cd ../../.. > /bin/cp -pr Linux exec/tranCO2_test/. /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test > ls -l build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50249836 2019-05-23 11:18 build/Linux/bin/GEOSldas.x* Run for 5 days. Times for DATAATM TOTAL : 53.637 Initialize : 1.776 Run_GetForcing : 53.580 Run_RepairForcing : 0.000 GenInitTot : 0.175 --GenInitMine : 0.175 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.002 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for CATCHCN TOTAL : 422.984 RUN1 : 23.561 -RUN0 : 9.450 -SURF : 4.200 RUN2 : 328.562 -CATCHCN : 320.331 -ALBEDO : 2.495 -COMPUTE_RC : 60.043 -CN_DRIVER : 150.847 -CALL_CATCHCN : 85.277 GenInitTot : 66.753 --GenInitMine : 66.753 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.114 --GenFinalMine : 5.114 GenRecordTot : 0.002 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 427.563 INITIALIZE : 68.124 RUN1 : 26.701 RUN2 : 328.621 VEGDYN : 4.423 CATCHCN : 423.057 GenInitTot : 68.124 --GenInitMine : 0.080 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.114 --GenFinalMine : 0.000 GenRecordTot : 0.018 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 578.254 Initialize : 107.446 -LocStreamCreate : 5.566 Run : 415.143 DATAATM : 55.393 LANDPERT : 1.692 LAND : 427.595 ENSAVG : 0.042 GenInitTot : 92.100 --GenInitMine : 22.112 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.115 --GenFinalMine : 0.000 GenRecordTot : 0.049 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 12.106 Initialize : 1.438 Finalize : 0.003 Run : 10.670 --Couplers : 4.374 --I/O : 8.115 ----IO Create : 0.294 ----IO Write : 7.821 -----IO Post : 2.348 -----IO Wait : 3.443 -----IO Write : 5.248 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 So CN_Driver is the slowest step. Talked to Sarith. He told me to add print statement to see if CN_Driver is called every 5400s or 450s. Modified GEOS_CatchCNGridComp.F90 to do so. "gmake install" in: GEOScatchCN_GridComp Applications/LDAS_App > cd ../../.. > /bin/cp -pr Linux exec/tranCO2_test/. /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test > ls -l build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50250749 2019-05-23 11:48 build/Linux/bin/GEOSldas.x* Run for 1 days. "compute_rc" and "CATCHCN" are called every 450s, which is as expected. However, "CN_Driver" is called every 3600s (hour), while we want to call it every 5400s. Why? It's because AGCM_S is actually the second in each hour not the second in each day, so at the beginning of each hour AGCM_S is 0, and this "if(mod(AGCM_S,nint(dtcn)) == 0)" statement is met and therefore CN_Driver is called every hour. This bug is alreay fixed in the newer tags. 2. Checked out Sarith's latest tag: GEOSldas_m4-17_UCatchCN > cd /discover/nobackup/fzeng/offline_code/tmp > cvs co -r GEOSldas_m4-17_UCatchCN GEOSldas_m4 > mv GEOSldas ../GEOSldas_m4-17_UCatchCN > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/src/GEOSldas_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp Checked the timers. Added "print *, 'calling CN_Driver:',AGCM_YY,AGCM_MM,AGCM_DD,AGCM_HH,AGCM_MI,AGCM_S" before calling CN_Driver. Compiled: > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/src > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN > source $ESMADIR/src/g5_modules > make -j 8 install > cd .. > mkdir -p exec/tranCO2_test2 > /bin/cp -pr Linux exec/tranCO2_test2/. Set up the GEOSldas_CF90_test run again. > cd /discover/nobackup/fzeng/Catchment/M2n5P/run > nedit CF0090x6C.exec & EXP_ID : GEOSldas_CF90_test2 BEG_DATE : 19810101 000000 END_DATE : 19810106 000000 JOB_SGMT : 00000600 000000 > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/exec/tranCO2_test2/Linux/bin > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/ > source $ESMADIR/src/g5_modules > ./ldas_setup setup --runmodel /discover/nobackup/fzeng/Catchment/M2n5P/ /discover/nobackup/fzeng/Catchment/M2n5P/run/CF0090x6C.exec /discover/nobackup/fzeng/Catchment/M2n5P/run/CF0090x6C.bat Printed out on screen: 2 360 1981 12 27 creating dir structure creating restart and bc Correct the tile file if it is an old EASE tile format... cmd: ./preprocess_ldas.x correctease /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out//CF0090x6C_DE0360xPE0180-Pfafstetter.til /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out//MAPL_CF0090x6C_DE0360xPE0180-Pfafstetter.til Creating f2g.txt.... cmd: ./preprocess_ldas.x c_f2g /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out//CF0090x6C_DE0360xPE0180-Pfafstetter.til LDAS_domain_def.nml /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C /discover/nobackup/ltakacs/bcs/Icarus-NL/Icarus-NL_Reynolds/CF0090x6C_DE0360xPE0180/clsm/catchment.def GEOSldas_CF90_test2 198101010000 Creating domain..., reading white and black lists if there have ones... Finish domain setup... Writing catparam file : /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_ test2/output/CF0090x6C/rc_out///Y1981/M01/GEOSldas_CF90_test2.ldas_catparam.198 10101_0000z.bin linking bcs... Creating and lining restart... cmd: ./process_rst.csh g0620 GEOSldas_CF90_test2 /discover/nobackup/fzeng/Catchment/M2n5P /discover/nobackup/ltakacs/bcs/Icarus-NL/Icarus-NL_Reynolds/CF0090x6C_DE0360xPE0180/ CF0090x6C_DE0360xPE0180-Pfafstetter.til 2 1 19810101 e0004s_transientCO2_05 SMAP_EASEv2_M09_GLOBAL /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/ 1 0 Please hold on for a while until the restart file is created ..... restart: 1 catcRstFile1: /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/e0004s_transientCO2_05/output/SMAP_EASEv2_M09_GLOBAL/rs/ens0000/Y1981/M01/e0004s_transientCO2_05.catchcn_internal_rst.19810101_0000 catcRstFile: /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1981/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19810101_0000 Updating restart path... creating RC Files Optimizing... decomposition of processes.... cmd: ./preprocess_ldas.x optimize /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/input/tile.data 168 total tiles 475330 land_distribute: 2742 3063 2816 3420 3633 2818 2867 2824 2807 2948 3090 3120 3149 3112 2635 3538 3735 3700 3825 3906 4089 4162 4318 4153 4205 3968 3801 3785 3466 3366 3056 3096 2881 1657 0 0 0 0 0 0 2509 2474 2517 2539 2669 2649 2678 2654 2461 2844 3264 2738 3353 3983 4805 2487 2497 2744 2680 2651 2820 2692 2779 2778 2787 2760 2870 2880 2846 2927 2873 2763 2686 2707 2494 2486 2532 2589 4930 4849 2648 2601 2660 2656 2684 2609 4996 5187 2895 3119 3060 3133 3137 3112 3183 3260 3149 3255 3168 3128 3096 4757 4294 4291 3710 2799 3058 2693 2819 3400 3468 2697 3622 3595 2628 2857 2927 3133 3031 3191 3109 2995 2777 2721 0 0 0 0 0 0 0 3200 3289 2846 2969 3333 2656 2582 3213 2844 1655 2639 2416 2193 2679 3196 3294 2468 2320 2586 2889 2232 2784 2432 2532 2944 3804 3892 4041 3890 3708 3535 3155 3176 2819 0 1937 1879 JMS.rc 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 2 2 2 2 2 2 3 3 3 4 4 4 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 4 4 3 3 3 2 2 3 4 62 29 4 3 3 3 3 3 3 3 4 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 70 20 ['/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_CAP.rc', '/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_ExtData.rc', '/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_HIST.rc', '/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_LDAS.rc'] CAP.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run ExtData.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run HIST.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run PE90x540-CF LDAS.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run creating gcm style batch Run scripts lenkf.j Experiment directory: /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2 creating batch Run scripts > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run/ Check restart file: > cat cap_restart 19810101 000000 > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 153 2019-05-23 13:02 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1981/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19810101_0000 lrwxrwxrwx 1 fzeng g0620 128 2019-05-23 13:02 vegdyn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/GEOSldas_CF90_test2.vegdyn_internal_rst Check executable: > ls -l ../build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50809485 2019-05-23 12:47 ../build/Linux/bin/GEOSldas.x* CAP.rc: check the END_DATE, JOB_SGMT and NUM_SGMT. No change. MAPL_ENABLE_TIMERS: YES HISTORY.rc: tavg1_2D_lfs_Nx.resolution: 360 181, tavg1_2D_lnd_Nx.resolution: 360 181, Added to L236: 'CNCO2' , 'CATCHCN' , No longer need this below: LDAS.rc: added "IMPOSE_CT_CO2: 1" to the end (L38). Did an interactive run: > interactive.py -A sp3 -n 168 -a g0620 -X --debug > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > cp lenkf.j lenkf.j.orig > nedit lenkf.j & Add "exit" after L278 "mpirun -map-by core --mca btl ^vader -np $numprocs $GEOSBIN/GEOSldas.x" > ./lenkf.j Run 5 days. CN_Driver is called every 5400s. Correct! Times for DATAATM TOTAL : 60.175 Initialize : 2.445 Run_GetForcing : 60.123 Run_RepairForcing : 0.000 GenInitTot : 0.032 --GenInitMine : 0.032 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.002 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for CATCHCN TOTAL : 237.546 RUN1 : 14.897 -RUN0 : 5.708 -SURF : 2.355 RUN2 : 138.219 -CATCH : 119.782 -ALBEDO : 1.403 GenInitTot : 77.076 --GenInitMine : 77.076 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 7.722 --GenFinalMine : 7.722 GenRecordTot : 0.002 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 240.052 INITIALIZE : 78.458 RUN1 : 15.943 RUN2 : 138.287 VEGDYN : 2.351 CATCHCN : 237.630 GenInitTot : 78.458 --GenInitMine : 0.071 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 7.722 --GenFinalMine : 0.000 GenRecordTot : 0.019 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 382.713 Initialize : 120.339 -LocStreamCreate : 5.723 Run : 213.069 DATAATM : 62.608 LANDPERT : 1.127 LAND : 240.079 ENSAVG : 0.041 GenInitTot : 81.012 --GenInitMine : 0.025 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 7.723 --GenFinalMine : 0.000 GenRecordTot : 0.052 --GenRecordMine : 0.001 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 11.562 Initialize : 1.353 Finalize : 0.003 Run : 10.226 --Couplers : 3.345 --I/O : 8.022 ----IO Create : 0.278 ----IO Write : 7.744 -----IO Post : 2.330 -----IO Wait : 3.366 -----IO Write : 5.218 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 So it's much faster now. 3. > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 153 2019-05-23 13:02 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1981/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19810101_0000 lrwxrwxrwx 1 fzeng g0620 128 2019-05-23 13:02 vegdyn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/GEOSldas_CF90_test2.vegdyn_internal_rst > cat cap_restart 19810101 000000 I wonder why they are both still 19810101. Is it because I did the 5-day run interactively? Looks like it is. > nedit CAP.rc & END_DATE: 19820101 000000 > cp lenkf.j.orig lenkf.j > qsub lenkf.j ========== 20190524: 1. For Sarith's GEOSldas_m4-17_UCatchCN, CN_Driver is called 5400s which is correct. It took about 4.5 hours to finish 1 year of simulation. Couldn't find log file(s) in rc_out. The scratch directory has "GEOSldas_log.txt" but this file only shows the time taken but no other information about the simulation. Also I couldn't find *.exec and *.bat files in Linux/bin or Applications/LDAS_App. Used the ones from the previous GEOSldas_CF90_test run. Need to ask Weiyuan. Talked to Sarith. See below. 2. In GEOS_CatchCNGridComp.F90: call MAPL_GetResource ( MAPL, ATM_CO2, Label="ATM_CO2:",DEFAULT=2, RC=STATUS) VERIFY_(STATUS) ! 0: uses a fix value defined by CO2 ! 1: CT tracker monthly mean diurnal cycle ! 2: CT tracker monthly mean diurnal cycle scaled to match EEA global average CO2 ! 3: spatially fixed interannually varyiing CMIP from getco2.F90 look up table (AGCM only) ! 4: import AGCM model CO2 (AGCM only) So "ATM_CO2: 2" in LDAS.rc is good. 3. Plan to run for another year using 3-hour DTCN. > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > ls -l ../input/restart/ total 0 lrwxrwxrwx 1 fzeng g0620 153 2019-05-23 23:09 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1982/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19820101_0000 > cat cap_restart 19820101 000000 > nedit CAP.rc & END_DATE: 19830101 000000 JOB_SGMT: 00010000 000000 NUM_SGMT: 1 > nedit lenkf.j & setenv NODAILIES 2 [was 0 for the simulation of 1981 yesterday. Now 2 means only keep monthly and delete daily] Talked to Sarith: the way I used the old *.exec and *.bat files is not good. Re-do GEOSldas_CF90_test2. /discover/nobackup/fzeng/Catchment/M2n5P > mv GEOSldas_CF90_test2 GEOSldas_CF90_test2_old 4. Set up GEOSldas_CF90_test2 following GEOSldas_tutorial. > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/exec/tranCO2_test2/Linux/bin > source /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/src/g5_modules > ./ldas_setup sample --exeinp > GEOSldas_CF90_test2.exec > ./ldas_setup sample --batinp > GEOSldas_CF90_test2.bat > mv GEOSldas_CF90_test2.exec GEOSldas_CF90_test2.bat /discover/nobackup/fzeng/Catchment/M2n5P/run/. > cd /discover/nobackup/fzeng/Catchment/M2n5P/run > nedit GEOSldas_CF90_test2.bat GEOSldas_CF90_test2.exec & [following GEOSldas_CF90.bat and GEOSldas_CF90.exec] GEOSldas_CF90_test2.exec: EXP_ID : GEOSldas_CF90_test2 EXP_DOMAIN : CF0090x6C NUM_ENSEMBLE : 1 BEG_DATE : 19810101 000000 END_DATE : 19810106 000000 RESTART : 1 RESTART_ID : e0004s_transientCO2 RESTART_PATH : /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09 RESTART_DOMAIN : SMAP_EASEv2_M09_GLOBAL /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/e0004s_transientCO2_05/output/SMAP_EASEv2_M09_GLOBAL/rs/ens0000/Y1981/M01 > ls -l lrwxrwxrwx 1 fzeng g0620 127 2018-07-23 16:28 e0004s_transientCO2_05.catchcn_internal_rst.19810101_0000 -> /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/rst_CF90_offline/Icarus-NL_Reynolds_CF90/catchcn_internal_rst.19810101_0000z lrwxrwxrwx 1 fzeng g0620 110 2018-07-05 16:13 e0004s_transientCO2_05.catchcn_internal_rst.19810101_0000.DE_02880x01440 -> /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/rst_pLAI/DE_02880x01440/catchcn_internal_rst.19810101_0000z -rw-r--r-- 1 fzeng g0620 7547615847 2017-07-10 14:14 e0004s_transientCO2.ens0000.catchcn_ldas_rst.19810101_0000z > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/exec/tranCO2_test2/Linux/bin > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/ > source $ESMADIR/src/g5_modules The ldas_setup needs updated: > cvs upd -r 1.1.2.2.2.66 ldas_setup > ./ldas_setup setup --runmodel /discover/nobackup/fzeng/Catchment/M2n5P/ /discover/nobackup/fzeng/Catchment/M2n5P/run/GEOSldas_CF90_test2.exec /discover/nobackup/fzeng/Catchment/M2n5P/run/GEOSldas_CF90_test2.bat 11 180 1981 6 30 180 1981 12 27 180 1982 6 25 180 1982 12 22 180 1983 6 20 180 1983 12 17 180 1984 6 14 180 1984 12 11 180 1985 6 9 180 1985 12 6 180 1986 6 4 creating dir structure creating restart and bc Correct the tile file if it is an old EASE tile format... cmd: ./preprocess_ldas.x correctease /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out//CF0090x6C_DE0360xPE0180-Pfafstetter.til /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out//MAPL_CF0090x6C_DE0360xPE0180-Pfafstetter.til Creating f2g.txt.... cmd: ./preprocess_ldas.x c_f2g /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out//CF0090x6C_DE0360xPE0180-Pfafstetter.til LDAS_domain_def.nml /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C /discover/nobackup/ltakacs/bcs/Icarus-NL/Icarus-NL_Reynolds/CF0090x6C_DE0360xPE0180/clsm/catchment.def GEOSldas_CF90_test2 198101010000 Creating domain..., reading white and black lists if there have ones... Finish domain setup... Writing catparam file : /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_ test2/output/CF0090x6C/rc_out///Y1981/M01/GEOSldas_CF90_test2.ldas_catparam.198 10101_0000z.bin linking bcs... Creating and lining restart... cmd: ./process_rst.csh g0620 GEOSldas_CF90_test2 /discover/nobackup/fzeng/Catchment/M2n5P /discover/nobackup/ltakacs/bcs/Icarus-NL/Icarus-NL_Reynolds/CF0090x6C_DE0360xPE0180/ CF0090x6C_DE0360xPE0180-Pfafstetter.til 2 1 19810101 e0004s_transientCO2_05 SMAP_EASEv2_M09_GLOBAL /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/ 1 0 Please hold on for a while until the restart file is created ..... restart: 1 catcRstFile1: /discover/nobackup/fzeng/Catchment/SMAP_EASEv2_M09/e0004s_transientCO2_05/output/SMAP_EASEv2_M09_GLOBAL/rs/ens0000/Y1981/M01/e0004s_transientCO2_05.catchcn_internal_rst.19810101_0000 catcRstFile: /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1981/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19810101_0000 Updating restart path... creating RC Files Optimizing... decomposition of processes.... cmd: ./preprocess_ldas.x optimize /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/input/tile.data 168 total tiles 475330 land_distribute: 2742 3063 2816 3420 3633 2818 2867 2824 2807 2948 3090 3120 3149 3112 2635 3538 3735 3700 3825 3906 4089 4162 4318 4153 4205 3968 3801 3785 3466 3366 3056 3096 2881 1657 0 0 0 0 0 0 2509 2474 2517 2539 2669 2649 2678 2654 2461 2844 3264 2738 3353 3983 4805 2487 2497 2744 2680 2651 2820 2692 2779 2778 2787 2760 2870 2880 2846 2927 2873 2763 2686 2707 2494 2486 2532 2589 4930 4849 2648 2601 2660 2656 2684 2609 4996 5187 2895 3119 3060 3133 3137 3112 3183 3260 3149 3255 3168 3128 3096 4757 4294 4291 3710 2799 3058 2693 2819 3400 3468 2697 3622 3595 2628 2857 2927 3133 3031 3191 3109 2995 2777 2721 0 0 0 0 0 0 0 3200 3289 2846 2969 3333 2656 2582 3213 2844 1655 2639 2416 2193 2679 3196 3294 2468 2320 2586 2889 2232 2784 2432 2532 2944 3804 3892 4041 3890 3708 3535 3155 3176 2819 0 1937 1879 JMS.rc 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 2 2 2 2 2 2 3 3 3 4 4 4 3 3 3 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 4 4 3 3 3 2 2 3 4 62 29 4 3 3 3 3 3 3 3 4 3 3 2 2 2 2 2 2 2 2 2 2 2 2 2 70 20 ['/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_CAP.rc', '/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_ExtData.rc', '/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_HIST.rc', '/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc/GEOSldas_LDAS.rc'] CAP.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run ExtData.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run HIST.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run PE90x540-CF LDAS.rc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/build/Linux/etc /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run creating gcm style batch Run scripts lenkf.j Experiment directory: /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2 creating batch Run scripts 11 > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run/ Check restart file: > cat cap_restart 19810101 000000 > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 153 2019-05-24 16:40 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1981/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19810101_0000 lrwxrwxrwx 1 fzeng g0620 128 2019-05-24 16:40 vegdyn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/GEOSldas_CF90_test2.vegdyn_internal_rst Check executable: > ls -l ../build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50809485 2019-05-23 12:47 ../build/Linux/bin/GEOSldas.x* CAP.rc: check the END_DATE, JOB_SGMT and NUM_SGMT. No change. MAPL_ENABLE_TIMERS: YES HISTORY.rc: tavg1_2D_lfs_Nx.resolution: 360 181, tavg1_2D_lnd_Nx.resolution: 360 181, Added to L236: 'CNCO2' , 'CATCHCN' , No longer need this below: LDAS.rc: added "IMPOSE_CT_CO2: 1" to the end (L38). Did an interactive run: > interactive.py -A sp3 -n 168 -a g0620 -X --debug > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > cp lenkf.j lenkf.j.orig > nedit lenkf.j & Add "exit" after L278 "mpirun -map-by core --mca btl ^vader -np $numprocs $GEOSBIN/GEOSldas.x" > ./lenkf.j It's running. Need to comment out the print statement I added right before calling CN_Driver in GEOS_CatchCNGridComp.F90. Compiled GEOScatchCN_GridComp and Applications/LDAS_App. /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN > /bin/cp -pr Linux exec/tranCO2_test2/. > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > ls -l ../build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50809139 2019-05-24 19:05 ../build/Linux/bin/GEOSldas.x* > interactive.py -A sp3 -n 168 -a g0620 -X --debug > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > ./lenkf.j It's running. Stopped it. > mv lenkf.j.orig lenkf.j > ./ldas_batchrun.j ========== 20190528: 1. The GEOSldas_CF90_test2 finished 5 years of simulation, but I couldn't find the log files. -rw-r--r-- 1 fzeng g0620 1471771 2019-05-24 22:04 GEOSldas_CF90_test2.tavg1_2D_lfs_Nx.monthly.198101.nc4 -rw-r--r-- 1 fzeng g0620 5671641 2019-05-24 22:04 GEOSldas_CF90_test2.tavg1_2D_lnd_Nx.monthly.198101.nc4 -rw-r--r-- 1 fzeng g0620 1468594 2019-05-25 03:04 GEOSldas_CF90_test2.tavg1_2D_lfs_Nx.monthly.198201.nc4 -rw-r--r-- 1 fzeng g0620 5601975 2019-05-25 03:04 GEOSldas_CF90_test2.tavg1_2D_lnd_Nx.monthly.198201.nc4 -rw-r--r-- 1 fzeng g0620 1476608 2019-05-25 10:08 GEOSldas_CF90_test2.tavg1_2D_lfs_Nx.monthly.198301.nc4 -rw-r--r-- 1 fzeng g0620 5638563 2019-05-25 10:07 GEOSldas_CF90_test2.tavg1_2D_lnd_Nx.monthly.198301.nc4 -rw-r--r-- 1 fzeng g0620 1473759 2019-05-25 15:12 GEOSldas_CF90_test2.tavg1_2D_lfs_Nx.monthly.198401.nc4 -rw-r--r-- 1 fzeng g0620 5629143 2019-05-25 15:12 GEOSldas_CF90_test2.tavg1_2D_lnd_Nx.monthly.198401.nc4 -rw-r--r-- 1 fzeng g0620 1472133 2019-05-25 22:04 GEOSldas_CF90_test2.tavg1_2D_lfs_Nx.monthly.198501.nc4 -rw-r--r-- 1 fzeng g0620 5623746 2019-05-25 22:04 GEOSldas_CF90_test2.tavg1_2D_lnd_Nx.monthly.198501.nc4 It took sometimes 5 hours and sometimes 7 hours to finish 1 year of simulation. Asked Weiyuan why there is no log file in rc_out. He said he doesn't know why, but suggested me to make some change to lenkf.j. > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > cp -p lenkf.j lenkf.j.orig > nedit lenkf.j & Changed L579-585 from if (-f GEOSldas_log.txt) then /bin/mv GEOSldas_log.txt $logfile endif if(-f GEOSldas_err.txt) then /bin/mv GEOSldas_err.txt $errfile endif to /bin/cp -p GEOSldas_log.txt $logfile /bin/cp -p GEOSldas_err.txt $errfile /bin/rm GEOSldas_log.txt /bin/rm GEOSldas_err.txt > cat cap_restart 19860101 000000 > ls -l ../input/restart/ total 0 lrwxrwxrwx 1 fzeng g0620 153 2019-05-26 01:31 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1986/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19860101_0000 lrwxrwxrwx 1 fzeng g0620 128 2019-05-24 16:40 vegdyn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/GEOSldas_CF90_test2.vegdyn_internal_rst > nedit CAP.rc & END_DATE: 19860106 000000 > qsub lenkf.j However, still don't see the log file in rc_out. Read lenkf.j to understand why. > nedit lenkf.j & Changed L579-585 from /bin/cp -p GEOSldas_log.txt $logfile /bin/cp -p GEOSldas_err.txt $errfile /bin/rm GEOSldas_log.txt /bin/rm GEOSldas_err.txt to /bin/cp -p $EXPDIR/scratch/GEOSldas_log.txt $logfile /bin/cp -p $EXPDIR/scratch/GEOSldas_err.txt $errfile /bin/rm $EXPDIR/scratch/GEOSldas_log.txt /bin/rm $EXPDIR/scratch/GEOSldas_err.txt > cat cap_restart 19860106 000000 > ls -l ../input/restart/ total 0 lrwxrwxrwx 1 fzeng g0620 153 2019-05-28 11:25 catchcn_internal_rst -> /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rs/ens0000/Y1986/M01/GEOSldas_CF90_test2.catchcn_internal_rst.19860106_0000 > nedit CAP.rc & END_DATE: 19860108 000000 > qsub lenkf.j Still not working. Asked Weiyuan. He told me to echo $logfile and see where the log file went. > interactive.py -A sp3 -n 168 -a g0620 -X --debug > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > nedit lenkf.j & Added "echo $logfile" to line 580. > ./lenkf.j /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out/Y1986/M01/GEOSldas_CF90_test2.ldas_log.19860108_0000z.txt /bin/cp: cannot stat `/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/scratch/GEOSldas_log.txt': No such file or directory /bin/cp: cannot stat `/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/scratch/GEOSldas_err.txt': No such file or directory /bin/rm: cannot remove `/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/scratch/GEOSldas_log.txt': No such file or directory /bin/rm: cannot remove `/discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/scratch/GEOSldas_err.txt': No such file or directory > nedit lenkf.j & Changed L579-585 from /bin/cp -p $EXPDIR/scratch/GEOSldas_log.txt $logfile /bin/cp -p $EXPDIR/scratch/GEOSldas_err.txt $errfile /bin/rm $EXPDIR/scratch/GEOSldas_log.txt /bin/rm $EXPDIR/scratch/GEOSldas_err.txt back to /bin/cp -p GEOSldas_log.txt $logfile /bin/cp -p GEOSldas_err.txt $errfile /bin/rm GEOSldas_log.txt /bin/rm GEOSldas_err.txt > ./lenkf.j /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out/Y1986/M01/GEOSldas_CF90_test2.ldas_log.19860108_0000z.txt /bin/cp: cannot stat `GEOSldas_log.txt': No such file or directory /bin/cp: cannot stat `GEOSldas_err.txt': No such file or directory /bin/rm: cannot remove `GEOSldas_log.txt': No such file or directory /bin/rm: cannot remove `GEOSldas_err.txt': No such file or directory > nedit lenkf.j & Changed /bin/cp -p GEOSldas_log.txt $logfile /bin/cp -p GEOSldas_err.txt $errfile /bin/rm GEOSldas_log.txt /bin/rm GEOSldas_err.txt to /bin/cp -p GEOSldas_log.txt ${EXPID}.ldas_log.${logYEAR}${logMON}${logDAY}_${logHour}${logMin}z.txt /bin/cp -p GEOSldas_err.txt ${EXPID}.ldas_err.${logYEAR}${logMON}${logDAY}_${logHour}${logMin}z.txt > nedit CAP.rc & END_DATE: 19860110 000000 NUM_SGMT: 1 > qsub lenkf.j Don't see the copied files under "scratch" > nedit lenkf.j & Added "echo $PWD" in line 581. > interactive.py -A sp3 -n 168 -a g0620 -X --debug > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > ./lenkf.j /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out/Y1986/M01/GEOSldas_CF90_test2.ldas_log.19860110_0000z.txt /gpfsm/dnb31/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/scratch /bin/cp: cannot stat `GEOSldas_log.txt': No such file or directory /bin/cp: cannot stat `GEOSldas_err.txt': No such file or directory > nedit CAP.rc & END_DATE: 19860112 000000 NUM_SGMT: 1 > nedit lenkf.j & Added "echo ${NUM_SGMT}" to line 582. > qsub lenkf.j Doesn't work. > nedit lenkf.j & #SBATCH --output=../scratch/GEOSldas_log_txt #SBATCH --error=../scratch/GEOSldas_err_txt /bin/cp -p GEOSldas_log_txt ${EXPID}.ldas_log.${logYEAR}${logMON}${logDAY}_${logHour}${logMin}z.txt /bin/cp -p GEOSldas_err_txt ${EXPID}.ldas_err.${logYEAR}${logMON}${logDAY}_${logHour}${logMin}z.txt > nedit CAP.rc & END_DATE: 19860113 000000 > qsub lenkf.j Still not working. > nedit lenkf.j & L74: /bin/rm -rf *.* > nedit CAP.rc & END_DATE: 19860114 000000 > qsub lenkf.j Now it's working. > nedit lenkf.j & Changed /bin/cp -p GEOSldas_log_txt ${EXPID}.ldas_log.${logYEAR}${logMON}${logDAY}_${logHour}${logMin}z.txt /bin/cp -p GEOSldas_err_txt ${EXPID}.ldas_err.${logYEAR}${logMON}${logDAY}_${logHour}${logMin}z.txt to if (-f GEOSldas_log_txt) then /bin/mv GEOSldas_log_txt $logfile endif if(-f GEOSldas_err_txt) then /bin/mv GEOSldas_err_txt $errfile endif > nedit CAP.rc & END_DATE: 19860115 000000 > qsub lenkf.j Now I see /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/rc_out/Y1986/M01/GEOSldas_CF90_test2.ldas_log.19860114_0000z.txt. One more thing to fix: the log file doesn't show the time taken by the job segment at the end. > nedit lenkf.j & ##SBATCH --output=../scratch/GEOSldas_log_txt ##SBATCH --error=../scratch/GEOSldas_err_txt Rmoved: echo $logfile echo $PWD echo ${NUM_SGMT} > nedit CAP.rc & END_DATE: 19860116 000000 > qsub lenkf.j I got the slurm-JOBID.out in the "run" directory and it has the time info at the end. Great! 2. Fixed the bug in lenkf.j.template: > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/src/Applications/LDAS_App > cp -p lenkf.j.template lenkf.j.template.orig > nedit lenkf.j.template & L7-8: ##SBATCH --output=../scratch/GEOSldas_log_txt ##SBATCH --error=../scratch/GEOSldas_err_txt L74: /bin/rm -rf *.* L580-586: if (-f GEOSldas_log_txt) then /bin/mv GEOSldas_log_txt $logfile endif if(-f GEOSldas_err_txt) then /bin/mv GEOSldas_err_txt $errfile endif 3. Test runs: > cd /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN/src/GEOSldas_GridComp/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit GEOS_CatchCNGridComp.F90 & Added a few timers: call MAPL_TimerAdd(GC, name="-compute_rc" ,RC=STATUS) VERIFY_(STATUS) call MAPL_TimerAdd(GC, name="-FPAR_scaling" ,RC=STATUS) VERIFY_(STATUS) call MAPL_TimerAdd(GC, name="-CN_Driver" ,RC=STATUS) VERIFY_(STATUS) call MAPL_TimerAdd(GC, name="-call_CATCHCN" ,RC=STATUS) VERIFY_(STATUS) Added "if (MAPL_AM_I_Root(VM)) print *, 'DTCN: ', dtcn" to L7143. > setenv ESMADIR /discover/nobackup/fzeng/offline_code/GEOSldas_m4-17_UCatchCN > source $ESMADIR/src/g5_modules > gmake install > cd ../../../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux exec/tranCO2_test2/. > cd /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/run > ls -l ../build/Linux/bin/GEOSldas.x -rwxr-xr-x 1 fzeng g0620 50817987 2019-05-28 16:14 ../build/Linux/bin/GEOSldas.x* > nedit CAP.rc & END_DATE: 19860201 000000 > qsub lenkf.j After it's finished. Checked the log file and DTCN is 5400s. Good. > nedit CAP.rc & END_DATE: 19870201 000000 JOB_SGMT: 00000600 000000 NUM_SGMT: 1 > qsub lenkf.j [Changed the requested time to 12 hours before submitting the job] 4. Processed output: clm4.5_DE720: 39, 40, 41, 42 clm4.5_Princeton_DE720: 72, 73, 74, 75, 76, 77, 78 Both reached reasonable equilibrium. Commented out "./ldas_batchrun.j" in submit_next_batch_DE720 and submit_next_batch_Princeton_DE720 under ~/Catchment/CLM4.5. Archived these cycles. ========== 20190529: 1. The GEOSldas_CF90_test2 run stopped at 19860801. It's because in CAP.rc I have only one job segment and each job segment is 6 months. It took 2.5 hours to finish 6 months of simulation with DTCN = 5400s. Times for CATCHCN TOTAL : 6247.435 RUN1 : 468.387 -RUN0 : 191.229 -SURF : 86.338 RUN2 : 5718.882 -CATCH : 5067.536 -compute_rc : 1098.059 -FPAR_scaling : 0.000 -CN_Driver : 2315.159 -call_CATCHCN : 1455.549 -ALBEDO : 50.613 GenInitTot : 69.298 --GenInitMine : 69.298 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 11.549 --GenFinalMine : 11.549 GenRecordTot : 0.038 --GenRecordMine : 0.016 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 6279.731 INITIALIZE : 71.255 RUN1 : 495.225 RUN2 : 5721.334 VEGDYN : 29.107 CATCHCN : 6250.465 GenInitTot : 71.255 --GenInitMine : 0.080 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 11.550 --GenFinalMine : 0.000 GenRecordTot : 0.652 --GenRecordMine : 0.013 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 8231.767 Initialize : 111.289 -LocStreamCreate : 5.608 Run : 8071.153 DATAATM : 1853.687 LANDPERT : 39.619 LAND : 6280.705 ENSAVG : 0.296 GenInitTot : 73.737 --GenInitMine : 0.045 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 11.616 --GenFinalMine : 0.000 GenRecordTot : 1.897 --GenRecordMine : 0.013 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 243.741 Initialize : 1.425 Finalize : 0.076 Run : 242.316 --Couplers : 123.312 --I/O : 174.069 ----IO Create : 9.896 ----IO Write : 164.161 -----IO Post : 4.322 -----IO Wait : 19.059 -----IO Write : 150.151 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 2. Checked LAI, GPP and NEE in /discover/nobackup/fzeng/Catchment/M2n5P/GEOSldas_CF90_test2/output/CF0090x6C/cat/ens0000/Y1986/M07/GEOSldas_CF90_test2.tavg1_2D_lnd_Nx.monthly.198607.nc4. The patterns look correct. 3. Changed DTCN to 3 hours and see if it's faster. > nedit CAP.rc & END_DATE: 19870201 000000 JOB_SGMT: 00000600 000000 NUM_SGMT: 1 > nedit LDAS.rc & DTCN: 10800 Checked the log file and made sure DTCN is 10800. It took 2 hours and 7 minutes to finish 6 months of simulation. Times for CATCHCN TOTAL : 4970.829 RUN1 : 473.453 -RUN0 : 193.819 -SURF : 85.628 RUN2 : 4450.814 -CATCH : 3789.061 -compute_rc : 965.896 -FPAR_scaling : 0.000 -CN_Driver : 1171.756 -call_CATCHCN : 1533.686 -ALBEDO : 51.486 GenInitTot : 66.257 --GenInitMine : 66.257 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 4.304 --GenFinalMine : 4.304 GenRecordTot : 0.038 --GenRecordMine : 0.015 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 5000.001 INITIALIZE : 67.610 RUN1 : 497.689 RUN2 : 4453.328 VEGDYN : 25.967 CATCHCN : 4973.900 GenInitTot : 67.610 --GenInitMine : 0.088 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 4.304 --GenFinalMine : 0.000 GenRecordTot : 0.675 --GenRecordMine : 0.013 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 6947.362 Initialize : 102.523 -LocStreamCreate : 5.073 Run : 6801.225 DATAATM : 1850.497 LANDPERT : 39.943 LAND : 5000.984 ENSAVG : 0.286 GenInitTot : 69.768 --GenInitMine : 0.025 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.055 --GenFinalMine : 0.000 GenRecordTot : 1.914 --GenRecordMine : 0.014 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 218.604 Initialize : 1.354 Finalize : 0.027 Run : 217.221 --Couplers : 125.212 --I/O : 133.261 ----IO Create : 10.090 ----IO Write : 123.161 -----IO Post : 4.348 -----IO Wait : 19.416 -----IO Write : 109.190 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 4. Tried turning on FPAR and ALBEDO scaling. In GEOSldas_CF90_test2.exec: # Scale CATCHCN ALBEDO and FPAR # 0--NO # 1-- Scale albedo to match interannually varying MODIS NIRDF and VISDF anomaly # 2-- Scale albedo to match interannually varying MODIS NIRDF and VISDF plus FPAR to match MODIS FPAR CDF SCALE_ALBFPAR: 0 > nedit LDAS.rc & SCALE_ALBFPAR: 2 DTCN is 10800. > nedit CAP.rc & END_DATE: 19870801 000000 JOB_SGMT: 00000600 000000 NUM_SGMT: 1 > qsub lenkf.j 5. Created ~/python/clm4-to-clm4.5/compare_ts_GEOSldas.py to analyze output from GEOSldas simulations. ========== 20190530: 1. With SCALE_ALBFPAR: 2 DTCN is 10800. It took 3 hours 14 minutes (about 3.25 hours) to finish 6 months of simulation. Times for CATCHCN TOTAL : 7640.303 RUN1 : 467.808 -RUN0 : 190.885 -SURF : 86.537 RUN2 : 7101.779 -CATCH : 6444.884 -compute_rc : 1066.774 -FPAR_scaling : 3155.137 -CN_Driver : 1123.399 -call_CATCHCN : 1476.404 -ALBEDO : 69.863 GenInitTot : 65.451 --GenInitMine : 65.451 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.211 --GenFinalMine : 5.211 GenRecordTot : 0.037 --GenRecordMine : 0.015 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for LAND TOTAL : 7667.869 INITIALIZE : 66.792 RUN1 : 491.120 RUN2 : 7104.204 VEGDYN : 24.991 CATCHCN : 7643.256 GenInitTot : 66.792 --GenInitMine : 0.070 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.212 --GenFinalMine : 0.000 GenRecordTot : 0.655 --GenRecordMine : 0.012 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for GCS TOTAL : 9646.152 Initialize : 103.048 -LocStreamCreate : 6.746 Run : 9502.135 DATAATM : 1891.251 LANDPERT : 39.354 LAND : 7668.785 ENSAVG : 0.278 GenInitTot : 68.603 --GenInitMine : 0.050 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 5.313 --GenFinalMine : 0.000 GenRecordTot : 1.874 --GenRecordMine : 0.013 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 Times for HIST TOTAL : 217.000 Initialize : 1.358 Finalize : 0.024 Run : 215.614 --Couplers : 121.052 --I/O : 133.999 ----IO Create : 9.832 ----IO Write : 124.155 -----IO Post : 4.333 -----IO Wait : 18.836 -----IO Write : 110.153 -ParserRun : 0.000 GenInitTot : 0.000 --GenInitMine : 0.000 GenRunTot : 0.000 --GenRunMine : 0.000 GenFinalTot : 0.000 --GenFinalMine : 0.000 GenRecordTot : 0.000 --GenRecordMine : 0.000 GenRefreshTot : 0.000 --GenRefreshMine : 0.000 ========== 20190531: 1. Plotted time series and mean seasonal cycles of nfire, burn, closs, sfmc and rzmc of a few grid cells in S America using ~/Catchment/CLM4.5/check_fire_bug_fix.gs. 2. Prepared meeting slides. 3. Read Fang Li's 2012 paper about fire parameterization. In the calculation of the effect of soil moisture on fuel cumbustibility, the soil wetness (for surface 5mm according to the code) has to be fraction of water holding capacity. In Catchment-CN4.5, this is SWSRF1, SWSRF2 and SWSRF4 computed in PARTITION. It looks to me they are fraction of water holding capacity since they are in range of 0 to 1 in Subroutine Partition. Double check with Randy. ========== 20190603: 1. Randy suggests an experiment to make the soil water saturate and see if this reduces the burned area in SH South America. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp -p CNFireMod.F90 CNFireMod.F90.restore 2. Exp #1: set surface soil moisture (wf) to be saturated in the fire model > nedit CNFireMod.F90 & L653-654: ! m = max(0._r8,wf(c)) m = max(0._r8,1.) This sets surface soil moisture to be saturate. > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_wetSF > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_wetSF/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_wetSF_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_wetSF_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_wetSF/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetSF_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetSF_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-06-03 14:23 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetSF_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetSF_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800102_020730z. > qsub lenkf.0.j 3. Exp #2: set root zone soil moisture (wf2) to be saturated in the fire model > nedit CNFireMod.F90 & L613-614: ! baf_peatf(c) = boreal_peatfire_c/secsphr*exp(-SHR_CONST_PI*(max(wf2(c),0._r8)/0.3_r8))* & ! removed *dt, see clm4_6_00, fzeng, 7 May 2019 baf_peatf(c) = boreal_peatfire_c/secsphr*exp(-SHR_CONST_PI*(max(1.0,0._r8)/0.3_r8))* & L654: m = max(0._r8,wf(c)) Compiled following 2 above. > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_wetRZ > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_wetRZ/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_wetRZ_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_wetRZ_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_wetRZ/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetRZ_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetRZ_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-06-03 14:53 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetRZ_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_wetRZ_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_112230z. > qsub lenkf.0.j Restored the file: > cp CNFireMod.F90.restore CNFireMod.F90 ========== 20190604: 1. Processed the output: > cd ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_wetSF_DE720 > grid_restore_hdeg clm4.5_wetRZ_DE720 Analyzed the output: > cp clm4.5_DE720.ctl clm4.5_wetSF_DE720.ctl > cp clm4.5_DE720.ctl clm4.5_wetRZ_DE720.ctl > nedit clm4.5_wetSF_DE720.ctl clm4.5_wetRZ_DE720.ctl & > nedit check_fire_bug_fix.gs & ========== 20190606: 1. Read CLM4.5 Tech Note about accelerated decomposition. ========== 20190607: 1. Created ~/python/soil_wetness_fire.py and used it to plot equation 9 in Li et al., 2012 and change the coefficient and power scale to see the sensitivity. ========== 20190613: 1. For the 4 fire experiments: modified code and set up experiments (1) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-2. *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_coeff2 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_coeff2/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_sfmc_coeff2_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_sfmc_coeff2_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_sfmc_coeff2/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff2_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff2_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-13 12:34 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff2_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff2_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_180730z. > qsub lenkf.0.j (2) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-5. *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_coeff5 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_coeff5/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_sfmc_coeff5_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_sfmc_coeff5_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_sfmc_coeff5/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff5_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff5_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-13 12:49 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff5_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_coeff5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800103_020730z. > qsub lenkf.0.j (3) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**1.5)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_power1pt5 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_power1pt5/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_sfmc_power1pt5_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_sfmc_power1pt5_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_sfmc_power1pt5/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-13 13:16 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_090730z. > qsub lenkf.0.j (4) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2.5)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_power2pt5 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_power2pt5/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_sfmc_power2pt5_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_sfmc_power2pt5_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_sfmc_power2pt5/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-13 13:23 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_060000z. > qsub lenkf.0.j 2. For the 4 fire experiments: processed output > cd ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_sfmc_coeff2_DE720 > grid_restore_hdeg clm4.5_sfmc_coeff5_DE720 > grid_restore_hdeg clm4.5_sfmc_power1pt5_DE720 > grid_restore_hdeg clm4.5_sfmc_power2pt5_DE720 Created the GrADS control files for them. ========== 20190614: 1. Analyzed the output of the 4 fire experiments using ~/Catchment/CLM4.5/check_fire_bug_fix.gs The results from clm4.5_sfmc_coeff2_DE720 and clm4.5_sfmc_coeff5_DE720 are as expected, but those from clm4.5_sfmc_power1pt5_DE720 and clm4.5_sfmc_power2pt5_DE720 are not. 2. Repeat clm4.5_sfmc_power1pt5_DE720 and clm4.5_sfmc_power2pt5_DE720 following (3) and (4) in 1 on 20190613: (3) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**1.5)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_power1pt5 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_power1pt5/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/run > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-14 10:43 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x* > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 80 2019-06-13 13:19 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power1pt5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_040730z. > qsub lenkf.0.j (4) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2.5)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_power2pt5 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_power2pt5/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/run > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-14 10:52 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/build/Linux/bin/LDASsaCN_mpi.x* > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 80 2019-06-13 13:25 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_power2pt5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_060730z. > qsub lenkf.0.j 3. The clm4.5_sfmc_power1pt5_DE720 and clm4.5_sfmc_power2pt5_DE720 are finished. Checked and made sure the output files in "cat_avg" are new. Processed output > cd ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_sfmc_power1pt5_DE720 > grid_restore_hdeg clm4.5_sfmc_power2pt5_DE720 Checked and made sure the processed output files in "post" are new. Checked the GrADS control files. Found that the control file for clm4.5_sfmc_power1pt5_DE720 should be for clm4.5_sfmc_power2pt5_DE720, and the one for clm4.5_sfmc_power2pt5_DE720 is actually for clm4.5_wetRZ_DE720 (I haven't modified it!). Correct these! The GrADS control files for clm4.5_sfmc_coeff2_DE720 and clm4.5_sfmc_coeff5_DE720 are correct. Run ~/Catchment/CLM4.5/check_fire_bug_fix.gs again. 4. Another fire experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! fire_m = exp(-SHR_CONST_PI *(m/0.69_r8)**2)*(1.0_r8 - max(0._r8, & fire_m = exp(-SHR_CONST_PI *(m/0.5_r8)**4)*(1.0_r8 - max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8))))* & min(1._r8,exp(SHR_CONST_PI*(forc_t(g)-SHR_CONST_TKFRZ)/10._r8)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_sfmc_thetaEpt5_power4 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_sfmc_thetaEpt5_power4/. Setup the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & D0.5_cn.exe: exp_id = clm4.5_sfmc_thetaEpt5_power4_DE720 start_time = 1980-01-01-00-00-00 end_time = 1981-01-01-00-00-00 restart_id = clm4.5_DE720 D0.5.bat: job-name = clm4.5_sfmc_thetaEpt5_power4_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_sfmc_thetaEpt5_power4/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_thetaEpt5_power4_DE720/run > nedit lenkf.0.j & #SBATCH --time=02:00:00 -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_thetaEpt5_power4_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-14 16:02 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_thetaEpt5_power4_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_42/output > rm rst_clm4.5 > cd ../../run/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_sfmc_thetaEpt5_power4_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_050730z. > qsub lenkf.0.j Restore the file: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp CNFireMod.F90.restore CNFireMod.F90 ========== 20190617: 1. Processed and analyzed the clm4.5_sfmc_thetaEpt5_power4_DE720 data. > cd ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_sfmc_thetaEpt5_power4_DE720 Created the GrADS control files for them. Analyzed the output of the fire experiment using ~/Catchment/CLM4.5/check_fire_bug_fix.gs ========== 20190618: 1. For nitrif_denitrif: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5 > > /bin/cp -pr LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_nitrif ========== 20190625: 1. For debugging fire code: Used ~/Catchment/princeton_tiles_location.f90 to look for grids of forest and grassland. It shows that the PFTs at (0.25N, 20.25E) is 4  4 18 19 0.9892 0.0000 0.0000 0.0108. According to the fire code: if (trotr1_col(c)+trotr2_col(c)>0.6_r8) then farea_burned(c)=min(1.0_r8,baf_crop(c)+baf_peatf(c)) else This grid cell should have zero burned area outside the agricultural burn month (ABM) because (1) Agriculture is only burned in the ABM. (2) I think the peat fraction is 0 here. (3) This grid cell is 98.92% BET. However, the burned area is >5% in multiple months of 2006. Why? Run the code on debugger to figure out. Compiled the code with "BOPT=g": > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake clean > gmake install BOPT=g > cd ../../../GEOSlana_GridComp > gmake install BOPT=g > cd ../../Applications/LDAS_App/ > gmake install BOPT=g > cd ../../.. > /bin/cp -pr Linux exec/clm4.5_debug/. > cd src > gmake realclean > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > nedit driver_inputs_0.5D_GLOBAL.nml & minlon = 20. ! min longitude maxlon = 20.5 ! max longitude minlat = 0. ! min latitude maxlat = 0.5 ! max latitude > cd ../input/restart > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_43/output Don't have 20060101 restart file in clm4.5_DE720_43. Have to start from 1980. > cd ../../run/ > lenkf.0.i ========== 20190626: 1. Found a bug in the fire code. cwtgcell is very close to zero, which is wrong. This causes trotr1_col and trotr2_col to be very close to 0. CNFireMod.F90 195: real(r8), pointer :: cwtgcell(:) ! column's weight relative to corresponding gridcell In original CLM4.5 main/subgridAveMod.F90: 833: real(r8), pointer :: wtgcell(:) ! weight of pfts relative to gridcells 850: wtgcell =>pft%wtgcell 1224: real(r8), pointer :: wtgcell(:) ! weight of columns relative to gridcells 1238: wtgcell =>col%wtgcell Currently in CN_DriverMod.F90: wtgcell => pft%wtgcell pwtgcell(p) = fveg(nc,nv,nz)*wtzone(nc,nz) ! PFT weight in catchment tile Follow this to set cwtgcell in CN_DriverMod.F90. real, pointer :: cwtgcell(:) !weight of columns relative to gridcells cwtgcell => col%wtgcell cwtgcell(n) = wtzone(nc,nz) ! weight of columns relative to gridcells Compiled the code with "BOPT=g": > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install BOPT=g > cd .. > /bin/cp -pr Linux exec/clm4.5_debug/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_debug/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 111472141 2019-06-26 13:30 ../build/Linux/bin/LDASsaCN_mpi.x* > /bin/rm ../output/global/rc_out/* > lenkf.0.i This grid cell: PFT frac 4 0.989 19 0.0108 trotr1_col is now 0.0989 0.445 0.445 In our CNFireMod.F90: if( ivt(p) == nbrdlf_evr_trp_tree .and. wtcol(p) .gt. 0._r8 )then trotr1_col(c)=trotr1_col(c)+wtcol(p)*cwtgcell(c) end if if( ivt(p) == nbrdlf_dcd_trp_tree .and. wtcol(p) .gt. 0._r8 )then trotr2_col(c)=trotr2_col(c)+wtcol(p)*cwtgcell(c) end if This is consistent with clm4_6_00 CNFireMod.F90. However, since our cwtgcell is 0.1, 0.45 and 0.45, our trotr1_col(c) + trotr2_col(c) will never be >0.45 and thus will never be >0.6. Because of this, the bug doesn't affect model output. Three cycles (44 to 46) are done for clm4.5_DE720 after this bug fix. GPP and NEE values show that the model is still in equilibrium. In clm4_6_00 CNFireMod.F90: trotr1_col => cps%trotr1_col , & ! Input: [real(r8) (:)] pft weight of BET on the gridcell (0-1) trotr2_col => cps%trotr2_col , & ! input: [real(r8) (:)] pft weight of BDT on the gridcell (0-1) Why use pft weight of BET on the gridcell instead of pft weight of BET on the column, since the burned area is calculated for each column? Compiled the code without "BOPT=g": > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install > cd .. > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-06-26 14:59 ../build/Linux/bin/LDASsaCN_mpi.x* > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_120730z. > ./ldas_batchrun.j 2. Used ~/Catchment/princeton_tiles_location.f90 to look for grids of forest and grassland and used ~/Catchment/CLM4.5/check_fire_bug_fix.gs to plot the vegc, burn, and closs time series. 25.75 -6.75 1.00000 6 6 10 11 0.5807 0.0000 0.0000 0.4193 4 0.0000 0.0000 0.0000 CNFireMod.F90 L642-651: if( cropf_col(c) .lt. 1.0 )then ! see clm4_6_00 and Li et al., BG 2014, fzeng, 7 May 2019 if (trotr1_col(c)+trotr2_col(c)>0.6_r8) then farea_burned(c)=min(1.0_r8,baf_crop(c)+baf_peatf(c)) else fuelc(c) = totlitc(c)+totvegc_col(c)-rootc_col(c)-fuelc_crop(c)*cropf_col(c) do j = 1, nlevdecomp fuelc(c) = fuelc(c)+decomp_cpools_vr(c,j,i_cwd) * dzsoi_decomp(j) end do fuelc(c) = fuelc(c)/(1._r8-cropf_col(c)) ========== 20190627: 1. Created ~/python/clm4-to-clm4.5/tile2grid_fveg_frac.py to plot the major vegetation types and the corresponding fractions. 2. Checked the avaible output of the current clm4.5_DE720 cycle. ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_DE720 Compared the fire output to that in the 43rd cycle. Not much change. 3. Used ~/Catchment/princeton_tiles_location.f90 to look for grids of forest and grassland and used ~/Catchment/CLM4.5/check_fire_bug_fix.gs to plot the vegc, burn, and closs time series. Used ~/python/clm4-to-clm4.5/tile2grid_fveg_frac.py to find grid cells 100% covered by forest, grass, or crop. % cover type lat lon (100.0, 15.0, -27.75, 15.25) (100.0, 15.0, -27.25, 15.25) (100.0, 17.0, -21.75, 22.75) (100.0, 17.0, -21.75, 23.25) (100.0, 15.0, -19.25, 12.25) (100.0, 15.0, -18.75, 12.75) (100.0, 15.0, -18.25, 11.75) (100.0, 15.0, -18.25, 12.25) (100.0, 15.0, -17.75, 11.75) (100.0, 17.0, -17.75, 15.25) (100.0, 15.0, -17.25, 11.75) (100.0, 15.0, -16.75, 11.75) (100.0, 17.0, -9.25, 12.75) (100.0, 4.0, -2.25, 24.75) (100.0, 4.0, -1.75, 24.75) (100.0, 19.0, -1.25, 32.75) No fire in 1997-2016 ... Picked some of these and plotted the vegc, burn, and closs time series using ~/Catchment/CLM4.5/check_fire_bug_fix.gs. For the ~10 grid cells 100% covered by crop that I picked, all have 0 burned area. (100.0, 17.0, -9.25, 12.75) ========== 20190628: 1. Read Fang Li's 2012, 2013 and 2014 papers to understand how the fire C emissions are computed. 2. Read the CASA-GFED3 code about combustion completeness. ========== 20190701: 1. Processed output using ~/python/clm4-to-clm4.5/compare_ts.py: clm4.5_DE720: 44 and 45 2. Made more plots using ~/Catchment/CLM4.5/check_fire_bug_fix.gs. ========== 20190703: 1. Organized some files: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp -p compute_rc.F90 compute_rc.F90.restore 2. For an experiment: (1) Code modification and compile: > nedit CNFireMod.F90 & L469-476: if( ivt(p) == nbrdlf_evr_trp_tree .and. wtcol(p) .gt. 0._r8 )then ! trotr1_col(c)=trotr1_col(c)+wtcol(p)*cwtgcell(c) trotr1_col(c)=trotr1_col(c)+wtcol(p) !*cwtgcell(c) end if if( ivt(p) == nbrdlf_dcd_trp_tree .and. wtcol(p) .gt. 0._r8 )then ! trotr2_col(c)=trotr2_col(c)+wtcol(p)*cwtgcell(c) trotr2_col(c)=trotr2_col(c)+wtcol(p) !*cwtgcell(c) end if > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake clean > gmake install > cd ../../../GEOSlana_GridComp/ > gmake clean > gmake install > cd ../../Applications/LDAS_App/ > gmake clean > gmake install > cd ../../../ > mkdir -p exec/clm4.5_Kn11_Ndiags_fire > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire/. (2) Experiment setup: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/ > rm -rf clm4.5_fire_DE720_keep drwxr-xr-x 6 fzeng g0620 512 2019-05-09 17:21 clm4.5_fire_DE720_keep/ clm4.5_fire_DE720_keep was the test run after I came back from NCAR and fixed the fire bugs. No longer needed. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-03 16:16 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_44/output > cd ../../run/ > cp ../../clm4.5_DE720/run/ldas_batchrun.j . > nedit lenkf.0.j & -end_year 1988 -end_month 1 -end_day 1 -end_hour 0 -end_min 0 -end_sec 0 \ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_040730z. > ./ldas_batchrun.j ========== 20190705: 1. Created and used ~/python/clm4-to-clm4.5/plot_GFED4_fields.py to plot GFED4.1s burned area fraction and fire C emissions. 2. Processed output using ~/python/clm4-to-clm4.5/compare_ts.py: clm4.5_DE720: 46 3. The clm4.5_fire_DE720 run I set up on 20190703 wrote output every 3 hourly. The output files filled up my disk space. Need to remove it and re-do. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > cp ../clm4.5_fire_DE720/run/D0.5.bat . > cp ../clm4.5_fire_DE720/run/D0.5_cn.exe . > cd .. > rm -rf clm4.5_fire_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire/Linux/bin > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > ls -l /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-03 16:16 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_45/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_DE720/run/ldas_batchrun.j . > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_040730z. > ./ldas_batchrun.j 4. Created ~/python/clm4-to-clm4.5/plot_GFED3_fields.py and ~/python/clm4-to-clm4.5/plot_CatCN_AGB.py to plot AGB of CASA-GFED3 and Catchment-CN4.5. ========== 20190708: 1. The clm4.5_fire_DE720 run is finished. Processed the output: ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_fire_DE720 Modified ~/Catchment/CLM4.5/check_fire_bug_fix.gs and plotted fire output of clm4.5_fire_DE720. 2. For another experiment: (1) Code modification and compile: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & L469-476: if( ivt(p) == nbrdlf_evr_trp_tree .and. wtcol(p) .gt. 0._r8 )then ! trotr1_col(c)=trotr1_col(c)+wtcol(p)*cwtgcell(c) trotr1_col(c)=trotr1_col(c)+wtcol(p) !*cwtgcell(c) end if if( ivt(p) == nbrdlf_dcd_trp_tree .and. wtcol(p) .gt. 0._r8 )then ! trotr2_col(c)=trotr2_col(c)+wtcol(p)*cwtgcell(c) trotr2_col(c)=trotr2_col(c)+wtcol(p) !*cwtgcell(c) end if L646-648: ! if (trotr1_col(c)+trotr2_col(c)>0.6_r8) then if (trotr1_col(c)>0.6_r8) then farea_burned(c)=min(1.0_r8,baf_crop(c)+baf_peatf(c)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../../ > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire/. Override the clm4.5_fire_DE720 done over the weekend: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-08 16:09 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ls -l lrwxrwxrwx 1 fzeng g0620 80 2019-07-05 15:14 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_45/output/ > cd ../../run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_DE720 \ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_030730z. > ./ldas_batchrun.j ========== 20190710: 1. Processed the available output of clm4.5_fire_DE720. Plotted burned area and fire C emissions using ~/Catchment/CLM4.5/check_fire_bug_fix.gs 2. Another fire experiment: only at a grid cell in northern Australia > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & L656: ! m = max(0._r8,wf(c)) m = max(0._r8,0.) L666: if ( wtlf(c) > 0.0_r8 )then ! spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & spread_m = (1.0_r8 - max(0._r8,min(1._r8,(0./wtlf(c)-0.3_r8)/ & (0.7_r8-0.3_r8))))*(1.0-max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) else spread_m = 0.0_r8 end if > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_drySM > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_drySM/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_drySM/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720_drySM/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-10 15:31 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > nedit driver_inputs_0.5D_GLOBAL.nml & minlon = 133. ! min longitude maxlon = 133.5 ! max longitude minlat = -13.5 ! min latitude maxlat = -13. ! max latitude This is in northern Australia. out_select_ensavg%tile%inst = .false. out_select_ensavg%tile%xhourly = .false. out_select_ensavg%tile%daily = .false. out_select_ensavg%tile%pentad = .false. out_select_ensavg%tile%monthly = .true. out_select_ensavg%grid%inst = .false. out_select_ensavg%grid%xhourly = .false. out_select_ensavg%grid%daily = .false. out_select_ensavg%grid%pentad = .false. out_select_ensavg%grid%monthly = .false. > nedit lenkf.*.j & #SBATCH --ntasks=1 mpirun -np 1 \ [for all job files] > cp ../../clm4.5_fire_DE720/run/ldas_batchrun.j . > nedit ldas_batchrun.j & Deleted the lines for lenkf.3.j and lenkf.4.j since we don't have that here. > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_060730z. > ./ldas_batchrun.j ========== 20190711: 1. Created ~/python/clm4-to-clm4.5/plot_GFED3_maps.py and plotted GFED3 FTC and FHC maps. 2. Searched for other vegetation maps. ========== 20190712: 1. Created ~/python/clm4-to-clm4.5/plot_MODIS_landcover.py and plotted the land cover map in 1. 2. Searched for papers about fraction tree cover. 3. Another fire experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp CNFireMod.F90.trotr CNFireMod.F90 > cp -p CNFireMod.F90 CNFireMod.F90.restore [use the modified CNFireMod.F90 as default since the modification will be permanent] > nedit CN_DriverMod.F90 & ! data fsr_pfx / 0., 0.4, 0.43, 0.43, 0.4, 0.4, 0.4, 0.4, 0.4, 0.46, 0.46, 0.46, 0.46, 0.60, 0.60, 0.60, 0.60, 0.60, 0.60, 0.60/ data fsr_pfx / 0., 0.4, 0.43, 0.43, 0.4, 0.4, 0.53, 0.4, 0.4, 0.46, 0.46, 0.46, 0.46, 0.60, 0.60, 0.60, 0.60, 0.60, 0.60, 0.60/ > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire/. Re-do the clm4.5_fire_DE720 run starting from 1996 using the new executable: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-12 16:09 ../build/Linux/bin/LDASsaCN_mpi.x* > cp ldas_batchrun.j ldas_batchrun_part.j > nedit ldas_batchrun_part.j & #!/bin/bash -f jobid0=$(echo $(sbatch lenkf.2.j) | cut -d' ' -f 4) echo $jobid0 jobid1=$(echo $(sbatch --dependency=afterok:$jobid0 lenkf.3.j) | cut -d' ' -f 4) echo $jobid1 jobid2=$(echo $(sbatch --dependency=afterok:$jobid1 lenkf.4.j) | cut -d' ' -f 4) echo $jobid2 > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.2.j It's running. Stopped it at date_time_new 19960101_090730z. > ./ldas_batchrun_part.j ========== 20190716: 1. Processed and analyzed the output of the new clm4.5_fire_DE720 experiment. ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_fire_DE720 ~/Catchment/CLM4.5 > grads -lc "run check_fire_bug_fix.gs" 2. Created ~/python/clm4-to-clm4.5/tile2grid_peat_frac.py and used it to plot peat fraction, GDP etc. 3. Documented in /discover/nobackup/fzeng/clm4-to-clm4.5/notes/notes why we allow vegetation type 6 (deciduous broadleaf tropical trees) to brun. ========== 20190717: 1. Another fire experiment: use a very large lightning rate (maximum of the input data) globally > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp -p CN_DriverMod.F90 CN_DriverMod.F90.restore [use fsr=0.53 for veg type 6 permanently] > nedit CNFireMod.F90 & ! ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) ! our forc_lnfm is counts/km2/day, so divided by 24 to get per hr, fzeng, 7 May 2019 ig = (lh+0.6/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_maxLNFM > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_maxLNFM/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_maxLNFM/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720_maxLNFM/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-17 15:57 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire_DE720/run/ldas_batchrun.j . > nedit ldas_batchrun.j & Deleted the lines for lenkf.3.j and lenkf.4.j since we don't have that here. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720_maxLNFM/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_230730z. > ./ldas_batchrun.j ========== 20190718: 1. The clm4.5_fire_DE720_maxLNFM run is ongoing. Processed available output: > cd ~/Catchment/CLM4.5 Modified grid_restore_hdeg.f90 to only process 1997-2016 output and compiled it. > grid_restore_hdeg clm4.5_fire_DE720_maxLNFM Created clm4.5_maxLNFM_DE720.ctl. Modified check_fire_bug_fix.gs and used it to do some preliminary analysis. 2. Fire experiment: set wf (similar to surface soil moisture) to 0 in CNFireMod.F90. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made yesterday. L656: ! m = max(0._r8,wf(c)) m = 0. > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_drySF > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_drySF/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_drySF/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_drySF_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-18 11:54 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire_DE720/run/ldas_batchrun.j . > nedit ldas_batchrun.j & Deleted the lines for lenkf.3.j and lenkf.4.j since we don't have that here. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_drySF_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_103000z. > ./ldas_batchrun.j 3. Fire experiment: set wf2 (similar to root zone soil moisture) to 0 in CNFireMod.F90. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made at 2 above today. L615: ! baf_peatf(c) = boreal_peatfire_c/secsphr*exp(-SHR_CONST_PI*(max(wf2(c),0._r8)/0.3_r8))* & ! removed *dt, see clm4_6_00, fzeng, 7 May 2019 baf_peatf(c) = boreal_peatfire_c/secsphr*exp(-SHR_CONST_PI*(max(0.,0._r8)/0.3_r8))* & > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_dryRZ > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_dryRZ/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_dryRZ/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_dryRZ_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-18 12:17 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire_DE720/run/ldas_batchrun.j . > nedit ldas_batchrun.j & Deleted the lines for lenkf.3.j and lenkf.4.j since we don't have that here. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_dryRZ_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_220730z > ./ldas_batchrun.j 4. Fire experiment: set btran to 0 in CNFireMod.F90. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made at 3 above today. L666: ! spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & spread_m = (1.0_r8 - max(0._r8,min(1._r8,(0./wtlf(c)-0.3_r8)/ & > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_0btran > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_0btran/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_0btran/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_0btran_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-18 12:30 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire_DE720/run/ldas_batchrun.j . > nedit ldas_batchrun.j & Deleted the lines for lenkf.3.j and lenkf.4.j since we don't have that here. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_0btran_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_091500z > ./ldas_batchrun.j Restored the change made above. ========== 20190719: 1. Analyzed the output of the 3 fire experiments set up yesterday. > cd ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_drySF_DE720 > grid_restore_hdeg clm4.5_dryRZ_DE720 > grid_restore_hdeg clm4.5_0btran_DE720 ========== 20190724: 1. Fix a bug in the fire model: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & L660-661: ! lh = 0.0035_r8*6.8_r8*hdmlf**(0.43_r8)/30._r8/24._r8 lh = 0.0035_r8*6.8_r8*hdmlf**(0.43_r8)/(30._r8*24._r8*secsphr/dt) > cp -p CNFireMod.F90 CNFireMod.F90.restore > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire1_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-24 16:26 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire_DE720/run/ldas_batchrun.j . > nedit ldas_batchrun.j & Deleted the lines for lenkf.3.j and lenkf.4.j since we don't have that here. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire1_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970102_030730z. > ./ldas_batchrun.j 2. Fire experiment: set the lower threshold of btran to 0.4 in CNFireMod.F90. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made at 3 above today. L668-673: ! spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & ! (0.7_r8-0.3_r8))))*(1.0-max(0._r8, & ! min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.4_r8)/ & (0.7_r8-0.4_r8))))*(1.0-max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_btranLOWpt4 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_btranLOWpt4/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_btranLOWpt4/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_btranLOWpt4_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-24 16:34 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire1_DE720/run/ldas_batchrun.j . > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_btranLOWpt4_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_190730z > ./ldas_batchrun.j 3. Fire experiment: set the upper threshold of btran to 0.8 in CNFireMod.F90. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made at 3 above today. L668-673: ! spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & ! (0.7_r8-0.3_r8))))*(1.0-max(0._r8, & ! min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & (0.8_r8-0.3_r8))))*(1.0-max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_btranUPpt8 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_btranUPpt8/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_btranUPpt8/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_btranUPpt8_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-24 16:44 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire1_DE720/run/ldas_batchrun.j . > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_btranUPpt8_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_140730z > ./ldas_batchrun.j 4. Fire experiment: set the lower threshold of btran to 0.4 and the upper threshold of btran to 0.8 in CNFireMod.F90. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made at 3 above today. L668-673: ! spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & ! (0.7_r8-0.3_r8))))*(1.0-max(0._r8, & ! min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.4_r8)/ & (0.8_r8-0.4_r8))))*(1.0-max(0._r8, & min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_btranLpt4Upt8 > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_btranLpt4Upt8/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_btranLpt4Upt8/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_btranLpt4Upt8_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-24 16:52 ../build/Linux/bin/LDASsaCN_mpi.x*s > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire1_DE720/run/ldas_batchrun.j . > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_btranLpt4Upt8_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_200730z > ./ldas_batchrun.j ========== 20190729: 1. Analyzed the output of the clm4.5_fire1_DE720 run after bug fix and the 3 fire experiments set up last Wednesday. > cd ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_fire1_DE720 > grid_restore_hdeg clm4.5_btranLOWpt4_DE720 > grid_restore_hdeg clm4.5_btranUPpt8_DE720 > grid_restore_hdeg clm4.5_btranLpt4Upt8_DE720 2. Continued to spin up clm4.5_DE720 after these permanent changes: (1) Allow type 6 (broadleaf deciduous tropical tree) to burn: CNFireMod.F90 (2) Change fire spread rate (fsr) of type 6 from 0.4 to 0.53 since this type in Africa actually is mostly grassland: CN_DriverMod.F90 (3) Bug fix in CNFireMod.F90: NOTE - THIS CHANGE WAS RESTORED ON 20190731. L660-661: ! lh = 0.0035_r8*6.8_r8*hdmlf**(0.43_r8)/30._r8/24._r8 lh = 0.0035_r8*6.8_r8*hdmlf**(0.43_r8)/(30._r8*24._r8*secsphr/dt) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp CNFireMod.F90.restore CNFireMod.F90 [restore CNFireMod.F90] > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71238786 2019-07-29 15:19 ../build/Linux/bin/LDASsaCN_mpi.x* > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 80 2019-07-05 10:50 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_46/output/ Good. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800101_080730z > ./ldas_batchrun.j ========== 20190730: 1. Used ~/python/clm4-to-clm4.5/tile2grid_fveg_frac.py to locate the grid cells dominated by grass. fveg ityp lat lon (100.0, 17.0, -21.75, 22.75) (100.0, 17.0, -21.75, 23.25): results very close to the first grid (85.72, 14.0, 49.25, -105.25) (80.20, 14.0, 50.25, 64.75) Plotted burned area of these grids using ~/Catchment/CLM4.5/check_fire_bug_fix.gs 2. Read Prentice and Mackerras, 1977 and Latham and Schlieter, 1989 to understand the fire model. 3. Added the diagnostic "fuelc" to the output of Catchment-CN4.5. CNFireMod.F90: fuelc => ccs%fuelc Modified these files to add this diagnostic: CN_DriverMod.F90 [follow 'gross_nmin'] process_cn.F90 [follow 'gross_nmin'] clsm_ensdrv_out_routines.F90 [follow 'gross_nmin'] catch_types.F90 [update 'N_cn_diagn' from 53 to 54 and follow 'gross_nmin'] clsm_ensdrv_glob_param.F90 [change 'N_CatchCN_OutFields = 58' to 'N_CatchCN_OutFields = 59'] Updated the corresponding restore files. Compiled: Used kn = 0.11 in compute_rc.F90. After adding a new diagnostic, the model should be rebuilt from a clean state: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install |& tee make.install.log > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > /bin/cp -pr Linux/ exec/clm4.5_Kn11/. > ls -l exec/clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-07-30 16:04 exec/clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x* Once the current cycle is finished: Need to change "build" in both the experiment dir and the temperate!! Update the submit_next_batch file. Need to remove the executable to avoid confusion: clm4.5_Kn11_Ndiags clm4.5_Kn11_Ndiags_20190419 5. Fire experiment: cloud-to-ground lightning fraction. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & Restored the change made on 20190724. L664-665: ! ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) ! our forc_lnfm is counts/km2/day, so divided by 24 to get per hr, fzeng, 7 May 2019 ig = (lh+forc_lnfm(g)/24/(4.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_LNFMfrac > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_LNFMfrac/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_LNFMfrac/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_LNFMfrac_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-07-30 16:18 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire1_DE720/run/ldas_batchrun.j . > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_LNFMfrac_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970101_180730z > ./ldas_batchrun.j ========== 20190731: 1. The clm4.5_LNFMfrac_DE720 crashed at date_time_new 19970102_193000z: forrtl: error (72): floating overflow Image PC Routine Line Source libirc.so 00002AAAAB590961 Unknown Unknown Unknown libirc.so 00002AAAAB58F0B7 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E3362 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2E31B6 Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2D1A0C Unknown Unknown Unknown libmpi_usempif08. 00002AAAAC2B3392 Unknown Unknown Unknown libc.so.6 00002AAAABA16910 Unknown Unknown Unknown LDASsaCN_mpi.x 00000000004D763B catch_types_mp_ad 1760 catch_types.F90 LDASsaCN_mpi.x 00000000005043F0 clsm_ensdrv_out_r 725 clsm_ensdrv_out_routines.F90 LDASsaCN_mpi.x 00000000004B0129 MAIN__ 2414 cnlsm_ensdrv_main.F90 L1760 catch_types.F90: add_cn_diagn%lai = cn_diagn_1%lai + cn_diagn_2%lai ... add_cn_diagn%fuelc = cn_diagn_1%fuelc + cn_diagn_2%fuelc In the code: fuelc is initialized to spval and spval = 1.e36_r8. In CNFireMod.F90: fuelc(c) = totlitc(c)+totvegc_col(c)-rootc_col(c)-fuelc_crop(c)*cropf_col(c) This is only done for non-crop areas outside tropical closed forests. So for crop and tropical closet forest areas, fuelc is still 1.e36_r8. This is likely what caused the floating overflow issue. Solution: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp In CNFireMod.F90: Added "fuelc = 0." to L639. > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_LNFMfrac/. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_LNFMfrac_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-07-31 11:15 ../build/Linux/bin/LDASsaCN_mpi.x* > ./lenkf.0.j It's running. Stopped it at date_time_new 19970104_070730z > ./ldas_batchrun.j 2. Also correct this for the executable of clm4.5_DE720. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp In CNFireMod.F90.restore: Added "fuelc = 0." to L639. > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux/ exec/clm4.5_Kn11/. > ls -l exec/clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-07-31 11:27 exec/clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x* 3. It turned out that the bug fix I made to CNFireMod.F90 is not needed. The original code is correct, so L661: lh = 0.0035_r8*6.8_r8*hdmlf**(0.43_r8)/30._r8/24._r8 Re-compiled the code for both clm4.5_LNFMfrac_DE720 and clm4.5_DE720. Discard the clm4.5_DE720 cycle that is just completed: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm -rf ../clm4.5_DE720_33706561 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-07-31 14:36 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart/ > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_46/output > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-07-31 14:36 ../build/Linux/bin/LDASsaCN_mpi.x* > ./lenkf.0.j It's running. Stopped it at date_time_new 19800105_040730z. > ./ldas_batchrun.j 4. Compared Li et al., 2012, 2013 and 2014 papers with the model code. ========== 20190801: 1. Processed and analyzed the available clm4.5_LNFMfrac_DE720 output. Changed "nrec" from 102 to 103 in grid_restore_hdeg.f90 to include "fuelc" and recompiled. grid_restore_hdeg.f90 works for 1980-2016. Made another copy of grid_restore_hdeg.f90 and renamed it to grid_restore_hdeg_part.f90. The latter works for 1997-2016 for the fire experiments. Compiled grid_restore_hdeg_part.f90. ~/Catchment/CLM4.5 > grid_restore_hdeg_part clm4.5_LNFMfrac_DE720 ~/Catchment/CLM4.5 > grads -lc "run check_fire_bug_fix.gs" 2. Continued to compare Li et al., 2012, 2013 and 2014 papers with the model code. 3. In L663 of CNFireMod.F90: ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) It seems to me that there is a bug in cos(3._r8*grc%latdeg(g)). In Fortran, the "cos" function needs the input to be in radian, but here grc%latdeg(g) is in degree. Talked to Randy. Randy told me to ask Matt if this is the case in the Fortran library we are using. Matt just left. Sent him an email. Matt's reply on 20190802: ~~~~~~~~~~~~~ Per the standard, the input of COS(): (Input) Must be of type real or complex. It must be in radians and is treated as modulo 2*pi. So, radians it is. If it's getting degrees, that becomes fun! I break my brain every time I try to figure out the floating point modulus let alone that of a transcendental real! ~~~~~~~~~~~~~ 4. Fire experiment: fix the cos bug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & L663-664: ! ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) ! our forc_lnfm is counts/km2/day, so divided by 24 to get per hr, fzeng, 7 May 2019 ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)*(SHR_CONST_PI/180._r8)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_cos > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_cos/. Set up the experiment: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run > nedit D0.5_cn.exe D0.5.bat & > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11_Ndiags_fire_cos/Linux/bin > ./ldsetup setup /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5_cn.exe /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/run/D0.5.bat --runmodel --monthsperjob 96 --landmodel catchCN > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_cos_DE720/run > nedit lenkf.0.j & -restart_path ../input/restart/output \ -restart_id clm4.5_fire_DE720 \ > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-08-01 16:22 ../build/Linux/bin/LDASsaCN_mpi.x* > cd ../input/restart/ > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/output > rm rst_clm4.5 > cd ../../run/ > cp ../../clm4.5_fire1_DE720/run/ldas_batchrun.j . > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_cos_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19970103_120730z > ./ldas_batchrun.j 5. Checked the entire Catchment-CN to see if the "cos" is used correctly everywhere else. Yes. ========== 20190802: 1. Processed the available output of clm4.5_cos_DE720 and analyzed it. ~/Catchment/CLM4.5 > grid_restore_hdeg_part clm4.5_cos_DE720 ~/Catchment/CLM4.5 > grads -lc "run check_fire_bug_fix.gs" 2. Fixed the bug in the fire model code. Commented "./ldas_batchrun.j" in ~/Catchment/CLM4.5/submit_next_batch_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & ! ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) ! There is a bug here. The input of "cos" must be in radians!! fzeng, 2 Aug 2019 ig = (lh+forc_lnfm(g)/24/(4.16_r8+2.16_r8*cos(3._r8*grc%latdeg(g)*(SHR_CONST_PI/180._r8)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) ! our forc_lnfm is counts/km2/day, so divided by 24 to get per hr, fzeng, 7 May 2019 > cp -p CNFireMod.F90 CNFireMod.F90.restore > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_new > /bin/cp -pr Linux/ exec/clm4.5_Kn11_new/. > ls -l exec/clm4.5_Kn11_new/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-08-02 14:54 exec/clm4.5_Kn11_new/Linux/bin/LDASsaCN_mpi.x* The current cycle of clm4.5_DE720 is completed in the evening. Did these below: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec > rm -rf clm4.5_Kn11 > mv clm4.5_Kn11_new clm4.5_Kn11 > ls -l clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-08-02 14:54 clm4.5_Kn11/Linux/bin/LDASsaCN_mpi.x* > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-08-02 14:54 build/Linux/bin/LDASsaCN_mpi.x* > cd run > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 86 2019-08-02 19:11 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_33752945/output/ > ./ldas_batchrun.j Uncomment "./ldas_batchrun.j" in ~/Catchment/CLM4.5/submit_next_batch_DE720 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_template > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-08-02 14:54 build/Linux/bin/LDASsaCN_mpi.x* ========== 20190805: 1. Created a Python script ~/python/clm4-to-clm4.5/plot_burn_bias.py to compute burned area biases with respect to GFED4. GFED4 data in hdf5 files. It took a loooong time to read these files in Python. Tried to search for solutions but had no luck. ========== 20190806: 1. Since I used Matlab to read hdf files before, I created a Matlab script ~/matlab/clm4-to-clm4.5/read_GFED4_fields.m to read the GFED4 hdf5 files. It's also very slow. But I used it to extract the burned area and fire C emission data and saved the data in .mat files which can be easily read by Python. 2. Computed and plotted the burned area biases using ~/python/clm4-to-clm4.5/plot_burn_bias.py. ========== 20190807: 1. Investigated why different locations (e.g. southern DRC vs. Mozambique) in S Africa with the same dominant vegetation type but have very different burned area. 2. Read the code to see if our btran is computed in the same way as that in CLM4.5. It looks the same to me. ========== 20190808: 1. Plotted fuelc for the 3 grassland sites: one in N America, one in S Africa, and one in central Asia. Found strange behavior at all 3 sites around year 2010. Why? 2. Read a few papers about lightning. 3. During the meeting with Randy, Melanie and Eunjee, we found from the plots Melanie made that surface soil moisture is lagging behind the water flux (rainf + snmelt - evap) by 1-2 months. It could be because the Catchment-CN4.5 code I am using is old and surface soil moisture is strongly coupled to root zone soil moisture as Randy suggested. Sarith helped me modify the Catchment code to include the new land physics updates. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5 > cp -pr LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3 LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 Files modified: GEOSsurface_GridComp/Shared/StieglitzSnow.F90 GEOSsurface_GridComp/GEOSland_GridComp/Shared/lsm_routines.F90 GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp/catchmentCN.F90 These updates will decouple surface soil moisture and root zone soil moisture. ========== 20190809: 1. Updated CNFireMod.F90: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CNFireMod.F90 & (1) In the lightning fraction calculation: changed 4.16 back to 5.16, and limit grc%latdeg to be within 60S and 60N as done in CLM5 (https://github.com/ESCOMP/ctsm/blob/release-clm5.0/src/biogeochem/CNFireLi2016Mod.F90): ig = (lh+forc_lnfm(g)/24/(5.16_r8+2.16_r8*cos(3._r8*min(60.,abs(grc%latdeg(g)))*(SHR_CONST_PI/180._r8)))*0.25_r8)*(1._r8-fs)*(1._r8-cropf_col(c)) (2) Changed the btran lower threshold from 0.3 to 0.4, and the btran upper threshold from 0.7 to 0.8 ! spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.3_r8)/ & ! (0.7_r8-0.3_r8))))*(1.0-max(0._r8, & ! min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) spread_m = (1.0_r8 - max(0._r8,min(1._r8,(btran_col(c)/wtlf(c)-0.4_r8)/ & ! To reduce burned area biases, changed the btran lower threshold from 0.3 to 0.4, (0.8_r8-0.4_r8))))*(1.0-max(0._r8, & ! and the btran upper threshold from 0.7 to 0.8, fzeng, 9 Aug 2019 min(1._r8,(forc_rh(g)-30._r8)/(80._r8-30._r8)))) > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > source $ESMADIR/src/g5_modules > gmake install > cd ../exec > mv clm4.5_Kn11 clm4.5_Kn11_20190802 > cd .. > mkdir -p exec/clm4.5_Kn11 > /bin/cp -pr Linux exec/clm4.5_Kn11/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec/clm4.5_Kn11 build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71242882 2019-08-09 12:48 build/Linux/bin/LDASsaCN_mpi.x* > ls -l input/restart/ lrwxrwxrwx 1 fzeng g0620 80 2019-08-08 10:22 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_50/output/ > cd run > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800102_020730z. > ./ldas_batchrun.j ========== 20190812: 1. The clm4.5_DE720_51 is finished, but the next cycle was not submitted. It's because the "./ldas_batchrun.j" was commented out in the submit_next_batch_DE720 file. Also, the new clm4.5_DE720 is still pointing to the old executable because I haven't updated that "build" in the template. Corrected this. Manually submitted the jobs. Uncommented "./ldas_batchrun.j" in submit_next_batch_DE720. Updated "build" in the template. ========== 20190813: 1. Investigated the strange behavior of fuelc in 2010 in cycle 48. ========== 20190814: 1. Talked to Weiyuan about the strange behavior of fuelc in 2010 in cycle 48. He said there could be some numerical instability in the code. 2. Added the diagnostics below to the output of Catchment-CN4.5. totlitc cwdc rootc CN_DriverMod.F90: real, pointer :: totlitc(:) ! (gC/m2) total litter carbon totlitc => ccs%totlitc frootc => pcs%frootc frootc_storage => pcs%frootc_storage frootc_xfer => pcs%frootc_xfer livecrootc => pcs%livecrootc livecrootc_storage => pcs%livecrootc_storage livecrootc_xfer => pcs%livecrootc_xfer deadcrootc => pcs%deadcrootc deadcrootc_storage => pcs%deadcrootc_storage deadcrootc_xfer => pcs%deadcrootc_xfer CNiniTimeVar.F90: cwdc(:) ! (gC/m2) coarse woody debris C cwdc => ccs%cwdc Modified these files to add the diagnostics: CN_DriverMod.F90 [follow 'fuelc'] process_cn.F90 [follow 'fuelc'] clsm_ensdrv_out_routines.F90 [follow 'fuelc'] catch_types.F90 [update 'N_cn_diagn' from 54 to 57 and follow 'fuelc'] clsm_ensdrv_glob_param.F90 [change 'N_CatchCN_OutFields = 59' to 'N_CatchCN_OutFields = 62'] Compiled: Used kn = 0.11 in compute_rc.F90. After adding a new diagnostic, the model should be rebuilt from a clean state: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > source $ESMADIR/src/g5_modules > gmake realclean > gmake install |& tee make.install.log > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > mkdir -p exec/clm4.5_Kn11_new > /bin/cp -pr Linux/ exec/clm4.5_Kn11_new/. > ls -l exec/clm4.5_Kn11_new/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71255170 2019-08-14 12:21 exec/clm4.5_Kn11_new/Linux/bin/LDASsaCN_mpi.x* After the current cycle was finished: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec > mv clm4.5_Kn11 clm4.5_Kn11_20190809 > mv clm4.5_Kn11_new clm4.5_Kn11 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71255170 2019-08-14 12:21 ../build/Linux/bin/LDASsaCN_mpi.x* > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 86 2019-08-14 14:09 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_33997998/output/ > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800103_010730z. > ./ldas_batchrun.j > nedit /home/fzeng/Catchment/CLM4.5/submit_next_batch_DE720 & Uncomment "./ldas_batchrun.j" Updated and compiled ~/Catchment/CLM4.5/grid_restore_hdeg.f90 Checked and made sure /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_template is using the right executable. Updated the corresponding restore files. 3. Analyzed and compared the output of cycles 50 and 51 using ~/Catchment/CLM4.5/check_fire_bug_fix.gs and ~/python/clm4-to-clm4.5/plot_burn_bias.py ========== 20190815: 1. Continued investigating the strange behavior of fuelc in 2010 in cycle 48. 2. Compute btran from sfmc and use it in the fire model. Modified: process_cn.F90 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSlana_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > source $ESMADIR/src/g5_modules > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_btran_sf > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_btran_sf/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec/clm4.5_Kn11_Ndiags_fire_btran_sf build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71259266 2019-08-15 16:27 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_52/output Will compare the output from this run to that of clm4.5_DE720_53. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800103_120730z > ./ldas_batchrun.j Restored process_cn.F90. ========== 20190816: 1. Checked the output of the clm4.5_fire_DE720 run set up yesterday. ========== 20190819: 1. Processed and analyzed the output of clm4.5_fire_DE720. ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_fire_DE720 ~/Catchment/CLM4.5/check_fire_bug_fix.gs ~/python/clm4-to-clm4.5/plot_burn_bias.py ========== 20190823: 1. Fire experiment: use sfmc to compute btran for grass, shrub and type 6; use rzmc to compute btran for trees expect type 6 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSlana_GridComp > nedit process_cn.F90 & > cp -p process_cn.F90 process_cn.F90.btran_rzNsf > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CN_DriverMod.F90 & > cp -p CN_DriverMod.F90 CN_DriverMod.F90.btran_rzNsf > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > mkdir -p exec/clm4.5_Kn11_Ndiags_fire_btran_rzNsf > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire_btran_rzNsf/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5 > /bin/cp -pr clm4.5_fire_DE720 clm4.5_fire_DE720_20190815 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec/clm4.5_Kn11_Ndiags_fire_btran_rzNsf build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71263362 2019-08-23 13:16 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart > ls -l lrwxrwxrwx 1 fzeng g0620 80 2019-08-15 16:38 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_52/output/ Will compare the output from this run to that of clm4.5_DE720_53. > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800103_000730z > ./ldas_batchrun.j ========== 20190824: 1. Checked the output of the fire experiment set up on 20190823. ========== 20190826: 1. Analyzed the output of the fire experiment set up on 20190823. ~/Catchment/CLM4.5 > grid_restore_hdeg clm4.5_fire_DE720 ~/Catchment/CLM4.5 > grads -lc "run check_fire_bug_fix.gs" ========== 20190903: 1. We will make the modification to btran calculation as done on 20190823 permanent, so spin up the model. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec > mv clm4.5_Kn11 clm4.5_Kn11_20190814 > mv clm4.5_Kn11_Ndiags_fire_btran_rzNsf clm4.5_Kn11 > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720 > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71263362 2019-08-23 13:16 build/Linux/bin/LDASsaCN_mpi.x* > cd run > ls -l ../input/restart/ lrwxrwxrwx 1 fzeng g0620 80 2019-08-28 10:55 output -> /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_57/output/ > ./ldas_batchrun.j > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSlana_GridComp > cp -p process_cn.F90 process_cn.F90.restore > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > cp -p CN_DriverMod.F90 CN_DriverMod.F90.restore ========== 20190904: 1. The clm4.5_DE720 run crashed. forrtl: severe (24): end-of-file during read, unit 10, file /discover/nobackup/fzeng/bcs/Icarus-NL/Icarus-NL_Reynolds//DE_00720x00360_PE_0720x0360/DE_00720x00360_PE_0720x0360.til Image PC Routine Line Source libmpi_usempif08. 00002AAAAC2D0F47 Unknown Unknown Unknown LDASsaCN_mpi.x 0000000003893F10 Unknown Unknown Unknown LDASsaCN_mpi.x 00000000006D3557 tile_coord_routin 460 tile_coord.F90 LDASsaCN_mpi.x 00000000006D181B tile_coord_routin 845 tile_coord.F90 LDASsaCN_mpi.x 00000000004D4422 MAIN__ 830 cnlsm_ensdrv_main.F90 This is because Weiyuan and I changed this tile file for use in the GEOSldas4.5_DE720 run. Solution: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > nedit driver_inputs_0.5D_GLOBAL.nml & L416-417: !tile_coord_file = 'DE_00720x00360_PE_0720x0360.til' tile_coord_file = 'DE_00720x00360_PE_0720x0360_2grid.til' > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSlana_GridComp > nedit tile_coord.F90 & L263-264: !case('DE_00720x00360_PE_0720x0360.til') ! 0.5-degree grid, grid = tile case('DE_00720x00360_PE_0720x0360_2grid.til') ! 0.5-degree grid, grid = tile L456-457: !if (ease_grid .or. (index(tile_coord_file,'DE_00720x00360_PE_0720x0360')/=0)) then if (ease_grid .or. (index(tile_coord_file,'DE_00720x00360_PE_0720x0360_2grid')/=0)) then Compiled: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec > mv clm4.5_Kn11 clm4.5_Kn11_20190823 > mkdir clm4.5_Kn11 > cd .. > /bin/cp -pr Linux exec/clm4.5_Kn11/. Test: > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720/run > ls -l ../build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71263362 2019-09-04 10:55 ../build/Linux/bin/LDASsaCN_mpi.x* > ./lenkf.0.j It's running. Stopped it at date_time_new 19800102_180730z > ./ldas_batchrun.j Updated the template: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_template > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71263362 2019-09-04 10:55 build/Linux/bin/LDASsaCN_mpi.x* > cd run > nedit driver_inputs_0.5D_GLOBAL.nml & L416-417: !tile_coord_file = 'DE_00720x00360_PE_0720x0360.til' tile_coord_file = 'DE_00720x00360_PE_0720x0360_2grid.til' 2. Melanie showed me the analyses she did and the negative biases in carbon emissions in the type 6 area. What we are going to do is we compute combustion completeness and fire related mortality of type 6 assuming that type 6 is 35% tree and 65% grass, as what we did for maximum fire spread rate. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > nedit CN_DriverMod.F90 & L1080-1100: ! data cc_dstex / 0., 0.22, 0.25, 0.25, 0.22, 0.22, 0.22, 0.22, 0.22, 0.3, 0.3, 0.3, 0.3, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data cc_leax / 0., 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data cc_lstex / 0., 0.22, 0.25, 0.25, 0.22, 0.22, 0.22, 0.22, 0.22, 0.3, 0.3, 0.3, 0.3, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data cc_othex / 0., 0.45, 0.5, 0.5, 0.45, 0.45, 0.45, 0.45, 0.45, 0.55, 0.55, 0.55, 0.55, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data fm_leax / 0., 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data fm_lstex / 0., 0.45, 0.5, 0.5, 0.45, 0.45, 0.35, 0.35, 0.45, 0.55, 0.55, 0.55, 0.55, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data fm_othex / 0., 0.45, 0.5, 0.5, 0.45, 0.45, 0.35, 0.35, 0.45, 0.55, 0.55, 0.55, 0.55, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! original CLM4.5 values ! data fm_roox / 0., 0.13, 0.15, 0.15, 0.13, 0.13, 0.1, 0.1, 0.13, 0.17, 0.17, 0.17, 0.17, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2/ ! original CLM4.5 values ! data fm_lroox / 0., 0.13, 0.15, 0.15, 0.13, 0.13, 0.1, 0.1, 0.13, 0.17, 0.17, 0.17, 0.17, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2/ ! original CLM4.5 values ! data fm_droox / 0., 0.13, 0.15, 0.15, 0.13, 0.13, 0.1, 0.1, 0.13, 0.17, 0.17, 0.17, 0.17, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2/ ! original CLM4.5 values data cc_dstex / 0., 0.22, 0.25, 0.25, 0.22, 0.22, 0.60, 0.22, 0.22, 0.3, 0.3, 0.3, 0.3, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data cc_leax / 0., 0.8, 0.8, 0.8, 0.8, 0.8, 0.80, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data cc_lstex / 0., 0.22, 0.25, 0.25, 0.22, 0.22, 0.60, 0.22, 0.22, 0.3, 0.3, 0.3, 0.3, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data cc_othex / 0., 0.45, 0.5, 0.5, 0.45, 0.45, 0.68, 0.45, 0.45, 0.55, 0.55, 0.55, 0.55, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data fm_leax / 0., 0.8, 0.8, 0.8, 0.8, 0.8, 0.80, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data fm_lstex / 0., 0.45, 0.5, 0.5, 0.45, 0.45, 0.64, 0.35, 0.45, 0.55, 0.55, 0.55, 0.55, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data fm_othex / 0., 0.45, 0.5, 0.5, 0.45, 0.45, 0.64, 0.35, 0.45, 0.55, 0.55, 0.55, 0.55, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8, 0.8/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data fm_roox / 0., 0.13, 0.15, 0.15, 0.13, 0.13, 0.17, 0.1, 0.13, 0.17, 0.17, 0.17, 0.17, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data fm_lroox / 0., 0.13, 0.15, 0.15, 0.13, 0.13, 0.17, 0.1, 0.13, 0.17, 0.17, 0.17, 0.17, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 data fm_droox / 0., 0.13, 0.15, 0.15, 0.13, 0.13, 0.17, 0.1, 0.13, 0.17, 0.17, 0.17, 0.17, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2/ ! assume type 6 is 35% tree and 65% grass, fzeng, 2019 Compiled: > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/src/Components/GEOSsurface_GridComp/GEOSland_GridComp/GEOScatchCN_GridComp > setenv ESMADIR /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05 > source $ESMADIR/src/g5_modules > gmake install > cd ../../../GEOSlana_GridComp/ > gmake install > cd ../../Applications/LDAS_App/ > gmake install > cd ../../.. > /bin/cp -pr Linux exec/clm4.5_Kn11_Ndiags_fire/. > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720 > rm build > ln -s /gpfsm/dnb31/fzeng/clm4-to-clm4.5/LDAS/clm4.5/LDASsa_m3-16_0_p2_CatchCatchCN_for_MERRA3_V24_C05/exec/clm4.5_Kn11_Ndiags_fire build > ls -l build/Linux/bin/LDASsaCN_mpi.x -rwxr-xr-x 1 fzeng g0620 71263362 2019-09-04 15:54 build/Linux/bin/LDASsaCN_mpi.x* > cd input/restart > rm output > ln -s /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_DE720_57/output > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > nedit driver_inputs_0.5D_GLOBAL.nml & L416-417: !tile_coord_file = 'DE_00720x00360_PE_0720x0360.til' tile_coord_file = 'DE_00720x00360_PE_0720x0360_2grid.til' Test: > interactive.py -A sp3 -n 308 -a g0620 -X --debug > cd /discover/nobackup/fzeng/clm4-to-clm4.5/LDAS/tests/clm4.5/clm4.5_fire_DE720/run > ./lenkf.0.j It's running. Stopped it at date_time_new 19800102_020730z > ./ldas_batchrun.j