CLM run failed with dynamic landuse input

[leihm@tsinghua_edu_cn]

Dear Sir/Madam,

I’m trying to use dynamic landuse data for CLM 3.5 modeling. As a simplest test, I used one year landuse data to generate the surface data (using mksurfdata tool). The namelist for mksurfdata tool is:

&clmexp
mksrf_fgrid = '/home/lhm/clm/clm_model/tools/mkgriddata/griddata_0170x0135.nc'
mksrf_fsoitex = '/home/lhm/clm/inputdata/rawdata/mksrf_soitex.10level.nc'
mksrf_fvegtyp = '/home/lhm/clm/inputdata/rawdata/mksrf_pft.060929.nc'
mksrf_fsoicol = '/home/lhm/clm/inputdata/rawdata/mksrf_soilcol.060929.nc'
mksrf_flanwat = '/home/lhm/clm/inputdata/rawdata/mksrf_lanwat.060929.nc'
mksrf_fmax = '/home/lhm/clm/inputdata/rawdata/mksrf_fmax.070406.nc'
mksrf_fglacier = '/home/lhm/clm/inputdata/rawdata/mksrf_glacier.060929.nc'
mksrf_furban = '/home/lhm/clm/inputdata/rawdata/mksrf_urban.060929.nc'
mksrf_flai = '/home/lhm/clm/inputdata/rawdata/mksrf_lai.060929.nc'
mksrf_fdynuse = 'pftdyn_1982.txt'
mksrf_gridtype = 'regional'
/

--------------------
In the pftdyn_1982.txt file, the content is:
"/home/lhm/clm/inputdata/rawdata/mksrf_pft.060929.nc" 1993

The mksurfdata tool runs successfully, and generates two files:
surface-data.dynpft.0135x0170.nc
surfdata_0170x0135.nc

Then, I set the CLM namelist as:
&clm_inparm
caseid = '$case'
ctitle = '$case'
dtime = 3600
start_ymd = 19820101
start_tod = 0
stop_ymd = 19851231
stop_tod = 86400
irad = -1
wrtdia = .true.
mss_irt = 0
hist_dov2xy = .true.
hist_nhtfrq = 0
hist_mfilt = 1
hist_crtinic = 'MONTHLY'
finidat = ''
fsurdat = '$CSMDATA/surfdata/$fsurdat'
fpftdyn = '$CSMDATA/surfdata/$dynludat'
fatmgrid = '$CSMDATA/griddata/griddata_0170x0135.nc'
fatmlndfrc = '$CSMDATA/griddata/fracdata_0170x0135.nc'
fpftcon = '$CSMDATA/pftdata/pft-physiology.c070207'
offline_atmdir = '$CSMDATA/mydata'
frivinp_rtm = '$CSMDATA/rtmdata/rdirc.05.061026'
nsrest = 0
hist_empty_htapes = .true.
hist_fincl1 = 'TBOT','FSDS','FSA','QBOT','WIND','RAIN','FCTR','FCEV','FGEV','FSH','FGR','FIRA','FSR'
rtm_nsteps = 1
hist_avgflag_pertape = 'A'

/


However, CLM run failed, and shows:

GPTLstop: thread 0 timer for init had not been started.
clm_mapa2l subroutine
clm_mapa2l mapping complete
soil balance e nstep = 1 point = 22165 imbalance = -3.95 W/m2
clm model is stopping
ENDRUN: called without a message string
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 15[cli_15]: aborting job:
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 15
rank 15 in job 37 node1_54385 caused collective abort of all ranks
exit status of rank 15: return code 1

I seems that the energy balance over soil is not conservative. If I set fpftdyn in the namelist as fpftdyn = ‘’, the model runs successfully.
I don’t know why. Thank you for your help!

Best Wishes.

 

[slevis]

Dear CLM3.5 User,

The 3.5 version is rather old, and we do not support it any more. Also, transient land use was a very new and unreliable option in 3.5.

Our best recommendation is that you upgrade to the CLM4 model. This will involve a learning phase: reading the User's Guide, becoming familiar with the new scripts and the latest mksurfdata tool, etc.; however, the CLM4 includes various improvements and the transient land use capability is more robust.

Sam Levis