[RegCNET] question when running regcm4 with CLM scheme

lizk lizk at climate.sh.cn
Fri Jan 28 06:36:20 CET 2011


Dear all,

         Is there anyone who run Regcm4 with CLM successfully? I have compiled  the model and preprocessed  the input data with no problem, but some errors occured when I run it in parallel. A part of  log file is as follows,

 Attempting to initialize the land model .....

 Attempting to initialize run control settings .....
 nsrest =             0  dtime =           600
 start_ymd =      20010201  start_tod =             0
 stop_ymd =      20011031  stop_tod =             0
 calendar = 
 NO_LEAP                                                                                                                         
 wrtdia =   T
 irad =             3
 hist_nhtfrq =           -12          -12          -12          -12 
          -12          -12
 error: logname not defined
 ENDRUN: called without a message string
p0_27458:  p4_error: : 1
rm_l_7_30687: (0.664062) net_send: could not write to fd=5, errno = 32
rm_l_4_27522: (1.191406) net_send: could not write to fd=5, errno = 32
rm_l_2_27498: (1.515625) net_send: could not write to fd=7, errno = 32
rm_l_5_30668: (0.871094) net_send: could not write to fd=5, errno = 32
 .....
 ........................................................................................................................
Here is my regcm.in:
&dimparam
 iy     = 80,
 jx     = 112,
 kz     = 18,
 nsg    = 1,
 /
 &geoparam
 iproj = 'LAMCON',
 ds = 60.0,
 ptop = 5.0,
 clat = 33.5,
 clon = 100.0,
 plat = 33.5,
 plon = 100.0,
 truelatl = 30.0,
 truelath = 60.0,
 /
 &terrainparam
 domname = 'eastasia',
 itype_in = 1,
 ntypec = 10,
 ntypec_s = 10,
 ifanal = .true.,
 smthbdy = .false.,
 lakadj = .false.,
 fudge_lnd   = .false.,
 fudge_lnd_s = .false.,
 fudge_tex   = .false.,
 fudge_tex_s = .false.,
 ntex = 17,
 h2opct = 75.,
 dirter = './Input',
 inpter = './DATA',
 /
 &ioparam
 igrads = 1,
 ibigend = 1,
 ibyte = 4,
 iotyp = 1,       
 /
 &debugparam
 debug_level = 1
 /
 &boundaryparam
 nspgx  = 12,
 nspgd  = 12,
 /
 &modesparam
 nsplit = 2,
 /
 &lakemodparam
 lkpts  = 10,
 /
 &globdatparam
 ssttyp = 'OISST',
 dattyp = 'NNRP2',
 ehso4 = .false.,
 globidate1 = 2001020100,
 globidate2 = 2001103100,
 dirglob = './Input',
 inpglob = './DATA',
 /
 &lsmparam
 lsmtyp = 'BATS',
 /
 &aerosolparam
 aertyp = 'AER00D1'
 ntr = 10,
 nbin = 4,
 /
&restartparam
 ifrest  = .false. ,
 idate0  = 2001020100,
 idate1  = 2001020100,
 idate2  = 2001103100,
 nslice  =        120,
 /
 &timeparam
 radfrq  =    30.,
 abemh   =    18.,
 abatm   =   600.,
 dt      =   100.,
 ibdyfrq =          6,
 /
 &outparam
 ifsave  = .true. ,
   savfrq  =  7200.,
 iftape  = .true. ,
   tapfrq  =     6.,
 ifrad   = .true. ,
   radisp  =     6.,
 ifbat   = .true. ,
 ifsub   = .true. ,
   batfrq  =     3.,
 ifprt   = .false.,
   prtfrq  =    12.,
   kxout   =          6,
   jxsex   =         40,
 ifchem  = .true.,
   chemfrq =     6.,
 dirout = './output',
 /
 &physicsparam
 iboudy  =          5,
 ibltyp  =          1,
 icup    =          2,
   igcc    =          2,
 ipptls  =          1,
 iocnflx =          2,
 ipgf    =          1,
 iemiss  =          0,
 lakemod =          0,
 ichem   =          0,
 /
 &subexparam
 qck1land=   .250E-03,
 qck1oce =   .250E-03,
 cevap   =   .100E-02,
 caccr   =      3.000,
 /
 &grellparam
 /
 &emanparam
 /
 &chemparam
 idirect   =    1,
 inpchtrname  =   '',
 inpchtrsol   =   .00,
 inpchtrdpv   =  .00000, .00000,
 inpdustbsiz  =   .00,  .00,
 /
 &clmparam
 dirclm = './Input',
 clmfrq = 12.,
 imask = 1,
/
....................................................................................................
I have run CLM standalone before and I suspect that Regcm4 with CLM need a namelist file for CLM model specially in terms of past experiences. Also I find the "error: logname not defined
" is in controlMOD.F90 in ..\RegCM-4.0\Main\clm3.5\src\main\, and there is some lines like " namelist / clm_inparm/ &", no "clm_inparm" however can be found in regcm.in. Any advice will be appreciated. Thank you!

                                                                                                                 Li Zhenkun


More information about the RegCNET mailing list