fv history error

[boyle5@llnl_gov]

I am running the CAM3.1p2, fv2x2.5 on Thunder at LLNL.
This is the standard distribution on a verified platform.

NHTFRQ(1) determines the frequency of the default history file.
If I set NHTFRQ(1) to anything but 0 (the default), I get an error involving memory allocation as the model tries to write out the history file.
If I use 0 all goes OK.
I have run the model in the T42 configuration, on the same platform and this error does not occur.

This appears to be a bug - Possibly it might be a symptom of a problem in the accumulations for the history files, and thus could affect the monthly mean values ( NHTFRQ(1) = 0) that are produced.

 

[rneale]

Coule you send the namelist you use that leads to this error.
Thanks

 

[boyle5@llnl_gov]

From some tests it appears that the problems arises when I try to use the output option that restricts the output to specific regions rather than global i.e. I have the line:
fincl1lonlat = '151e:161e_7s:2n'

it appears that
fincl2lonlat = '151e:161e_7s:2n'
is OK it is just the default history that has this difficulty.

A simple workaround is to avoid use of the lonlat for the default history. Although this problem might point to something more fundamental.

--Jim

 

[eaton]

Jim,

I can't reproduce this problem. fincl1lonlat = '151e:161e_7s:2n' works for me with cam3.1.p2. In recent communications with Art Mirin I've learned that the current cam development code is not passing the port tests on thunder due to the upgrade to version 9.1 intel compiler. It's possible that this is also causing the problem you're seeing with cam3.1.p2. Art says that the compiler version 8.1 is still giving correct answers on thunder, so you might try that as well.

Brian

 

[boyle5@llnl_gov]

Brian,
Thanks for the quick reply - Just to try and clarify a bit.
I now built the model on a NERSC machine - Bassi. This is a IBM p575 Power 5, similar to blue vista, blue ice.

I built using the vanilla config with the -dyn fv -res 2x2.5

below is the namelist I have tested. If I run with this namelist the model hangs - on batch jobs the time runs out.
if I delete the line: fincl1lonlat = '149e:161e_5s:2n',
all goes well in that the model runs in a reasonable time.

--Jim


&CAMEXP
BNDTVS = '/scratch/scratchdirs/u14201/modelData/atm/cam2/sst/sst_HadOIBl_bc_2x2.5_1949_2001_c020531.nc'
bnd_topo = '/scratch/scratchdirs/u14201/modelData/atm/cam2/topo/topo-from-cami_0000-09-01_2x2.5_L26_c040615.nc'
absems_data = '/scratch/scratchdirs/u14201/modelData/atm/cam2/rad/abs_ems_factors_fastvx.c030508.nc'
aeroptics = '/scratch/scratchdirs/u14201/modelData/atm/cam2/rad/AerosolOptics_c040105.nc'
bndtvaer = '/scratch/scratchdirs/u14201/modelData/atm/cam2/rad/AerosolMass_V_2x2.5_clim_c031022.nc'
bndtvdms = '/scratch/scratchdirs/u14201/modelData/atm/cam2/scyc/DMS_emissions_64x128_c030722.nc'
bndtvghg = '/scratch/scratchdirs/u14201/modelData/atm/cam2/ggas/ghg_1870_2100_c040122.nc'
bndtvo = '/scratch/scratchdirs/u14201/modelData/atm/cam2/ozone/pcmdio3.r8.64x1_L60_clim_c970515.nc'
bndtvoxid = '/scratch/scratchdirs/u14201/modelData/atm/cam2/scyc/oxid_3d_64x128_L26_c030722.nc'
bndtvscon = '/scratch/scratchdirs/u14201/modelData/atm/cam2/rad/scon_1870_2100_c040122.nc'
bndtvsox = '/scratch/scratchdirs/u14201/modelData/atm/cam2/scyc/SOx_emissions_64x128_L2_1870-1871_c040520.nc'
bndtvvolc = '/scratch/scratchdirs/u14201/modelData/atm/cam2/rad/VolcanicMass_1870-1999_64x1_L18_c040115.nc'
isccpdata = '/scratch/scratchdirs/u14201/modelData/atm/cam2/rad/isccp.tautab_invtau.nc'
CTITLE = 'case FA_1992-12-15-00000',
NCDATA = '/scratch/scratchdirs/u14201/runDir_fv2x25L26_TC/FA_CAM31_TC_fv2x2.5L26_initialData/FA_1992-12-14-64800.cam2.i.1992-12-15-00000.nc'
CASEID = 'FA_1992-12-15-00000',
MSS_IRT = 0,
NSREST = 0,
NREFRQ = 0,
INITHIST = 'NONE',
NDENS = 2,2,2
START_YMD = 19921215,
START_TOD = 00000,
REF_YMD = 19920601,
REF_TOD = 0,
CALENDAR = 'NO_LEAP'
MFILT = 30,481,241,81,
DTIME = 1800,
NELAPSE = -10,
NHTFRQ = 1,1,-1,-3,
FINCL2 = 'CLDICE:I','CLDLIQ:I','CME','CMEICE','CMELIQ','CMFDICE','CMFDLIQ','CMFDQ',
'CMFDQR','CMFDT','CMFMC','CONCLD','DCCLDICE','DCCLDLIQ','DCQ','DFCLDICE',
'DFCLDLIQ','DFQ','DISED','DLSED','DQSED','DTCOND','DTH','DTV',
'EVAPPREC','EVAPQCM','EVAPQZM','EVAPSNOW','EVAPTCM','EVAPTZM','FICE',
'HACLDICE','HACLDLIQ','HAQ','HCME','HEVAP','HFREEZ','HMELT','HPROGCLD',
'HREPART','HSED','ICE2PR','ICEFRAC','KTOOP','LANDFRAC','LHFLX','LHFLXOI',
'LIQ2PR','OCNFRAC','OMEGA','PBLH','PHIS','PRECC','PRECL','PRECSC',
'PRECSH','PRECSL','PRECZ','PS:I','PSL','Q:I','QC','QFLX','RELHUM','REPARTICE',
'REPARTLIQ','SFCLDICE','SFCLDLIQ','SFQ','SHFLX','SHFLXOI','SNOWHICE',
'SNOWHLND','SRFRAD','T:I','TACLDICE','TACLDLIQ','TAQ','TAUX','TAUY',
'TMQ','TREFHT','TS:I','TSMN','TSMX','TTEND','U:I','UTEND','V:I','VACLDICE',
'VACLDLIQ','VAQ','VAT','VD01','VDCLDICE','VDCLDLIQ','VTEND','Z3',
'ZMDICE','ZMDLF','ZMDLIQ','ZMDQ','ZMDT',
fincl3 = 'CLDHGH','CLDLOW','CLDMED','CLDTOT','CLOUDY','CLOUD',
'FISCCP1',
'FLNSC','FLNSOI','FLNS','FLNTC','FLNT','FLUTC','FLUT',
'FSDSC','FSDS','FSNSC','FSNSOI','FSNS','FSNTC','FSNTOAC','FSNTOA','FSNT',
'GCLDLWP','ICLDIWP','ICLDLWP',
'LWCF','SWCF',
'MEANPTOP','MEANTAU','MEANTTOP','TCLDAREA',
'QRL','QRS',
'SOLIN',
'TGCLDIWP','TGCLDLWP','ICIMR','ICWMR','PS'
fincl4 = 'PRECC','PRECL','PSL','TREFHT','Z500:I','CLDTOT','FLUT','PS','U200','U850','V200','V850',
'T300','T850','Q200','Q850','TAUX','TAUY','SHFLX','LHFLX','FSNT',
'ZBOT','TBOT','UBOT','VBOT','QBOT','PBOT',
fincl1lonlat = '149e:161e_5s:2n',
fincl2lonlat = '149e:161e_5s:2n',
fincl3lonlat = '149e:161e_5s:2n',
doisccp = .T.
IRADSW = -1,
IRADLW = -1,
IRADAE = -12,
SSTCYC = .F.,
OZNCYC = .T.,
IYEAR_AD = 1950,
SCON = 1.367e6,
KMXHDC = 5,
/
&clmexp
fpftcon = '/scratch/scratchdirs/u14201/modelData/lnd/clm2/pftdata/pft-physiology',
fsurdat = '/scratch/scratchdirs/u14201/modelData/lnd/clm2/srfdata/cam/clms_2x2.5_c031104.nc'
finidat = '/scratch/scratchdirs/u14201/runDir_fv2x25L26_TC/FA_CAM31_TC_fv2x2.5L26_initialData/FA_1992-12-14-64800.clm2.i.1992-12-15-00000.nc',
hist_fexcl1 = 'H2OSOI','SOILLIQ','SOILICE','TSOI','TLAKE',
/