User's Guide to NCAR CCM3.6 Search Page

2. Using CCM3.6

    2.4 Example Namelists

2.4.1 Example 1 -- Initial Run, One Day

 &LSMEXP
 finidat = 'arbitrary initialization'
 /

CASEID

The case identifier for this run is run01. It is recorded in the history file header record, the model printed output, and is used as part of the MSS path names for the history and restart files. CASEID must be 8 (or less) valid UNIX filename characters.

CTITLE

This title will appear in the model output and in the history file headers.

NCDATA

The, initial conditions, dataset is a netCDF file called SEP1.T42.0198. See "Model Input Datasets" .

BNDTVS

Specifies the time-variant sea surface temperature (SST) dataset.  This file must be in netCDF format.

BNDTVO

Specifies the time-variant ozone dataset ozn.0596.r8.nc. This dataset is defined on a pressure grid (unlike the hybrid grid of the model) and is in netCDF format.

NSWRPS

Mass Store, write password, for all output datasets from this run. This password must remain the same throughout a case.

IRT

Retention period of files written to the NCAR Mass Storage System. History files from this run will be written with a retention period of 100 days.

NSREST

Setting NSREST = 0 requests an initial run starting from the initial dataset, NCDATA. Continuation (restart and branch) runs may be specified via this variable. See Examples 3 and 5 for more information on continuation runs.

NDENS

Specifies the packing factor used for model history files.  Data packing is available only on an NCAR Cray and results in a loss of precision in the data values. In this example, the history files are packed 2 to 1.

NNBDAT, NNBSEC, NNDBAS, NNSBAS

These time variables allow the user to specify the base date and seconds as well as the base day (starting day number of run).  If these variables are not defined the model will use date  information from the initial dataset. In this example, NNBDAT sets the base date for this run to "00000901", or September 1, year 0000.

MFILT

Specifies the number of time samples to be written to each history file.  The model will open a history tape and write mfilt time samples before closing the file.  Setting MFILT to 20 means that each complete output history file will contain 20 time samples.

DTIME

This specifies the timestep for the run in seconds (floating point).  In general changes in resolution require a change in DTIME.

NESTEP

Specifies the ending timestep for the model run. This example will end at timestep 72.  Positive integer values denote timesteps while negative integer values denote model days.  The same ending time could have been specified as NESTEP=-1.

Hint: When starting a long run, it is a good idea to make a 1-day initial run, then restart the case for the remainder of the experiment.  If the system crashes and the job is rerun,  the model will restart from the last restart files.  If on the other hand, you specified your model run as one long, initial run,  the job would reload the initial namelist and start again from the beginning, duplicating all model results from the aborted job.

NHTFRQ

Specifies the number of timesteps (if positive) or the number of model hours (if negative) between history file writes. To write a file every day would require running 72 timesteps between model writes (72 *1200 seconds/step = 86400 seconds or 1 day).  Alternatively you could set NHTFRQ to -24 to get the same output frequency.  Timestep 0 data will always be written to the history file. MFILT set to 20 means that each history file will contain 20 time slices of data.

INCORBUF, INCORHST, INCORRAD

Specifies whether model buffers are stored in-core or out-of-core (written to disk). The target machine for this run is a Cray with SSD.    Availability of the high speed SSD allows CCM3.6 to run out-of-core with minimal overhead. When running on a machine with no external high-speed I/O device (SSD),  namelist variables INCORBUF, INCORHST, and INCORRAD should be .true. (Default).

IYEAR_AD

Calculate the earth's orbital parameters for 1950 AD.  See Example 4 or Table 2.5 for more information on how to set variable.  The year 1950 is used to most closely match the orbital parameters of previous releases of CCM3.

FINIDAT

LSM initial dataset. "arbitrary initialization" requests that the LSM use hard coded values for initial state information instead of reading the data from an input file.

 Search page

Questions on these pages can be sent to... erik@ucar.edu .