benchmark runs
A series of validation/test runs made with each release of the model. Namelist files and job scripts used to make these benchmark runs are available in subdirectories under the tests/ directory of the current release (e.g., namelist input files for the “control” test run are modeldir/tests/control/*.inp).
continuation run
A continuation run continues from the last output history of the previous run. That history (START time) must be on the first OUTPUT file provided by the namelist input file. A continuation run must not specify a SOURCE file or SOURCE_START time.
The directory containing start-up and other input data files required for running the model. A minimal set of datadir files are available via download. The datadir is sometimes referred to by the TGCMDATA environment variable. Additional data files are available via the NCAR Community Data Portal.
diagnostic fields
A list of diagnostic fields are available to be saved on secondary history files. See section Saving Diagnostic Fields.
Subdirectory under the modeldir containing documentation, e.g., the User’s Guide, Model Description, Release Notes, etc.
The model execution directory. This is the directory where the model is built and executed. It is typically, but not necessarilly, a subdirectory of your working directory workdir. When a job script is executed from a working directory, the execdir is created if it does not already exist. During a model run, output history files are written to the execdir.
A model history records the state of the model at a discrete instant in model time. One or more histories are stored in netCDF history files.
initial run
An initial run is started from a history on a SOURCE file (see also SOURCE_START). Subsequent continuation runs do not provide SOURCE or SOURCE_START, but rather search for the START time on the first OUTPUT history file provided in the namelist input, and continue the run from there.
job script
A csh script in the scripts/ directory which, when executed, will build and execute the model. The user defines a few shell variables in the job script, such as the modeldir, and the namelist input. For more details, please see job scripts.
model time
TIEGCM model time is represented by an integer triplet: day,hour,minute, where day is the julian day of the year, and hour is the ut. The variable for model time on history files is mtime(3,ntimes). For example, a history file may contain 24 hourly histories for day 80: mtime = 80,1,0, 80,2,0, ... 81,0,0.
The model root directory. This directory typically contains subdirectories src/ (model source code), scripts/ (utility scripts), doc/ (documentation), and tests/ (test runs). The modeldir is available via download, and is typically a subdirectory of the model working directory (workdir).
namelist input
The model reads user specified parameters from the namelist input file via f90 standard namelist read. Keyword/Value pairs are read from unit 5, and are validated by the input module (input.F).
NCAR Community Data Portal
The NCAR Community Data Portal is a public data repository with datasets from NCAR, UCAR, UOP, and participating organizations. To browse TIEGCM-related files (mostly netCDF history files for model start-up, or results of benchmark runs), click on the “Models” link, then to the “Thermospheric General Circulation Models” link, and finally to the desired model version. NetCDF Metadata is available without actually downloading files.
TIEGCM output history files are written in netCDF, a self-describing platform-independent data format written and maintained by the UCAR Unidata program.

The TIEGCM can be run in one of two resolutions:

  • 5 x 5 deg lat x lon, 2 grid levels per scale height (dz = 0.50)
  • 2.5 x 2.5 deg lat x lon, 4 grid levels per scale height (dz = 0.25)

The resolution is set by the “modelres” shell variable in the TIEGCM job script. See also the section on Grid Structure and Resolution.


The 2.5-degree resolution model is available in version 1.94, but it is not fully validated or supported by the public release.

Subdirectory under the modeldir containing supporting and utility scripts, including job scripts, the default namelist input file, several Make files, etc.
Subdirectory under the modeldir containing the model source code (*.F, *.h files).
Subdirectory under the modeldir. The tests directory contains subdirectories for benchmark runs that were made for the current release. The subdirectories contain job scripts and namelist input files that can be used to reproduce benchmark runs for testing and validation purposes. For more information, see the section on Benchmark Test Runs.
A unix environment variable that refers to the datadir. This environment variable may be used when referring to data files in the namelist read file, e.g., “GPI_NCFILE = $TGCMDATA/”. See namelist read files.
Post-processor and visualizer for TIEGCM netCDF history files. Written in f90, and available at the TGCM download site. See tgcmproc_f90.
Post-processor and visualizer for TIEGCM netCDF history files. This processor is Written in IDL with a GUI, and is available at the TGCM download site. See tgcmproc_idl.
Post-processor and visualizer for TIEGCM netCDF history files. This processor reads time-series history files and makes ut vs pressure and ut vs latitude contours. It is written in IDL with a GUI, and is available at the TGCM download site. See utproc.
Your local working directory. This will typically contain the model root directory modeldir, the execution directory execdir, and related namelist input files, job scripts, stdout files, etc. It may also contain a data subdirectory datadir.

Previous topic

Contact Information

This Page