this section describes how to download the CCSM3.0 source code and input datasets and the resulting unpacked directory structures.
CCSM3.0 is available via the web from:
www.cesm.ucar.edu/models/CCSM3.0
The CCSM3.0 distribution consists of a single tar file, (ccsm3.0.tar.gz) that contains the source code and accompanying scripts. In addition there are several input data tar files that need to be selectively downloaded by the user depending on the component set desired as well as the resolution needed.
Source code and scripts ccsm3.0.tar.gz source code and scripts Input data ccsm3.0.inputdata.atm_lnd.tar CAM/CLM data common to all resolutions ccsm3.0.inputdata.T85.tar CAM/CLM T85 specific data ccsm3.0.inputdata.T42.tar CAM/CLM T42 specific data ccsm3.0.inputdata.T31.tar CAM/CLM T31 specific data ccsm3.0.inputdata.gx1v3.tar POP/CSIM gx1 specific data ccsm3.0.inputdata.gx3v5.tar POP/CSIM gx3 specific data ccsm3.0.inputdata.cpl.tar CPL data ccsm3.0.inputdata.dxxx.tar data models ccsm3.0.inputdata.latm.tar LATM data ccsm3.0.inputdata_user.tar empty directory tree for user-created input datasets ccsm3.0.inputdata.CLM_RAW_SFC_GEN_FILES.tar optional files for the generation of CLM3.0 surface datasets at resolutions not included with this release. This is currently not officially supported in the CCSM3.0 release.
To uncompress and untar the file ccsm3.0.tar.gz, use the Unix
gunzip and
tar commands (note that >
below indicates a Unix prompt):
> gunzip -c ccsm3.0.tar.gz | tar -xf -
Once untarred, the resulting directory tree is:
$CCSMROOT/ | ccsm3/ | +------------+---------------------+ | | scripts/ models/ | | [build/run scripts] [model code] | | | | +----------------+--------------+ | | | | | ccsm_utils/ create_newcase* create_test* | | | +----------+---------------+-----------+----------+----+---+------+--------+-... | | | | | | | | cpl/ atm/ ocn/ ice/ lnd/ bld/ utils/ dead/ | | | | | | | | +-----+------+ +--+--+ +--+----+ +------+ +----+... | | | | | | | | | | | | | | cpl6/ cam/ datm6/ latm6/ pop/ docn6/ csim4/ dice6/ clm2/ dlnd6/ mct/ esmf/ dead6/
Note that not all of the directory names (e.g. clm2/ and csim4) have kept up with the official released versions of these models (e.g. CLM3 and CSIM5).
The code distribution consists of two top-level directories:
$CCSMROOT/ | ccsm3/ | +---------+----------+ | | scripts/ models/
In the scripts/ directory contains the scripts and tools required to construct the supported CCSM configurations. The models/ directory contains all the source code needed to build the CCSM. Underneath the models/ directory are directories for each of the CCSM components:
models/bld/ Gnumake Makefiles and machine-dependent macro files models/cpl Source code for the flux coupler models/atm/ Source code for the atmosphere models models/ice/ Source code for the sea-ice models models/ocn/ Source code for the ocean models models/lnd/ Source code for the land models models/utils/ Source code for common utilities (e.g. esmf, mct) models/csm_share/ Source code shared by multiple CCSM components (e.g. shr/ cpl/)
The various downloaded input data tar files should be untarred into a common inputdata directory. In addition, a CCSM resolution must first be selected before the inputdata is downloaded. For example, if a user wishes to run a T42_gx1v3 CCSM3 run with fully active components, they should untar the input data as follows (we assume the user will be untarring the data into /user/data):
> cd /user/data > tar -xf ccsm3.0.inputdata.T42_gx1v3.tar > tar -xf ccsm3.0.inputdata.gx1v3.tar > tar -xf ccsm3.0.inputdata.atm\_lnd.tar > tar -xf ccsm3.0.inputdata.cpl.tar
This will result in an inputdata/ directory tree as shown in section 3.3 but only containing sub-directories for cpl/, cam2/, pop/, csim4/ and clm2/.
The directory tree for the tarfile containing the CCSM3.0 input data is shown below.
$DIN_LOC_ROOT/ | inputdata/ | +----------+----------------+-------------+------------+ | | | | | cpl/ atm/ ocn/ ice/ lnd/ | | | | +--+--+------+ +--+--+ +---+--+ +--+--+ | | | | | | | | | cam2/ datm6/ latm6/ pop/ docn6/ csim4/ dice6/ clm2/ dlnd6/
Note that there is one input data directory for every model. Each model directory in turn contains sub-directories containing input data for particular model component. In addition, some of above component directories also contain subdirectories of data. These are listed below:
Directory | Description |
atm/cam2/inic/gaus | cam initial files for spectral dycores |
atm/cam2/inic/fv | cam initial files for finite-volume dycore |
atm/cam2/rad | cam solar constant ramps, absorptivity/emissivity, aerosol |
optical properties and prescribed aerosol distributions | |
atm/cam2/ozone | cam prescribed ozone distributions |
atm/cam2/ggas/ | cam greenhouse gas chemistry and ramping datasets |
atm/cam2/scyc/ | cam sulfur cycle chemistry datasets |
atm/datm6/ | datm cycling data |
atm/latm6/ | latm cycling data |
ocn/pop/gx1v3/ic/ | pop initial conditions datasets |
ocn/pop/gx3v5/ic/ | pop initial conditions datasets |
ocn/pop/gx1v3/forcing/ | pop forcing datasets |
ocn/pop/gx3v5/forcing/ | pop forcing datasets |
ocn/pop/gx1v3/grid/ | pop grid-related input datasets |
ocn/pop/gx3v5/grid/ | pop grid-related input datasets |
ocn/docn6/ | docn cycling data |
ice/csim4/ | csim data |
ice/dice6/ | dice cycling data |
lnd/clm2/inidata_2.1/ccsm | clm initial data |
lnd/clm2/srfdata/csm | clm surface data |
lnd/clm2/rtmdata | clm river-routing (RTM) data |
lnd/clm2/pftdata | clm pft-physiology data |
lnd/clm2/rawdata | clm raw data to create surface datasets at model resolution |
lnd/dlnd6 | dlnd cycling data |
cpl/cpl6/ | coupler data |
The initial and boundary datasets for each component will be found in their respective subdirectories. Because it could take up several Gigabytes of disk space depending upon the resolution of the input datasets downloaded, the input data tar files should be untarred on a disk with sufficient space.
Along with the released input datasets, we are also providing an empty directory tree, inputdata_user/, that has the same structure as inputdata/, but without any files. This tree is provided for user-created input datasets. It should be untarred in a directory parallel to the inputdata/ directory. It should not be untarred under inputdata/. If the user intends to replace an existing inputdata/ dataset with their own customized dataset, the new dataset should be give a unique filename and placed in the corresponding directory as the original inputdata/ file that it replaces. More details are provided in section 6.12.
After downloading the model source code and data sets, users are strongly advised to validate the download by running low-resolution tests described in section 7.3.1. T31 and gx3v5 data sets are required to run the recommended test suite.