job_settings

Owned by Burkhardt Rockel (Unlicensed)
Last updated: Apr 25, 2023 by Beate Geyer
6 min read

This file includes the main settings for the simulation and is called at the beginning of each script in the job chain.

Environment Variable

Description

EXPID

job identification

Main directory settings

SPDIR

Absolut parent path of the SPICE directory. SPDIR only occurs in job_settings. This is not necessary to set if you define the different paths by yourself.

PFDIR

Parent directory where the scripts for the experiment are stored. PFDIR should be on a disk partition that is saved frequently as a copy.

WORKDIR

NONE of the created files under this directory will be deleted at the end of the simulation. The disk partition should have enough remaining disk space.

SCRATCHDIR

ALL created files under this directory will be deleted at the end of the job chain. This disk partition is only for temporary file storage.

SRCDIR

directory of supplementary programs

DATADIR

directory holding supplementary data (grid descriptions etc.)

ARCHIVE_OUTDIR

Directory where the results should be archived. "Results" are the output of the ICON output streams define in the namelist block gribout.

RESDIR

Directory where the restart files will be stored

INIDIR

Directory where some initial data are saved (e.g. the external data of the driving model)

Information on the time period of the simulation

YDATE_START

Start date of simulation YYYYMMDDHH[mmss]

YDATE_STOP

End date of simulation YYYYMMDDHH[mmss]

INC_DATE

Time increment for the CCLM job. Standard is one month 01:00:00 . You may want to use a higher frequency, e.g. if a monthly increment exceeds the batch job limits of your computing system.
Valid values are 01:00:00, 00:01:00, 00:02:00, ... , 00:27:00

ITYPE_CALENDAR

Calendar settings. 0 = proleptic gregorian, 1 = 360 day year, 2 = 365 day year

Email and account settings

EMAIL_ADDRESS

Replace this with your email address or leave the environment variable empty to prevent getting emails from your batch system

NOTIFICATION_ADDRESS

The run scripts send notifications to this address, in case it is set. In case of mailx is used this is your email address.

NOTIFICATION_SCRIPT

The script holding the notification instruction including the whole absolute path to the script. It is called like so

${NOTIFICATION_SCRIPT} ${NOTIFICATION_ADDRESS} subject message-file

The following scripts are available under ${SPDIR)/src/notifications:
mailx.sh - for notifications with mailx
mattermost.sh - for notifications in a mattermost hook
pushover.sh - for notifications through pushover https://pushover.net/

You can also define your own notification script holding your favorite notification system.

PROJECT_ACCOUNT

On the computer systems at DKRZ you need to set you project account ID. If your computings system does not have this feature you can leave it open

Directory and binary path settings for utilities

CFU

cfu (climate fortran utilities) program including the absolute path. These are part of the CCLM starter package

CDO

cdo program including the absolute path.

NCO_BINDIR

Absolute path to the directory of NCO utility binaries (e.g. ncks, ncrcat, ncatted ...)

NC_BINDIR

Absolute path to the directory of netCDF standard binaries (e.g. ncdump, nccopy)

PIGZ

pigz program including the absolute path in case your computing system supports parallel gzip.

PYTHON

python interpreter including the absolute path

Special script settings

ITYPE_INT2LM

Choose whether to run int2lm (=1) or use already existing int2lm output (=0)

ITYPE_CCLM_ARCH_PAR

Submit icon.job.sh in arch.job.sh directly after checking the CCLM output (=1,  i.e. icon simulation and archiving job are running in parallel after the file checking) or at the end of arch.job.sh (=0, i.e. are running sequentially).

ITYPE_COMPRESS_POST

Compression type of output in post-processing
0 = no compression
1 = internal compression (compression in netCDF file, requires netCDF Library with HDF-lib and z-lib)
2 = external compression (compression with gzip, requires gzip version >=1.6 to be installed)
3 = external compression (compression with pigz, requires pigz to be installed), set TASKS_POST  below for the -p option accordingly

ITYPE_COMPRESS_ARCH

Compression type of output in archiving
0 = no compression
1 = internal compression (compression in netCDF file, requires netCDF Library with HDF-lib and z-lib)
2 = external compression (compression with gzip, requires gzip version >=1.6 to be installed)
3 = external compression (compression with pigz, requires pigz to be installed), set TASKS_ARCH below for the -p option accordingly

ITYPE_TS

How timeseries are combined
0 = no timeseries
1 = monthly
2 = yearly (in a format that can directly be used as input for the CLM-Community CMOR-Tool)
3 = 1 & 2 (this doubles the post output !!)

ITYPE_SAMOVAR

choose SAMOVAR check
0 = no SAMOVAR check
1 = short SAMOVAR check (just the output of the last day of the month in each "outNN" directory)
2 = long SAMOVAR check (all output files will be checked)

SAMOVAR_EXE

Name of SAMOVAR executable including full path

SAMOVAR_SH

Name of the SAMOVAR shell script including the path. Default: ${SPDIR}/src/samovar/samovar.sh

SAMOVAR_LIST

File in csv format including the maximal value ranges of the output quantities. Default: ${SPDIR}/src/samovar/samovar.csv

Initial and boundary data input

GCM_DATADIR

Directory of the initial and boundary data

GCM_PREFIX

prefix of the data (e.g., caf or cas)

GCM_EXTFILE

external data file (e.g. orography, land-sea mask, …) of the driving model

HINCBOUND

boundary data temporal increment in hours

Climatological and other supplemental data

EXTPAR

External parameters on CCLM grid (can be generated with WebPEP)

ICON specific settings

INT2LM_EXE

Full path and name of INT2LM executable

CCLM_EXE

Full path and name of CCLM executable

STARTLAT_TOT

transformed latitude of the lower left grid point of the total domain (in degrees, N>0)

STARTLON_TOT

transformed longitude of the lower left grid point of the total domain (in degrees, E>0)

POLLAT

latitude of the rotated north pole (in degrees, N>0)

POLLON

longitude of the rotated north pole (in degrees, E>0)

POLGAM

angle between the north poles of the systems

DLAT

grid point distance in meridional direction (in degrees)

DLON

grid point distance in zonal direction (in degrees)

IE_TOT

number of grid points in zonal direction

JE_TOT

number of grid points in meridional direction

KE_TOT

number of grid points in vertical direction

KE_SOIL

number of layers in the multi-layer soil model

CZML_SOIL

depth of the main soil layers in meters

LSSO

.TRUE. = forecast with sub-grid scale orography scheme

LFOREST

.TRUE. = to run with forest data (evergreen and deciduous)

LLAKE

.TRUE. = forecast with lake model

ITYPE_ALBEDO

albedo type

ITYPE_AEROSOL

aerosol type

LSTOMATA

.TRUE. = external minimum stomata resistance (given by EXTPAR)

LEMISS

.TRUE. = external surface emissivity map (given by EXTPAR)

NESTING_STREAM

If you want to use an output stream to save quantities as input for further downscaling put the number of the output streams here

HOUT_INC

time increment for each GRIBOUT hh:mm:ss  should be a multiple of the CCLM time step

HINCMXT

interval for min/max and average of temperature (to be consistent with HOUT_INC in global settings above)

HINCMXU

interval for min/max and average of wind speed (to be consistent with HOUT_INC in global settings above)

NBOUNDCUT

set number of boundary lines to be cut off in the time series data and also before evaluating the data with SAMOA

PLEVS

List of pressure levels in hPa. Must be the same as or a subset of the plev in CCLM namelist.

ZLEVS

List of altitude levels (height over NN). lzint_above_ground=.FALSE. Must be the same as or a subset of the zlev in CCLM namelist.

HLEVS

List of height levels (height over ground). lzint_above_ground=.TRUE. Must be the same as or a subset of the zlev in CCLM namelist.

Parallelization settings

PARTITION_PREP

Select specific node type/partition for prep.job.sh. This can be different from computing system to computing system. At DKRZ is can be compute or shared.

TASKS_PREP

Number of parallel tasks in prep.job.sh

TIME_PREP

requested time in batch queue for prep.job.sh

PARTITION_ARCH

Select specific node type/partition for arch.job.sh. This can be different from computing system to computing system. At DKRZ is can be compute or shared.

TASKS_ARCH

Number of parallel tasks in arch.job.sh

TIME_ARCH

requested time in batch queue for arch.job.sh

PARTITION_POST

Select specific node type/partition for post.job.sh. This can be different from computing system to computing system. At DKRZ is can be compute or shared.

TASKS_POST

Number of parallel tasks in post.job.sh

TIME_POST

requested time in batch queue for post.job.sh

TIME_POST_YEARLY

requested time in batch queue for post.job.sh when yearly time series are computed

PARTITION_INT2LM

Select specific node type/partition for int2lm.job.sh. This can be different from computing system to computing system. At DKRZ is can be compute or shared.

NODES_INT2LM

Number of nodes requested for int2lm.job.sh

NPX_INT2LM

number of grid partitioning in x-direction in INT2LM

NPY_INT2LM

number of grid partitioning in y-direction in INT2LM

TIME_INT2LM

requested time in batch queue for int2lm.job.sh

PARTITION_CCLM

Select specific node type/partition for cclm.job.sh. This can be different from computing system to computing system. At DKRZ is can be compute or shared.

NPX_CCLM

number of grid partitioning in x-direction in CCLM

NPY_CCLM 

number of grid partitioning in y-direction in CCLM

NPIO_CCLM

number of processors dedicated for asynchroneous output

TIME_CCLM

requested time in batch queue for cclm.job.sh