Preparing and running NEMO4 on occigen
Contributors: Nicolas Jourdain, Pierre Mathiot.
NB: the content of this page has been tested with NEMO’s trunk at revision 15469 (“NEMO-4.2 candidate release”) and XIOS’s trunk at revision 2251.
Content:
- Module environment
- Get the NEMO (ocean/sea-ice model) and XIOS (IO server) sources
- Compile XIOS
- Compile NEMO
- Create inputs for the new NEMO configuration
- Create coordinate and bathymetry files
- Create the DOMAINcfg and mesh_mask files
- Create initial state file
- Create files for lateral boundary conditions
- Create files for runoff and sea surface salinity restoring
- Create weights for on-the-fly interpolation of atmospheric forcing
- Running NEMO
If you are new to occigen, see this page for information on how to get an account and connect.
The official documentation on downloading and installing the NEMO code can be found on the NEMO website.
1. Module environment
The steps below can work with several modules, but here is the recommended list:
module purge
module load gcc/8.3.0
module load intel/19.4
module load intelmpi/5.1.3.258
module load netcdf-fortran/4.4.4-intel-19.0.4-intelmpi-2019.4.243
module load netcdf/4.6.3-intel-19.0.4-intelmpi-2019.4.243
module load hdf5/1.10.5-intel-19.0.4-intelmpi-2019.4.243
Note that if you change these modules, the netcdf paths can be found from:
nc-config --libs
2. Get the NEMO (ocean/sea-ice model) and XIOS (IO server) sources
cd $SCRATCHDIR
mkdir -pv models
cd models
svn co http://forge.ipsl.jussieu.fr/nemo/svn/NEMO/trunk nemo_v4_trunk
svn co https://forge.ipsl.jussieu.fr/ioserver/svn/XIOS/trunk xios_trunk
To get specific versions instead of the trunks (versions under development):
svn co https://forge.ipsl.jussieu.fr/nemo/svn/NEMO/releases/r4.0/r4.0.6 nemo_r4.0.6
svn co https://forge.ipsl.jussieu.fr/ioserver/svn/XIOS/branchs/xios-2.5
3. Compile XIOS
cd xios_trunk
We create three files in the arch directory to set up the pathways, compilation options and environment:
cat << EOF > arch/arch-X64_OCCIGENbis.env
#!/bin/bash
module unload netcdf hdf5 intel
module load gcc/8.3.0
module load intel/19.4
module load intelmpi/5.1.3.258
module load netcdf-fortran/4.4.4-intel-19.0.4-intelmpi-2019.4.243
module load netcdf/4.6.3-intel-19.0.4-intelmpi-2019.4.243
module load hdf5/1.10.5-intel-19.0.4-intelmpi-2019.4.243
export HDF5_INC_DIR=/opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/hdf5-1.10.5-nnfz/include
export HDF5_LIB_DIR=/opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/hdf5-1.10.5-nnfz/lib
EOF
cat << EOF > arch/arch-X64_OCCIGENbis.fcm
%CCOMPILER mpiicc
%FCOMPILER mpiifort
%LINKER mpiifort -nofor-main
%BASE_CFLAGS -std=c++11 -diag-disable 1125 -diag-disable 279
%PROD_CFLAGS -O3 -D BOOST_DISABLE_ASSERTS
%DEV_CFLAGS -g -traceback
%DEBUG_CFLAGS -DBZ_DEBUG -g -traceback -fno-inline
%BASE_FFLAGS -D__NONE__
%PROD_FFLAGS -O3
%DEV_FFLAGS -g -O2 -traceback
%DEBUG_FFLAGS -g -traceback
%BASE_INC -D__NONE__
%BASE_LD -lstdc++
%CPP mpicc -EP
%FPP cpp -P
%MAKE gmake
EOF
cat << EOF > arch/arch-X64_OCCIGENbis.path
NETCDF_INCDIR="-I/opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-fortran-4.4.4-hnmh/include -I/opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-4.6.3-ilty/include"
NETCDF_LIBDIR="-L/opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-fortran-4.4.4-hnmh/lib -L/opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-4.6.3-ilty/lib"
NETCDF_LIB="-lnetcdff -lnetcdf"
MPI_INCDIR=""
MPI_LIBDIR=""
MPI_LIB=""
HDF5_INCDIR="-I$HDF5_INC_DIR"
HDF5_LIBDIR="-L$HDF5_LIB_DIR"
HDF5_LIB="-lhdf5_hl -lhdf5 -lhdf5 -lz"
OASIS_INCDIR="-I$PWD/../../oasis3-mct/BLD/build/lib/psmile.MPI1"
OASIS_LIBDIR="-L$PWD/../../oasis3-mct/BLD/lib"
OASIS_LIB="-lpsmile.MPI1 -lscrip -lmct -lmpeu"
EOF
Then compile xios:
./make_xios --prod --full --arch X64_OCCIGENbis --job 8
Once completed, check that you have the executable:
ls bin/xios_server.exe
4. Compile NEMO
cd $SCRATCHDIR/models/nemo_v4_trunk # or any other release
cat << EOF > arch/arch-X64_OCCIGENbis.fcm
# NCDF_HOME root directory containing lib and include subdirectories for netcdf4
# HDF5_HOME root directory containing lib and include subdirectories for HDF5
# XIOS_HOME root directory containing lib for XIOS
# OASIS_HOME root directory containing lib for OASIS
#
# NCDF_INC netcdf4 include file
# NCDF_LIB netcdf4 library
# XIOS_INC xios include file (taken into accound only if key_iomput is activated)
# XIOS_LIB xios library (taken into accound only if key_iomput is activated)
# OASIS_INC oasis include file (taken into accound only if key_oasis3 is activated)
# OASIS_LIB oasis library (taken into accound only if key_oasis3 is activated)
#
# FC Fortran compiler command
# FCFLAGS Fortran compiler flags
# FFLAGS Fortran 77 compiler flags
# LD linker
# LDFLAGS linker flags, e.g. -L<lib dir> if you have libraries
# FPPFLAGS pre-processing flags
# AR assembler
# ARFLAGS assembler flags
# MK make
# USER_INC complete list of include files
# USER_LIB complete list of libraries to pass to the linker
# CC C compiler used to compile conv for AGRIF
# CFLAGS compiler flags used with CC
#
# Note that:
# - unix variables "$..." are accpeted and will be evaluated before calling fcm.
# - fcm variables are starting with a % (and not a $)
#
%NCDF_HOME /opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-4.6.3-ilty/
%NCDF_HOME_FORTRAN /opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-fortran-4.4.4-hnmh/
%HDF5_HOME /opt/software/occigen/spack/linux-rhel7-x86_64/intel-19.0.4/netcdf-fortran-4.4.4-hnmh/
%XIOS_HOME $SCRATCHDIR/models/xios_trunk
%OASIS_HOME $SCRATCHDIR/models/oa3mct
%NCDF_INC -I%NCDF_HOME/inc -I%NCDF_HOME_FORTRAN/inc
%NCDF_LIB -L%NCDF_HOME/lib -lnetcdf -L%NCDF_HOME_FORTRAN/lib -lnetcdff
%XIOS_INC -I%XIOS_HOME/inc
%XIOS_LIB -L%XIOS_HOME/lib -lxios -lstdc++
%OASIS_INC -I%OASIS_HOME/build/lib/mct -I%OASIS_HOME/build/lib/psmile.MPI1
%OASIS_LIB -L%OASIS_HOME/lib -lpsmile.MPI1 -lmct -lmpeu -lscrip
%CPP cpp
%FC mpif90 -c -cpp
%FCFLAGS -i4 -r8 -O3 -fp-model precise -xAVX -fno-alias -traceback
%FFLAGS %FCFLAGS
%LD mpif90
%FPPFLAGS -P -traditional
%LDFLAGS -lstdc++
%AR ar
%ARFLAGS rs
%MK gmake
%USER_INC %XIOS_INC %OASIS_INC %NCDF_INC
%USER_LIB %XIOS_LIB %OASIS_LIB %NCDF_LIB
%CC cc
%CFLAGS -O0
EOF
To know the options of makenemo (the NEMO compiler):
makenemo -h
You can try to compile one of the provided configuration to check that everything is well set up, e.g.:
makenemo -r WED025 -m X64_OCCIGENbis -j 8
ls cfgs/WED025/BLD/bin/nemo.exe # to check that compilation worked
To create and compile a new configuration (e.g. “AMU”, including ocean and sea-ice), do as follows:
export CONFEXE='AMU'
echo "$CONFEXE OCE ICE" >> cfgs/ref_cfgs.txt
mkdir cfgs/$CONFEXE
echo " bld::tool::fppkeys key_xios key_si3" > cfgs/${CONFEXE}/cpp_${CONFEXE}.fcm
makenemo -r ${CONFEXE} -m X64_OCCIGENbis -j 8
ls cfgs/${CONFEXE}/BLD/bin/nemo.exe
To modify some routines, copy them from cfgs/${CONFEXE}/WORK to cfgs/${CONFEXE}/MY_SRC, modify them and recompile (the files in MY_SRC will be compiled instead of those in WORK).
5. Create inputs for the new NEMO configuration
Choose a configuration name (typically
export CONFIG='AMUXL025.L75' # adapt
mkdir -pv ${SCRATCHDIR}/input
mkdir -pv ${SCRATCHDIR}/input/${CONFIG}
5.1. Create coordinate and bathymetry files
Now we are going to use Nico’s pre-processing toolbox to create new NEMO regional configurations (child domain) from parent configurations (global or regional). The following steps are to be done only once (not each time you create a configuration).
cd ${SCRATCHDIR}/input
git clone https://github.com/nicojourdain/BUILD_CONFIG_NEMO.git
# or git clone git@github.com:nicojourdain/BUILD_CONFIG_NEMO.git
cd BUILD_CONFIG_NEMO/GSW-Fortran/test
vi makefile # check FC and NETCDF_INCDIR
make
./gsw_check ## to check (some have to pass, maybe not all of them)
cd ../..
vi compile_ALL.sh # check FC
./compile_ALL.sh
The following steps rely on a preprocessing namelist. You can either use an existing one or create a new one:
vi namelist_${CONFIG}
ln -s -v namelist_${CONFIG} namelist_pre
namelist_pre is the one read by the processing tools.
NB: The file submit.sh is used to submit jobs with headers on clusters, so you may need to adapt it if you use a new machine.
Then, create the bathymetry (possibly including ice-shelf draft) and coordinate files:
./submit.sh extract_bathy_coord 01
ls ../nemo_${CONFIG}/bathy_meter_${CONFIG}.nc
ls ../nemo_${CONFIG}/coordinates_${CONFIG}.nc
If your are happy with the newly created bathymetry (which is from the parent grid), go directly to the next step. Otherwise, if you want to replace the bathymetry (and maybe ice draft) with an interpolation from a dataset, fill the &bathy_special section of the namelist, then remove the previous bathymetry file, e.g. :
rm -f ../nemo_${CONFIG}/bathy_meter_${CONFIG}.nc
Then, if the dataset is on a lon/lat grid :
./submit.sh extract_bathy_special_lonlat 03 30
or if the dataset is on a stereographic grid (you may need to use 60Gb instead of 30 for BedMachine):
./submit.sh extract_bathy_special_stereo 03 30
5.2. Create the DOMAINcfg and mesh_mask files
To make the following description qui general, we define $MY_NEMO as:
export MY_NEMO="${SCRATCHDIR}/models/nemo_v4_trunk" # adapt to the one you compiled
cd ${MY_NEMO}
Then, compile the NEMO tools called DOMAINcfg and REBUILD_NEMO (do not forget to re-load modules if necessary):
cd tools
./maketools -m X64_OCCIGENbis -n DOMAINcfg
ls -al DOMAINcfg/BLD/bin/make_domain_cfg.exe
ls -al DOMAINcfg/BLD/bin/dom_doc.exe
./maketools -m X64_OCCIGENbis -n REBUILD_NEMO
ls -al REBUILD_NEMO/BLD/bin/rebuild_nemo.exe
Then, create mesh_mask_${CONFIG}.nc and domain_cfg_${CONFIG}.nc as follows:
cd ${SCRATCHDIR}/input/nemo_${CONFIG}
mkdir DOMAINcfg
cd DOMAINcfg
ln -s -v ${MY_NEMO}/tools/DOMAINcfg/BLD/bin/make_domain_cfg.exe
ln -s -v ${MY_NEMO}/tools/DOMAINcfg/BLD/bin/dom_doc.exe
cp -p ${MY_NEMO}/tools/DOMAINcfg/namelist_ref .
cp -p ${MY_NEMO}/tools/DOMAINcfg/namelist_cfg .
vi namelist_ref # Look at default values for all namelist parameters (keep unchanged!).
vi namelist_cfg # Set values that should differ from namelist_ref.
# Fill &namdom : very important to keep ln_read_cfg = .false. and set nn_msh = 1
# fill cn_fcoord, cn_topo, cn_fisfd, cn_bath, cn_visfd, ... (see variables in namelist_ref)
# and put same value as BDY conditions for ppsur, ppa0, ppa1, ppkth, ...
# Fill &namcfg : set domain size in 3 dimensions, config name, etc.
# Fill entire &namzgr_isf section when using ice shelves (see namelist_ref).
# Set nn_msh=1 in &namdom section to obtain a mesh_mask file
# and put the seeds in the oceanic part of the domain (namzgr_isf and namclo)
# Adapt the number of nodes/tasks below (e.g. 112 tasks for eAMUXL12.L121 on BDW28 nodes):
cat <<EOF > run.sh
#!/bin/bash
#SBATCH -C BDW28
#SBATCH --nodes=1
#SBATCH --ntasks=8
#SBATCH --ntasks-per-node=8
#SBATCH --threads-per-core=1
#SBATCH -J run_DOMAINcfg
#SBATCH -e run_DOMAINcfg.e%j
#SBATCH -o run_DOMAINcfg.o%j
#SBATCH --time=00:09:00
mpirun -np 8 ./make_domain_cfg.exe
EOF
chmod +x run.sh
sbatch ./run.sh # then wait for the job completion
ls -al mesh_mask_00??.nc
ls -al domain_cfg_00??.nc
ln -s -v ${MY_NEMO}/tools/REBUILD_NEMO/BLD/bin/rebuild_nemo.exe
ln -s -v ${MY_NEMO}/tools/REBUILD_NEMO/rebuild_nemo
# change the 8 in the two lines below if you did not use 8 tasks in run.sh :
rebuild_nemo -d 1 -x 200 -y 200 -z 1 -t 1 mesh_mask 8
rebuild_nemo -d 1 -x 200 -y 200 -z 1 -t 1 domain_cfg 8
dom_doc.exe -n namelist_cfg -d domain_cfg.nc # to save namelist in domain_cfg.nc
mv mesh_mask.nc ../mesh_mask_${CONFIG}.nc
mv domain_cfg.nc ../domain_cfg_${CONFIG}.nc
rm -f mesh_mask_00??.nc domain_cfg_00??.nc
Note that the namelist_cfg can be re-extracted from domain_cfg_${CONFIG}.nc as follows:
ln -s -v ${MY_NEMO}/tools/REBUILD_NEMO/xtrac_namelist.bash
./xtrac_namelist.bash domain_cfg_${CONFIG}.nc restored_namelist_cfg
5.3. Create initial state file
To exctract the CHILD initial state (temperature and salinity) from the PARENT simulation, fill the &init section of the namelist:
./submit.sh extract_istate_TS 01 16
ls -al ../nemo_${CONFIG}/istate_TS_${CONFIG}.nc # check after completion of extract_istate_TS
If you also need an initial state for sea ice (concentration, ice thickness, snow thickness):
./submit.sh extract_istate_sea_ice 01 8
ls -al ../nemo_${CONFIG}/istate_sea_ice_${CONFIG}.nc # check after completion of extract_istate_sea_ice
You can control smoothing and nearest-neighbour interpolation through the namelist options.
5.4. Create files for lateral boundary conditions
First, create the file containing the coordinates of boundaries:
vi namelist_pre # check domain (x,y) size in ../nemo_${CONFIG}/mesh_mask_${CONFIG}.nc
# and fill &bdy, &bdy_east, &bdy_west, &bdy_north, &bdy_south (mind the fortran indexing).
./submit.sh build_coordinates_bdy 01
ls ../nemo_${CONFIG}/coordinates_bdy_${CONFIG}.nc
Note that you can define several segments for each boundary (staircases, see, e.g., namelist_WED12).
Then put data on these points to create the boundary conditions:
vi namelist_pre # fill &bdy_data section
./submit.sh extract_bdy_gridT 04 15 # adapt duration and memory requirements (typ. 5 min./yr for AMUXL025.L75 from ORCA025)
./submit.sh extract_bdy_gridU 05 15 # adapt duration and memory requirements (typ. 6 min./yr for AMUXL025.L75 from ORCA025)
./submit.sh extract_bdy_gridV 05 15 # adapt duration and memory requirements (typ. 7 min./yr for AMUXL025.L75 from ORCA025)
./submit.sh extract_bdy_icemod 01 8
./submit.sh extract_bdy_ssh 01 8
ls -lrt ../nemo_${CONFIG}/BDY
squeue
# once all these jobs have completed:
vi concatenate_yearly_BDY.sh # adapt CONFIG, BDY_DIR, YEARi, YEARf
./concatenate_yearly_BDY.sh # to get yearly files
ls ../nemo_${CONFIG}/BDY
If you want to prescribe barotropic tides at the boundaries:
vi namelist_pre # fill &bdy_tide section
./submit.sh extract_bdy_tides 01 8
ls ../nemo_${CONFIG}/BDY
5.5. Create files for runoff, SSS restoring, chlorophyll, tidal mixing
If you want to use sea surface salinity restoring:
vi namelist_pre # fill &sss_resto section
./submit.sh extract_SSS_restoring 03 15 # adapt duration and memory requirements
ls -lrt ../nemo_${CONFIG}/SSS
squeue
# once the job has completed:
vi concatenate_yearly_SSS.sh # adapt CONFIG, SSS_DIR, YEARi, YEARf
./concatenate_yearly_SSS.sh
ls ../nemo_${CONFIG}/SSS
If you want to prescribe runoff (from rivers and/or iceberg):
vi namelist_pre # fill &runoff section
# if you only have liquid runoff:
./submit.sh extract_runoff 03 8 # adapt duration (and memory) requirements
# if instead, you have iceberg runoff (possibly also liquid runoff):
./submit.sh extract_runoff_icb 03 8 # adapt duration (and memory) requirements
#
ls -lrt ../nemo_${CONFIG}/RNF
squeue
# once the job has completed:
vi concatenate_yearly_runoff.sh # adapt CONFIG, RNF_DIR, YEARi, YEARf
./concatenate_yearly_runoff.sh
ls ../nemo_${CONFIG}/RNF
If you need a chlorophyll file (for solar absorption), you can either extract it from the parent grid or from a regular lon-lat grid:
vi namelist_pre # fill &chloro
# to extract from parent grid:
./submit.sh extract_chloro 01 8
# to extract from a regular lon-lat grid:
./submit.sh extract_chloro_from_lonlat 01 8
ls ../nemo_${CONFIG}/chlorophyll_${CONFIG}.nc
If you use the internal wave mixing parameterisation (De Lavergne et al.), you can extract it from the parent grid as follows:
vi namelist_pre # fill &zdfiwm
./submit.sh extract_zdfiwm 01 8
ls ../nemo_${CONFIG}/zdfiwm_${CONFIG}.nc
5.6. Create weights for on-the-fly interpolation of atmospheric forcing
First compile the WEIGHTS tool:
cd ${MY_NEMO}/tools
./maketools -m X64_OCCIGENbis -n WEIGHTS
ls -al WEIGHTS/BLD/bin/*.exe
cd ${SCRATCHDIR}/input/nemo_${CONFIG}
mkdir WEIGHTS
cd WEIGHTS
for file in ${MY_NEMO}/tools/WEIGHTS/BLD/bin/*.exe ; do ln -s -v $file ; done
Then, prepare the namelist for the bilinear interpolation:
export REANALYSIS="JRA55" # or any other one, e.g. ERA5
cp -p ${MY_NEMO}/tools/WEIGHTS/namelist_bilin namelist_bilin_${REANALYSIS}_${CONFIG}
# link one of your forcing files, e.g.:
ln -s -v $[SCRATCHDIR}/FORCING_SETS/${REANALYSIS}/drowned_tas_${REANALYSIS}_y2005.nc
ncdump -h t2m_${REANALYSIS}_drowned_y2010.nc # check longitude & latitude names
vi namelist_bilin_${REANALYSIS}_${CONFIG} # fill the 3 sections as in the example below
#&grid_inputs
# input_file = "drowned_tas_${REANALYSIS}_y2005.nc"
# nemo_file = "../coordinates_${CONFIG}.nc"
# datagrid_file = "remap_${REANALYSIS}_grid.nc"
# nemogrid_file = "remap_${CONFIG}_grid.nc"
# method = 'regular'
# input_lon = 'lon'
# input_lat = 'lat'
# nemo_lon = 'glamt'
# nemo_lat = 'gphit'
# nemo_mask = 'none'
# nemo_mask_value = 10
# input_mask = 'none'
# input_mask_value = 10
#/
#&remap_inputs
# num_maps = 1
# grid1_file = "remap_${REANALYSIS}_grid.nc"
# grid2_file = "remap_${CONFIG}_grid.nc"
# interp_file1 = "${REANALYSIS}_${CONFIG}_bilin.nc"
# interp_file2 = "${CONFIG}_${REANALYSIS}_bilin.nc"
# map1_name = "${REANALYSIS} to ${CONFIG} bilin Mapping"
# map2_name = "${CONFIG} to ${REANALYSIS} bilin Mapping"
# map_method = 'bilinear'
# normalize_opt = 'frac'
# output_opt = 'scrip'
# restrict_type = 'latitude'
# num_srch_bins = 90
# luse_grid1_area = .false.
# luse_grid2_area = .false.
#/
#&shape_inputs
# interp_file = "${REANALYSIS}_${CONFIG}_bilin.nc"
# output_file = "weights_bilin_${REANALYSIS}_${CONFIG}.nc"
# ew_wrap = 0
#/
Then:
cat <<EOF > run_bilin.sh
#!/bin/bash
#SBATCH -C BDW28
#SBATCH --ntasks=1
#SBATCH --mem=55000
#SBATCH --threads-per-core=1
#SBATCH -J run_WEIGHT
#SBATCH -e run_WEIGHT.e%j
#SBATCH -o run_WEIGHT.o%j
#SBATCH --time=00:09:00
ulimit -s unlimited
./scripgrid.exe namelist_bilin_${REANALYSIS}_${CONFIG}
./scrip.exe namelist_bilin_${REANALYSIS}_${CONFIG}
./scripshape.exe namelist_bilin_${REANALYSIS}_${CONFIG}
EOF
chmod +x run_bilin.sh
sbatch run_bilin.sh
# when job has completed:
ls -al weights_bilin_${REANALYSIS}_${CONFIG}.nc
Then you can procede similarly with namelist_bicub instead of namelist_bilin:
sed -e "s/bilin/bicub/g ; s/bicubear/bicubic/g" namelist_bilin_${REANALYSIS}_${CONFIG} > namelist_bicub_${REANALYSIS}_${CONFIG}
sed -e "s/bilin/bicub/g" run_bilin.sh > run_bicub.sh
chmod +x run_bicub.sh
sbatch run_bicub.sh
# when job has completed:
ls -al weights_bicub_${REANALYSIS}_${CONFIG}.nc
Then to prepare the next steps:
cd ..
ln -s -v WEIGHTS/weights_bilin_${REANALYSIS}_${CONFIG}.nc
ln -s -v WEIGHTS/weights_bicub_${REANALYSIS}_${CONFIG}.nc
6. Running NEMO
Note that examples of namelists and xml files are provided with NEMO and can be found here:
ls -al ${MY_NEMO}/cfgs/WED025/EXP00/*.xml
ls -al ${MY_NEMO}/cfgs/WED025/EXP00/namelist*
To run long jobs, you can use Nico’s toolbox (so far only implemented for occigen), which you can get as follows:
cd ~
# if you work with NEMO4:
git clone git@github.com:nicojourdain/run_nemo.git # if you use github with SSH key
git clone https://github.com/nicojourdain/run_nemo.git # otherwise
# or if you work with NEMO3.6:
git clone --depth 1 --branch r3.6 git@github.com:nicojourdain/run_nemo.git
Then, follow indications on https://github.com/nicojourdain/run_nemo.