= Climatological Run: Read trajectories from files instead of calling TRAJ =

'''ATTENTION: high disk space requirements !!!'''

== MESSy-run without JUBE ==

=== Run with TRAJ, create trajectory files ===

  * clamstraj.nml:
 {{{
timestep_trajout=6,
timestep_trajin=0,
}}}

 * channel.nml
 {{{
OUT_PREC = 1, 2, 2, 1, 1, 1   ! for trajectory output 
OUT_CHANNEL(203)  = 'clamstraj', 2, 2,  1, T,F, T,F,F,F,F, F,F, , ,
}}}
   => output precision: ''NF90_DOUBLE'' <<BR>>
   => one step per output-file
   => output for clamstraj switched on

 * run messy-script

 * save trajectory files 


=== Use previously created trajectory files ===

 *  create new working directory 
  
 * use the same messy-script and namelist-directory 

 * Changes in namelist clamstraj.nml:
 {{{
timestep_trajout=0,
timestep_trajin=6,
dir_trajin='...',
}}}

  * The trajectories must be read in immediately before calling one of the other submodels <<BR>>
  => timestep_trajin must be divisor of all submodel timesteps
     (timestep_chem, timestep_mix, timstep_bmix, timestep_cirrus)
 
 * channel.nml
 {{{
OUT_PREC = 1, 1, 1, 1, 1, 1   ! for production 
OUT_CHANNEL(203)  = 'clamstraj', 2, 2,  1, T,F, F,F,F,F,F, F,F, , ,
}}}
   => output precision: ''NF90_FLOAT'' <<BR>>
   => no trajectory files are written out <<BR>>

 * Change working directory in messy-script
  
 * run messy-script 



== MESSy-run with JUBE-Script ==

=== Run with TRAJ, create trajectory files ===

 * copy JUBE-Script (e.g. ''messy-clams/messy/util/xmessy_jube_clim.xml'') to working directory ''workdir''

 * in workdir/xmessy_jube_clim.xml:

  * change comment: 
    <comment>messy/clams clim-run, create traj files</comment>

  * in parameterset "clamstraj_nml":
  {{{
timestep_trajout=6
timestep_trajin=0
}}}

  * in parameterset "timer_nml":
  {{{
delta_time = 1800
}}}

  * in parameterset "channel_nml":
  {{{
out_prec_netcdf = 2
lout_clamstraj  = T
}}}

 * jube run xmessy_jube_clim.xml

 * cp workdir/messy_bench/000000/000003_exe/work/climtest*_clamstraj.nc ''trajsave_dir''

=== Use previously created trajectory files ===

 * in workdir/xmessy_jube_clim.xml:

  * change comment:
   <comment>messy/clams clim-run, use traj files</comment>

  * in parameterset "clamstraj_nml":
  {{{
timestep_trajin=6
dir_trajin="trajsave_dir"
}}}
  
  * in parameterset "channel_nml":
  {{{
out_prec_netcdf = 1
lout_clamstraj  = F
}}}

 * jube run xmessy_jube_clim.xml