1.6. Build CROCO initial and boundary conditions#
This section presents the use of make_ini.py and make_bry.py
to create initial and boundary condtions (ibc) for
realistic regional configurations with CROCO.
The different steps to build ibc are:
Extrapolating values on mask
Interpolating each input for each Z-level on the CROCO horizontal grid
Interpolating on the vertical sigma-coordinates
1.6.1. Fill the Reader#
In this tutorial, we will use as initial and boundary conditions the
Mercator product GLORYS12, but already processed
in a CROCO-like format by tools available for downloading Mercator data in croco_tools.
An extraction from January 2013 to March 2013 is available in the tutorial section of CROCO website: https://www.croco-ocean.org/trainings/, in the section ‘Regional configuration (#define REGIONAL in cppdefs.h)’ under ‘Global datasets needed for preprocessing’. Here is the link to download them: https://data-croco.ifremer.fr/CONFIGS_EXAMPLES/DATASETS_GLOB_INTER/CROCO/MERCATOR_GLOB_201301.tar.gz
The following 3 files are provided:
mercator_Y2013M1.cdf, mercator_Y2013M2.cdf, mercator_Y2013M3.cdf.
These files are not “original” mercator files, and we will use the mercator_croco keyname to differentiate their formatting.
You can check the variables name from these input data by doing:
ncdump -h mercator_Y2013M1.cdf
You thus, first need to check/edit the ibc reader, Readers/ibc_reader.py, for these data.
Here we use the mercator_croco style of formatting:
if input == 'mercator_croco':
dico={ 'depth':'depth',\
'lonr':'lonT','lonu':'lonU','lonv':'lonV',\
'latr':'latT','latu':'latU','latv':'latV',\
'ssh':'ssh',\
'temp':'temp',\
'salt':'salt',\
'u': 'u',\
'v': 'v',\
'time': 'time',\
'time_dim':'time'\
}
Note
If you have downloaded mercator data on your own, with a more “mercator-native” format, you may rather use the mercator dictionnary:
if input == 'mercator':
dico={ 'depth':'depth',\
'lonr':'longitude','lonu':'longitude','lonv':'longitude',\
'latr':'latitude','latu':'latitude','latv':'latitude',\
'ssh':'zos',\
'temp':'thetao',\
'salt':'so',\
'u': 'uo',\
'v': 'vo',\
'time': 'time',\
'time_dim':'time'\
}
1.6.2. Using make_ini#
Before running make_ini.py, USER CHANGES section needs to be filled.
There are several parts in it:
Dates:
# Dates
# starting date
Yini, Mini, Dini, Hini = '2013','01','01','12' # Month and days need to be 2-digits format
# reference time (default = ini time)
Yorig, Morig, Dorig = Yini, Mini, Dini # Month and days need to be 2-digits format
Note
To start your simulation at midnight when the input data is at noon, simply set Hini to '00'. This will not interpolate the solution to the given time, but simply change the start time to midnight.
Note
Origin time and initial time can be different.
Input data informations contain the general informations about the inputs (reader keyword, directory, prefix)
# Input data information and formating inputdata = 'mercator_croco' # Input data dictionnary as defined in the Readers/ibc_reader.py input_dir = '../../MERCATOR_GLOB_2013/' input_prefix='mercator_' input_file = f'{input_dir}{input_prefix}Y2013M1.cdf' multi_files=False # If variables are in different netcdf if multi_files: # Mutiple files input_file = { 'ssh' : input_dir + input_prefix + 'ETAN.Y2013M01.nc',\ 'temp' : input_dir + input_prefix + 'THETA.Y2013M01.nc',\ 'salt' : input_dir + input_prefix + 'SALT.Y2013M01.nc',\ 'u' : input_dir + input_prefix + 'EVEL.Y2013M01.nc',\ 'v' : input_dir + input_prefix + 'NVEL.Y2013M01.nc'\ }
# time index to use in the file tndx = 0
inputdata needs to be set to the keyword to use in the reader, as defined in Readers/ibc_reader.py.
For input source in which variables are in different netcdf files, you can set multi_files to True, then you
can define one netcdf file for each variable in the following section.
input_file is the input filename (full path + filename). In the example, it is defined with the input_dir, input_prefix, and
simulation start date (Yini,Mini,Dini).
If input data files contains several times, the user must select the time index for starting the simulations with
tndx to select one (0 means first index).
Nzgoodmin defines a threshold above which a z-level in the input file is valid. If at a certain depth this threshold is not reached, the level will be considered as not intersecting the ocean.
tracers
# tracers tracers = ['temp','salt']
Set tracer names here, if any. Name defined here must also be defined in your reader.
CROCO grid informations
# CROCO grid informations croco_dir = '../../CROCO_FILES/' croco_grd = 'croco_grd.nc' sigma_params = dict(theta_s=7, theta_b=2, N=32, hc=200) # Vertical streching, sig_surf/sig_bot/ nb level/critical depth
Informations about your CROCO grid. Indicate the path (croco_dir),
the input grid to use (croco_grd), the parameters for the
sigma-coordinates
Ini filename to be generated (it will follow the pattern indicated + the date):
# Ini file informations
ini_filename = 'croco_ini.nc' # output will be put in croco_dir by default
Name of your output file. This file will be written the previoulsy set croco_dir.
To use make_ini.py, do:
python make_ini.py
You will find the ouput in croco_dir:
croco_ini_mercator_croco_Y2013M01.nc
1.6.3. Using make_bry#
make_bry is quite similar to make_ini, but you will have serveral time frames in the output bry file.
As for make_ini, USER CHANGES section of make_bry needs to be filled.
# Dates
Yorig = 2013 # year defining the origin of time as: days since Yorig-01-01
Ystart, Mstart = '2013', '01' # Starting month
Yend, Mend = '2013','03' # Ending month
# Input data information and formating
inputdata = 'mercator_croco' # Input data dictionnary as defined in the Readers/ibc_reader.py
input_dir = '../../MERCATOR_GLOB_2013/'
input_prefix = 'mercator_*' # Please use * to include all files
multi_files = False
if multi_files: # Multiple data files. Time is read in ssh file
input_file = {'ssh':sorted(glob.glob(input_dir+input_prefix+'ETAN.*.nc')),\
'temp':sorted(glob.glob(input_dir+input_prefix+'THETA.*.nc')),\
'salt':sorted(glob.glob(input_dir+input_prefix+'SALT.*.nc')),\
'u':sorted(glob.glob(input_dir+input_prefix+'EVEL.*.nc')),\
'v':sorted(glob.glob(input_dir+input_prefix+'NVEL.*.nc'))\
}
else: # glob all files
input_file = sorted(glob.glob(input_dir + input_prefix))
# default value to consider a z-level fine to be used
Nzgoodmin = 4
# Tracers
tracers = ['temp', 'salt']
# CROCO grid informations
croco_dir = '../../CROCO_FILES/'
croco_grd = 'croco_grd.nc'
sigma_params = dict(theta_s=7, theta_b=2, N=32, hc=200) # Vertical streching, sig_surf/sig_bot/ nb level/critical depth
# Bry file informations
bry_filename = 'croco_bry.nc' # output will be put in croco_dir by default
obc_dict = dict(south=1, west=1, east=1, north=1) # open boundaries (1=open , [S W E N])
output_file_format = "MONTHLY" # How outputs are spit (MONTHLY,YEARLY,FULL)
cycle_bry = 0.
In bry case , several days/months need to be given. To facilitate input
selection we use python glob module
which finds all the pathnames matching a specified pattern according to the
rules used by the Unix shell.
In input_prefix, select a specified
pattern and use Unix shell rules (*, $) to select all files following it.
In the last part, ‘Bry file informations’, output file name (bry_filename) is defined. Select
which boundary to open (by putting 1 to the corresponding boundary in
obc_dict). Several formats (output_file_format) exist to gather data
by month, year or put them all in the same file.
cycle is used if you want to create a cycle in number of days on bry conditions. Typically, to have a repetition of a year of boundary forcing cycle would be set to 365
Origin year is also defined in this part, along with starting and ending date.
To use make_bry.py do:
python make_bry.py