Execution of operadar

Tmatrix lookup tables

To run, operadar needs pre-computed tmatrix lookup tables. You can either create these tables yourself following the lookup tables wiki or request them to clotilde.augros@meteo.fr. The location of the tmatrix tables (directory) has to be written in the configuration file (see below).

Configuration file

The user must set up, beforehand, a configuration file based on the template provided ./configFile/template.py. Please, do not modify the template directly. Instead, make a copy of the file and name it differently. The configuration file options are explained here.

Activate your environment

See the Installation tutorial section. If not already activated, run :

$ source myvenv/bin/activate   <--- virtual env created with pyenv

If you forgot the name of your environment, you can list the available environments with :

find $HOME -name "*activate" -type f  <--- for pyenv

Quick execution

OPERADAR comes with a command line interface. This method is suitable for one file at a time and runs in a terminal. To show the help :

>>> $ python3 -m operadar --help

usage: python3 -m operadar [-h] [--append] [--verbose] filename config

OPERADAR

positional arguments:
  filename    Only the name of the input file, not the entire path. Please use the config file to provide the path to access the file.
  config      Name of the configuration file. Before running the code, you need to create a configuration file in ./configFiles/ based on the template provided.

options:
  -h, --help  show this help message and exit
  --append    Append the computed dual-polarimetric fields into the input file (default value: False).
  --verbose   If argument provided, show more details if activated (default value: False).

A bash script can also be used for the execution:

>>> $ ./exec_operadar.sh 
----------------------------------------------------------------
 This radar forward operator is developped at the CNRM, France. 
----------------------------------------------------------------

Usage: ./exec_operadar.sh -f FILENAME -c CONFIG [--verbose]

  -f FILENAME : Only the filename. Please use the config file to provide the path to access the file.
  -c CONFIG   : Before running the code, you need to create a configuration file in ./configFiles/ based on the template provided.
  --verbose   : Optional, show more details if activated (default value: False)
  --append    : Optional, append the computed dual-polarimetric fields into the input file (default value: False)
                /!\ If this option is activated, no netcdf output file will be saved. Only working for AROME .fa files

Call `operadar()`

operadar() can be called inside any python program. This method is best adapted to users who want to work with multiple files or with varying parameters/configuration across file(s). Here is a way of doing so.

In a novel python script, import operadar function :
```
from operadar.forward_operator import operadar
```
You will need to load the configuration file before calling operadar so the module can access or update the settings.
```
from operadar.load_config_file import load_configuration_file
```

Let's pretend you are working with two different study cases, that requires to read two different config files, and two different microphysics. Then :

cases = [('conv','2022-06-20 03:00','2022-06-20 11:00'),
         ('strat','2024-11-19 06:00','2024-11-19 22:00'),
        ]
microphysics_schemes = ['ICE3','LIMA']

for case_type,begin,end in cases :
for micro in microphysics_schemes :
    # Read tables and loop over the datetime (again) when changing microphysics and case stuy     
    read_tables = True 
    dict_tables = {}
    ech = pd.to_datetime(begin, format="%Y-%m-%d %H:%M")
    end = pd.to_datetime(end, format="%Y-%m-%d %H:%M")
    # Reload the corresponding configuration in the python environment
    config = load_configuration_file(f'conf_{micro}_{case_type}.py')
    # Execute operadar for each arome file
    while ech <= end :
        time = ech.strftime('%H:%M')
        print('\n','------------------------------------------',case_type,micro,time,'------------------------------------------')
        fname = f'historic.arome.franmg-01km30+00{time}.fa'
        read_tables, dict_tables = operadar(filename=fname,
                                            configuration=config,
                                            read_tables=read_tables,
                                            tables_content = dict_tables,
                                            get_more_details=False,
                                           )
        ech += dt.timedelta(minutes=5)

Please note that you can override all arguments of the configuration file if they are provided as input arguments of operadar.

`operadar()` arguments

This software has been designed so the user can loop on multiple files and eventually with varying configurations over files that share common config parameters. Thus, all the configuration parameters can be overwritten when calling operadar(). When arguments are optional, the default value is always red in the last copied configuration file :

Parameter	Status	Description
`filename`	mandatory	Only the name of the file.
`modelname`	optional	Can be either `'Arome'`, `'MesoNH'` or `'WRF'`.
`read_tables`	optional	Option to save computing time within a loop. Needs to always be `True` for the first iteration. Defaults to `True`.
`in_dir_path`	optional	Path where the input files are stored.
`out_dir_path`	optional	Path to store the output files.
`tables_path`	optional	Path where the Tmatrix lookup tables are stored.
`microphysics_scheme`	optional	Can be `ICE3`, `ICE4` or `LIMA` + a name extension (e.g. `LIMA_noHail` or `ICE3_CIBU_moins`), which is optional. Please note that only the three first characters are used to select the right scheme in the lookup tables. Then, the microphysics and correct computations are handles with the `hydrometeorMoments` argument.
`hydrometeorMoments`	optional	Dictionary of form `{'hydrometeor_key' : number of moment}` corresponding to the microphysics scheme.
`radar_band`	optional	Available bands : `'C'`, `'X'`, `'S'`, `'W'`, `'L'`, `'Ka'`, or `'K'`
`radarloc`	optional	Location of the radar to emulate radar geometry. Either `'center'` (i.e. center of the grid) or a `[lat_radar,lon_radar]` coordinate.
`distmax_rad`	optional	Maximum radar radius to compute pseudo-observations.
`tables_content`	optional	Dictionary containing the lookup tables parameters. Used to pass the dictionary throughout loop iterations. Please, also read the To go further section.
`mixed_phase_parametrization`	optional	Can be either : `'T_pos'` : the species content is transferred to the melting species only at positive temperatures. `'Fw_pos'` : the rain and graupel content are emptied and transferred into the wet graupel content within the melting layer. `'Fw_posg'` : only the graupel content is emptied and transferred to the wet graupel content within the melting layer.
`subDomain`	optional	This argument can be used to reduce the size of the output file, so you can chunk the output into smaller domain and thus lower the size of the outputs. Or, you want to work on different areas over the same file. Should be defined like `[lon_min,lon_max,lat_min,lat_max]` or set to `None` to work on all grid points.
`get_more_details`	optional	Boolean to print more details and computation steps.

To go further

radar_band and read_tables : the Tmatrix lookup tables are red once, for the radar band set in the config file, at the beginning of the code. To save computation time, it is not necessary to read the tables as long as the radar band remains the same, and thus, read_tables is automatically set to False after the first iteration that produces an output file. If you want to change the radar band over the iterations, you must also set read_tables=True. Based on the tutorial :

   import itertools

   fname_list = ['historic.arome.franmg-01km30+0006:00.fa',
                 'historic.arome.franmg-01km30+0006:05.fa',
                ]
   radar_band = ['C','S','X']
   read_tables = [True]

   # Creating the unique combinations
   combinations = list(itertools.product(fname_list,radar_band,read_tables))

will produce

>>> $ python3 operadar_multi.py
historic.arome.franmg-01km30+0006:00.fa C True
historic.arome.franmg-01km30+0006:00.fa S True
historic.arome.franmg-01km30+0006:00.fa X True
historic.arome.franmg-01km30+0006:05.fa C True
historic.arome.franmg-01km30+0006:05.fa S True
historic.arome.franmg-01km30+0006:05.fa X True

and can then be used like

   for fname,band,read_tmat in combinations :
      read_tmat,dict_Tmatrix = operadar(filename=fname,
                                        read_tables=read_tmat,
                                        radar_band=band,
                                        get_more_details=True,
                                        )

out_dir_path and in_dir_path : similarly, one may want to store the output files in different folders depending on the radar band...

   for band in radar_band :
      read_tables=True
      dict_Tmatrix={}
      outPath = f'/home/my_path/output/{band}band_simus/'
      for fname in fname_list :
         read_tables,dict_Tmatrix = operadar(filename=fname,
                                              read_tables=read_tmat,
                                              radar_band=band,
                                              Tmatrix_params=dict_Tmatrix,
                                              out_dir_path=outPath,
                                             )

... or provide files from different folders :

   fpath_list = ['xp_arome/GOV2/',
                 'xp_arome/GOV5/',
                ]
   fname_list = ['historic.arome.franmg-01km30+0006:00.fa',
                 'historic.arome.franmg-01km30+0006:05.fa',
                ]
   combinations = list(itertools.product(fpath_list,fname_list))
   band='X'
   read_tables=True
   dict_Tmatrix={}
   for fpath,fname in combinations :
      read_tables,dict_Tmatrix = operadar(filename=fname,
                                          read_tables=read_tmat,
                                          radar_band=band,
                                          Tmatrix_params=dict_Tmatrix,
                                          in_dir_path=fpath,
                                          )

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Execution of operadar

Tmatrix lookup tables

Configuration file

Activate your environment

Quick execution

Call `operadar()`

`operadar()` arguments

To go further

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Clone this wiki locally

Uh oh!

Execution of operadar

Tmatrix lookup tables

Configuration file

Activate your environment

Quick execution

Call operadar()

operadar() arguments

To go further

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Clone this wiki locally

Call `operadar()`

`operadar()` arguments