From 853b2838bc42d16d40a675b5413ced969e2e8dae Mon Sep 17 00:00:00 2001
From: Lukas Kugler <lukas.kugler@univie.ac.at>
Date: Tue, 11 Feb 2025 21:43:46 +0100
Subject: [PATCH] work in progress
---
config/jet.py | 107 ++++-----
dartwrf/assimilate.py | 12 +-
dartwrf/create_obskind_table.py | 34 +--
dartwrf/create_wbubble_wrfinput.py | 2 +-
dartwrf/exp_config.py | 173 ---------------
dartwrf/namelist_handler.py | 88 ++++++++
dartwrf/obs/__init__.py | 28 +++
dartwrf/obs/obskind.py | 338 -----------------------------
dartwrf/obs/obsseq.py | 4 +-
dartwrf/obs/run_obs_diag.py | 2 +-
dartwrf/prep_IC_prior.py | 4 +-
dartwrf/prepare_namelist.py | 128 ++++++-----
dartwrf/prepare_wrfrundir.py | 13 +-
dartwrf/run_ens.jet.sh | 25 ---
dartwrf/run_ens.vsc.gfortran.sh | 42 ----
dartwrf/run_ens.vsc.mem_0096.sh | 43 ----
dartwrf/run_ens.vsc.sh | 40 ----
dartwrf/run_ens.vsc.single.sh | 15 --
dartwrf/server_config.py | 107 ++++-----
dartwrf/update_IC.py | 2 +-
dartwrf/utils.py | 225 +++++++++++--------
dartwrf/workflows.py | 329 ++++++++++++++--------------
docs/source/tutorial1.rst | 10 +-
multiple_exps.py | 79 +++++++
24 files changed, 701 insertions(+), 1149 deletions(-)
delete mode 100644 dartwrf/exp_config.py
create mode 100644 dartwrf/namelist_handler.py
delete mode 100644 dartwrf/obs/obskind.py
delete mode 100755 dartwrf/run_ens.jet.sh
delete mode 100755 dartwrf/run_ens.vsc.gfortran.sh
delete mode 100755 dartwrf/run_ens.vsc.mem_0096.sh
delete mode 100755 dartwrf/run_ens.vsc.sh
delete mode 100755 dartwrf/run_ens.vsc.single.sh
create mode 100644 multiple_exps.py
diff --git a/config/jet.py b/config/jet.py
index ebb475d..df71c91 100755
--- a/config/jet.py
+++ b/config/jet.py
@@ -1,55 +1,56 @@
"""Cluster configuration file, see docstring of ClusterConfig class in dartwrf/utils.py for details"""
-from dartwrf import utils
-from dartwrf.exp_config import exp
-
-cluster = utils.ClusterConfig(exp)
-cluster.name = 'jet'
-cluster.max_nproc = 20
-cluster.use_slurm = True
-cluster.size_WRF_jobarray = exp.n_ens
-cluster.np_WRF = 16
-
-# binaries
-cluster.python = '/jetfs/home/lkugler/miniforge3/envs/verif/bin/python'
-cluster.ncks = '/jetfs/spack/opt/spack/linux-rhel8-skylake_avx512/intel-2021.7.1/nco-5.1.0-izrhxv24jqco5epjhf5ledsqwanojc5m/bin/ncks'
-cluster.ideal = '/jetfs/home/lkugler/data/compile/bin/ideal-v4.6.0_20250210_StS.exe'
-cluster.wrfexe = '/jetfs/home/lkugler/data/compile/bin/wrf-v4.6.0_20250210_StS.exe'
-cluster.dart_modules = 'module purge; module load rttov/v13.2-gcc-8.5.0'
-cluster.wrf_modules = """module purge; module load netcdf-fortran/4.5.3-intel-2021.7.1-27ldrnt"""
-
-# paths for data output
-cluster.wrf_rundir_base = '/jetfs/home/lkugler/data/run_WRF/' # path for temporary files
-cluster.dart_rundir_base = '/jetfs/home/lkugler/data/run_DART/' # path for temporary files
-cluster.archive_base = '/jetfs/home/lkugler/data/sim_archive/'
-
-# paths used as input
-cluster.srcdir = '/jetfs/home/lkugler/data/compile/WRF-4.3/run'
-cluster.dart_srcdir = '/jetfs/home/lkugler/data/compile/DART/DART-10.8.3_10pct/models/wrf/work/'
-cluster.rttov_srcdir = '/jetfs/home/lkugler/data/compile/RTTOV13/rtcoef_rttov13/'
-cluster.dartwrf_dir = '/jetfs/home/lkugler/DART-WRF/'
-
-# other inputs
-cluster.geo_em_nature = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.250m_1600x1600'
-cluster.geo_em_forecast = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.2km_200x200'
-#cluster.obs_impact_filename = cluster.dartwrf_dir+'/templates/impactfactor_T.txt'
-cluster.namelist = cluster.dartwrf_dir+'/config/templates/namelist.input_nat_exact'
-cluster.rttov_nml = cluster.dartwrf_dir + "/config/templates/obs_def_rttov.VIS+WV.nml"
-cluster.run_WRF = '/jetfs/home/lkugler/DART-WRF/dartwrf/run_ens.jet.sh'
-
-cluster.slurm_cfg = {"account": "lkugler", "partition": "all",
- "ntasks": "1", "ntasks-per-core": "1", "mem": "30G",
- "mail-type": "FAIL", "mail-user": "lukas.kugler@univie.ac.at"}
-
-# WRF file format, will only change if WRF changes
-cluster.wrfout_format = '/wrfout_d01_%Y-%m-%d_%H:%M:%S'
-
-# pattern for the init_time folder in sim_archive
-cluster.pattern_init_time = "/%Y-%m-%d_%H:%M/"
-
-# how an obs_seq.out file is archived
-cluster.pattern_obs_seq_out = cluster.archivedir + \
- "/diagnostics/%Y-%m-%d_%H:%M_obs_seq.out"
+from re import A
+from dartwrf.utils import ClusterConfig
+
+cluster = ClusterConfig(
+ max_nproc = 20,
+ max_nproc_for_each_ensemble_member = 16,
+ use_slurm = True,
+
+ # binaries
+ python = '/jetfs/home/lkugler/miniforge3/envs/verif/bin/python',
+ ncks = '/jetfs/spack/opt/spack/linux-rhel8-skylake_avx512/intel-2021.7.1/nco-5.1.0-izrhxv24jqco5epjhf5ledsqwanojc5m/bin/ncks',
+ ideal = '/jetfs/home/lkugler/data/compile/bin/ideal-v4.6.0_20250210_StS.exe',
+ wrfexe = '/jetfs/home/lkugler/data/compile/bin/wrf-v4.6.0_20250210_StS.exe',
+ dart_modules = 'module purge; module load rttov/v13.2-gcc-8.5.0',
+ wrf_modules = """module purge; module load netcdf-fortran/4.5.3-intel-2021.7.1-27ldrnt""",
+
+ # paths for data output
+ wrf_rundir_base = '/jetfs/home/lkugler/data/run_WRF/', # path for temporary files
+ dart_rundir_base = '/jetfs/home/lkugler/data/run_DART/', # path for temporary files
+ archive_base = '/jetfs/home/lkugler/data/sim_archive/',
+
+ # paths used as input
+ srcdir = '/jetfs/home/lkugler/data/compile/WRF-4.3/run',
+ dart_srcdir = '/jetfs/home/lkugler/data/compile/DART/DART-10.8.3_10pct/models/wrf/work/',
+ rttov_srcdir = '/jetfs/home/lkugler/data/compile/RTTOV13/rtcoef_rttov13/',
-# how an obs_seq.final file is archived
-cluster.pattern_obs_seq_final = cluster.archivedir + \
- "/diagnostics/%Y-%m-%d_%H:%M_obs_seq.final"
+ dartwrf_dir_dev = '/jetfs/home/lkugler/DART-WRF/',
+ WRF_namelist_template = '/jetfs/home/lkugler/DART-WRF/config/templates/namelist.input_nat_exact',
+ rttov_nml = "/jetfs/home/lkugler/DART-WRF/config/templates/obs_def_rttov.VIS+WV.nml",
+
+ # other inputs
+ geo_em_nature = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.250m_1600x1600',
+ geo_em_forecast = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.2km_200x200',
+ #obs_impact_filename = dartwrf_dir_dev+'/templates/impactfactor_T.txt'
+
+ WRF_exe_template = '/jetfs/home/lkugler/DART-WRF/config/templates/run_WRF.jet.sh',
+ WRF_ideal_template = '/jetfs/home/lkugler/DART-WRF/config/templates/run_WRF_ideal.sh',
+
+ slurm_cfg = {"account": "lkugler", "partition": "all",
+ "ntasks": "1", "ntasks-per-core": "1", "mem": "30G",
+ "mail-type": "FAIL", "mail-user": "lukas.kugler@univie.ac.at"},
+
+ # WRF file format, will only change if WRF changes
+ wrfout_format = '/wrfout_d01_%Y-%m-%d_%H:%M:%S',
+
+ # pattern for the init_time folder in sim_archive
+ pattern_init_time = "/%Y-%m-%d_%H:%M/",
+
+ # how an obs_seq.out file is archived
+ pattern_obs_seq_out = "<archivedir>/diagnostics/%Y-%m-%d_%H:%M_obs_seq.out",
+
+ # how an obs_seq.final file is archived
+ pattern_obs_seq_final = "<archivedir>/diagnostics/%Y-%m-%d_%H:%M_obs_seq.final",
+
+ )
\ No newline at end of file
diff --git a/dartwrf/assimilate.py b/dartwrf/assimilate.py
index 4059257..71ebb75 100755
--- a/dartwrf/assimilate.py
+++ b/dartwrf/assimilate.py
@@ -71,7 +71,7 @@ def prepare_prior_ensemble(assim_time, prior_init_time, prior_valid_time, prior_
- removes probably pre-existing files which could lead to problems
"""
print("prepare prior ensemble")
- for iens in range(1, exp.n_ens + 1):
+ for iens in range(1, exp.ensemble_size + 1):
print("link wrfout file to DART background file")
wrfout_run = (
@@ -115,7 +115,7 @@ def use_linked_files_as_prior():
"""Instruct DART to use the prior ensemble as input
"""
files = []
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
files.append("./prior_ens" + str(iens) + "/wrfout_d01")
write_txt(files, cluster.dart_rundir+'/input_list.txt')
@@ -124,7 +124,7 @@ def use_filter_output_as_prior():
"""Use the last posterior as input for DART, e.g. to evaluate the analysis in observation space
"""
files = []
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
f_new = cluster.dart_rundir+'/prior_ens'+str(iens)+'/wrfout_d01'
try:
os.remove(f_new)
@@ -139,7 +139,7 @@ def use_filter_output_as_prior():
def write_list_of_outputfiles():
files = []
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
files.append("./filter_restart_d01." + str(iens).zfill(4))
write_txt(files, cluster.dart_rundir+'/output_list.txt')
@@ -187,7 +187,7 @@ def archive_filteroutput(time):
copy(cluster.dart_rundir + "/input.nml", dir_out + "/input.nml")
# copy filter_restart files to archive (initial condition for next run)
- for iens in range(1, exp.n_ens + 1): # single members
+ for iens in range(1, exp.ensemble_size + 1): # single members
copy(
cluster.dart_rundir + "/filter_restart_d01." + str(iens).zfill(4),
dir_out + "/filter_restart_d01." + str(iens).zfill(4),
@@ -206,7 +206,7 @@ def archive_filteroutput(time):
try:
ftypes = ['preassim', 'postassim']
for ftype in ftypes:
- for iens in range(1, exp.n_ens + 1):
+ for iens in range(1, exp.ensemble_size + 1):
fname = "/"+ftype+"_member_" + str(iens).zfill(4) + ".nc"
copy(cluster.dart_rundir + fname, dir_out + fname)
except Exception as e:
diff --git a/dartwrf/create_obskind_table.py b/dartwrf/create_obskind_table.py
index 2c395bf..5994018 100644
--- a/dartwrf/create_obskind_table.py
+++ b/dartwrf/create_obskind_table.py
@@ -29,35 +29,6 @@ def _dict_to_py(d, outfile):
f.write(txt)
-def _obskind_read(dart_srcdir):
- """Read dictionary of observation types + ID numbers ("kind")
- from DART f90 script and return it as python dictionary
- """
- definitionfile = dart_srcdir + \
- '/../../../assimilation_code/modules/observations/obs_kind_mod.f90'
- with open(definitionfile, 'r') as f:
- kind_def_f = f.readlines()
-
- obskind_nrs = {}
- for i, line in enumerate(kind_def_f):
- if 'Integer definitions for DART OBS TYPES' in line:
- # data starts below this line
- i_start = i
- break
- for line in kind_def_f[i_start+1:]:
- if 'MAX_DEFINED_TYPES_OF_OBS' in line:
- # end of data
- break
- if '::' in line:
- # a line looks like this
- # integer, parameter, public :: MSG_4_SEVIRI_TB = 261
- data = line.split('::')[-1].split('=')
- kind_str = data[0].strip()
- kind_nr = int(data[1].strip())
- obskind_nrs[kind_str] = kind_nr
- return obskind_nrs
-
-
def _save_config_to_scriptsdir(server_config, original_scripts_dir):
try:
dir_path = os.path.dirname(os.path.realpath(__file__))
@@ -80,10 +51,9 @@ def run(server_config='jet.py'):
# import the config from scripts_dir
sys.path.append(original_scripts_dir)
from server_config import cluster
+ from dartwrf import obs
- dart_srcdir = cluster.dart_srcdir
-
- obskind_dictionary = _obskind_read(dart_srcdir)
+ obskind_dictionary = obs.obskind_read(cluster.dart_srcdir)
_dict_to_py(obskind_dictionary, original_scripts_dir+'/obs/obskind.py')
print('>>>', original_scripts_dir+'/obs/obskind.py', 'created')
diff --git a/dartwrf/create_wbubble_wrfinput.py b/dartwrf/create_wbubble_wrfinput.py
index bc79875..6fd8cac 100644
--- a/dartwrf/create_wbubble_wrfinput.py
+++ b/dartwrf/create_wbubble_wrfinput.py
@@ -23,7 +23,7 @@ if __name__ == "__main__":
print('perturb wbubble = ', perturbations)
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
print('iens', iens)
wrfin = cluster.wrf_rundir(iens)+'/wrfinput_d01'
diff --git a/dartwrf/exp_config.py b/dartwrf/exp_config.py
deleted file mode 100644
index e570351..0000000
--- a/dartwrf/exp_config.py
+++ /dev/null
@@ -1,173 +0,0 @@
-from dartwrf.utils import Experiment
-
-exp = Experiment()
-exp.expname = "exp_test"
-exp.model_dx = 2000
-exp.n_ens = 4
-exp.do_quality_control = False
-
-# path to the nature run, where we take observations from
-#exp.nature_wrfout_pattern = '/jetfs/home/lkugler/data/sim_archive/exp_v1.18_P1_nature+1/*/1/wrfout_d01_%Y-%m-%d_%H:%M:%S'
-#exp.nature_wrfout_pattern = '/jetfs/home/lkugler/data/sim_archive/exp_v1.18_P1_nature/*/1/wrfout_d01_%Y-%m-%d_%H:%M:%S'
-exp.nature_wrfout_pattern = '/jetfs/home/lkugler/data/sim_archive/exp_v1.19_P3_wbub7_nat/*/1/wrfout_d01_%Y-%m-%d_%H:%M:%S'
-
-exp.input_profile = '/mnt/jetfs/home/lkugler/data/initial_profiles/wrf/ens/2022-03-31/raso.fc.<iens>.wrfprof'
-
-
-exp.dart_nml = {'&assim_tools_nml':
- dict(filter_kind='1',
- sampling_error_correction='.true.',
- # obs_impact_filename='/jetfs/home/lkugler/DART-WRF/templates/impactfactor_T.txt',
- ),
- '&filter_nml':
- dict( ens_size=exp.n_ens,
- num_output_state_members=exp.n_ens,
- num_output_obs_members=exp.n_ens,
- inf_flavor=['0', '4'],
- inf_initial=[1.04, 0.5],
- inf_initial_from_restart='.false.',
- output_members='.true.',
- output_mean='.true.',
- output_sd='.true.',
- stages_to_write='output',
- ),
- '&quality_control_nml':
- dict(outlier_threshold='-1',
- ),
- '&obs_def_radar_mod_nml':
- dict(apply_ref_limit_to_obs = '.true.',
- reflectivity_limit_obs = 5.0,
- lowest_reflectivity_obs = 5.0,
- apply_ref_limit_to_fwd_op = '.true.',
- reflectivity_limit_fwd_op = 5.0,
- lowest_reflectivity_fwd_op = 5.0,
- microphysics_type = '5',
- ),
- '&location_nml':
- dict(horiz_dist_only='.false.',
- ),
- '&model_nml':
- dict(wrf_state_variables =
- [['U', 'QTY_U_WIND_COMPONENT', 'TYPE_U', 'UPDATE','999',],
- ['V', 'QTY_V_WIND_COMPONENT', 'TYPE_V', 'UPDATE','999',],
- ['W', 'QTY_VERTICAL_VELOCITY', 'TYPE_W', 'UPDATE','999',],
- ['PH', 'QTY_GEOPOTENTIAL_HEIGHT', 'TYPE_GZ', 'UPDATE','999',],
- ['THM', 'QTY_POTENTIAL_TEMPERATURE','TYPE_T', 'UPDATE','999',],
- ['MU', 'QTY_PRESSURE', 'TYPE_MU', 'UPDATE','999',],
-
- ['QVAPOR','QTY_VAPOR_MIXING_RATIO', 'TYPE_QV', 'UPDATE','999',],
- ['QCLOUD','QTY_CLOUDWATER_MIXING_RATIO','TYPE_QC', 'UPDATE','999',],
- ['QICE', 'QTY_ICE_MIXING_RATIO', 'TYPE_QI', 'UPDATE','999',],
- # ['QRAIN','QTY_RAINWATER_MIXING_RATIO','TYPE_QR', 'UPDATE','999',],
- # ['QSNOW','QTY_SNOW_MIXING_RATIO','TYPE_QS', 'UPDATE','999',],
- # ['QGRAUP','QTY_GRAUPEL_MIXING_RATIO','TYPE_QG', 'UPDATE','999',],
-
- ['CLDFRA','QTY_CLOUD_FRACTION', 'TYPE_CFRAC','UPDATE','999',],
- ['PSFC', 'QTY_SURFACE_PRESSURE', 'TYPE_PSFC', 'UPDATE','999',],
- ['T2', 'QTY_2M_TEMPERATURE', 'TYPE_T', 'UPDATE','999',],
- ['TSK', 'QTY_SKIN_TEMPERATURE', 'TYPE_T', 'UPDATE','999',],
- ['REFL_10CM','QTY_RADAR_REFLECTIVITY','TYPE_REFL', 'UPDATE','999',]],
-
- wrf_state_bounds =
- [['QVAPOR','0.0','NULL','CLAMP'],
- ['QCLOUD','0.0','NULL','CLAMP'],
- ['QICE','0.0','NULL','CLAMP'],
- ['CLDFRA','0.0','1.0','CLAMP'],
-
- # ['QRAIN','0.0','NULL','CLAMP'],
- # ['QSNOW','0.0','NULL','CLAMP'],
- # ['QGRAUP','0.0','NULL','CLAMP'],
- ],
- ),
- '&ensemble_manager_nml':
- dict(layout = 1,
- tasks_per_node = 12,
- communication_configuration = 1,
- ),
- }
-
-
-
-# n_obs can be 22500: 2km, 5776: 4km, 121: 30km, 256:16x16 (20km); 961: 10km resoltn
-# if radar: then n_obs is for each observation height level
-oeinf = 4.**.5
-
-vis = dict(var_name='VIS 0.6µm', unit='[1]',
- kind='MSG_4_SEVIRI_BDRF', sat_channel=1,
- n_obs=961, obs_locations='square_array_evenly_on_grid',
- # n_obs=1, obs_locations=[(44.141, -0.99)],
- error_generate=0.03, error_assimilate=0.03*oeinf,
- loc_horiz_km=20,
- #height=4000, loc_vert_km=3
- )
-
-wv73 = dict(var_name='Brightness temperature WV 7.3µm', unit='[K]',
- kind='MSG_4_SEVIRI_TB', sat_channel=6,
- n_obs=961, obs_locations='square_array_evenly_on_grid',
- error_generate=1, error_assimilate=1*oeinf,
- loc_horiz_km=20,
- #height=7000, loc_vert_km=3
- )
-
-wv62 = dict(var_name='Brightness temperature WV 6.2µm', unit='[K]',
- kind='MSG_4_SEVIRI_TB', sat_channel=5,
- n_obs=961, obs_locations='square_array_evenly_on_grid',
- # n_obs=1, obs_locations=[(44.141, -0.99)],
- error_generate=1, error_assimilate=1*oeinf,
- loc_horiz_km=20,
- #height=10000, loc_vert_km=3
- )
-
-ir108 = dict(var_name='Brightness temperature IR 10.8µm', unit='[K]',
- kind='MSG_4_SEVIRI_TB', sat_channel=9,
- n_obs=1, obs_locations='square_array_evenly_on_grid',
- error_generate=5., error_assimilate=10.,
- loc_horiz_km=32)
-
-radar = dict(var_name='Radar reflectivity', unit='[dBz]',
- kind='RADAR_REFLECTIVITY',
- n_obs=961, obs_locations='square_array_evenly_on_grid',
- # n_obs=2, obs_locations=[(45.332, 0.4735), (45.332, 0.53)],
- heights=range(2000, 14001, 2000),
- error_generate=2.5, error_assimilate=2.5*oeinf,
- loc_horiz_km=20, loc_vert_km=3)
-
-t = dict(var_name='Temperature', unit='[K]',
- kind='RADIOSONDE_TEMPERATURE',
- #n_obs=22500, obs_locations='square_array_evenly_on_grid',
- n_obs=1, obs_locations=[(45., 0.)],
- error_generate=0.2, error_assimilate=0.2,
- heights=[1000,], #range(1000, 17001, 2000),
- loc_horiz_km=50, loc_vert_km=2.5)
-
-q = dict(var_name='Specific humidity', unit='[kg/kg]',
- kind='RADIOSONDE_SPECIFIC_HUMIDITY', n_obs=1,
- error_generate=0., error_assimilate=5*1e-5,
- heights=[1000], #range(1000, 17001, 2000),
- loc_horiz_km=50, loc_vert_km=2.5)
-
-t2m = dict(var_name='SYNOP Temperature', unit='[K]',
- kind='SYNOP_TEMPERATURE',
- n_obs=256, obs_locations='square_array_evenly_on_grid',
- error_generate=0.3, error_assimilate=0.3,
- loc_horiz_km=10, loc_vert_km=2)
-
-psfc = dict(var_name='SYNOP Pressure', unit='[Pa]',
- kind='SYNOP_SURFACE_PRESSURE', n_obs=1,
- error_generate=50., error_assimilate=100.,
- loc_horiz_km=32, loc_vert_km=5)
-
-exp.observations = [t, radar, ]
-
-# the variables which will be replaced in the WRF initial conditions file
-exp.update_vars = ['U', 'V', 'W', 'THM', 'PH', 'MU', 'QVAPOR', 'QCLOUD', 'QICE', 'PSFC']
-#exp.update_vars = ['U', 'V', 'W', 'THM', 'PH', 'MU', 'QVAPOR', 'PSFC']
-#exp.update_vars = ['U', 'V', 'W', 'THM', 'PH', 'MU', 'QVAPOR', 'QCLOUD', 'QICE', 'QRAIN', 'QSNOW', 'QGRAUP', 'PSFC']
-#exp.update_vars = ['QVAPOR', 'QCLOUD', 'QICE', 'PSFC']
-
-exp.use_existing_obsseq = False
-# exp.use_existing_obsseq='/jetfs/home/lkugler/data/sim_archive/exp_v1.22_P2_rr_WV73_obs10_loc20_oe1/obs_seq_out/%Y-%m-%d_%H:%M_obs_seq.out'
-
-
-
-
diff --git a/dartwrf/namelist_handler.py b/dartwrf/namelist_handler.py
new file mode 100644
index 0000000..351f887
--- /dev/null
+++ b/dartwrf/namelist_handler.py
@@ -0,0 +1,88 @@
+import warnings
+import os
+
+# author: gthalassinos
+
+class WRF_namelist():
+ """Class to read and write WRF namelist files
+
+ Example:
+ n = WRF_namelist('/path_to_existing/namelist.input')
+ n.read()
+
+ n.namelist # is a dictionary
+ n.namelist['time_control']['run_days'] = 1
+
+ n.write('/path_to_new/namelist.input')
+ """
+ def __init__(self):
+ self.namelist = {}
+
+ def read(self, fname: str) -> None:
+ """Read a WRF namelist file and store it in a dictionary
+ """
+ with open(fname, 'r') as f:
+ lines = f.readlines()
+
+ subdomains = True
+ block_name = None
+ for l in lines:
+ l = l.strip()
+ if l.startswith("&"):
+ block_name = l.split("&")[1]
+ self.namelist[block_name] = {}
+ elif l.startswith("/"):
+ block_name = None
+ elif l.startswith("!") and block_name:
+ continue
+ elif (block_name) and ("=" in l):
+ variable, value = l.split("=")
+ variable = variable.strip()
+ value = value.strip().rstrip(",")
+
+ if "," in value:
+ try:
+ values = eval(value)
+ except:
+ raise ValueError(f"{variable} is not a tuple!")
+ subdomains = True
+ value = values
+ # check if we have single numbers
+ if not isinstance(value, tuple):
+ try:
+ value = int(value)
+ except ValueError:
+ try:
+ value = float(value)
+ except ValueError:
+ pass
+
+ self.namelist[
+ block_name][
+ variable] = value
+
+ subdomains = False
+
+ pass
+
+ def write(self, fname: str) -> None:
+ """Write a WRF namelist file
+ """
+ if os.path.exists(fname):
+ warnings.warn(f"{fname} already exists!")
+ if input('Continue? (Y/n) ') in ['Y', 'y']:
+ pass
+ else:
+ raise FileExistsError
+
+ with open(fname, 'w') as file:
+ for block, variables in self.namelist.items():
+ file.write(f" &{block}\n")
+ for variable, value in variables.items():
+ if isinstance(value, str) and not value.startswith('.'):
+ value = f'{value}'
+ if isinstance(value, tuple):
+ value = str(value)[1:-1]
+ file.write(f" {variable:<35} = {value},\n")
+ file.write(" /\n\n")
+ pass
\ No newline at end of file
diff --git a/dartwrf/obs/__init__.py b/dartwrf/obs/__init__.py
index e69de29..5c23b21 100644
--- a/dartwrf/obs/__init__.py
+++ b/dartwrf/obs/__init__.py
@@ -0,0 +1,28 @@
+
+def obskind_read(dart_srcdir: str) -> dict:
+ """Read dictionary of observation types + ID numbers ("kind")
+ from DART f90 script and return it as python dictionary
+ """
+ definitionfile = dart_srcdir + \
+ '/../../../assimilation_code/modules/observations/obs_kind_mod.f90'
+ with open(definitionfile, 'r') as f:
+ kind_def_f = f.readlines()
+
+ obskind_nrs = {}
+ for i, line in enumerate(kind_def_f):
+ if 'Integer definitions for DART OBS TYPES' in line:
+ # data starts below this line
+ i_start = i
+ break
+ for line in kind_def_f[i_start+1:]:
+ if 'MAX_DEFINED_TYPES_OF_OBS' in line:
+ # end of data
+ break
+ if '::' in line:
+ # a line looks like this
+ # integer, parameter, public :: MSG_4_SEVIRI_TB = 261
+ data = line.split('::')[-1].split('=')
+ kind_str = data[0].strip()
+ kind_nr = int(data[1].strip())
+ obskind_nrs[kind_str] = kind_nr
+ return obskind_nrs
\ No newline at end of file
diff --git a/dartwrf/obs/obskind.py b/dartwrf/obs/obskind.py
deleted file mode 100644
index 8011b5a..0000000
--- a/dartwrf/obs/obskind.py
+++ /dev/null
@@ -1,338 +0,0 @@
-""" NOTE: This file is autogenerated!
-Use dartwrf/create_obskind_table.py to regenerate!
-"""
-obs_kind_nrs = {
-"RADIOSONDE_U_WIND_COMPONENT": 1,
-"RADIOSONDE_V_WIND_COMPONENT": 2,
-"RADIOSONDE_GEOPOTENTIAL_HGT": 3,
-"RADIOSONDE_SURFACE_PRESSURE": 4,
-"RADIOSONDE_TEMPERATURE": 5,
-"RADIOSONDE_SPECIFIC_HUMIDITY": 6,
-"DROPSONDE_U_WIND_COMPONENT": 7,
-"DROPSONDE_V_WIND_COMPONENT": 8,
-"DROPSONDE_SURFACE_PRESSURE": 9,
-"DROPSONDE_TEMPERATURE": 10,
-"DROPSONDE_SPECIFIC_HUMIDITY": 11,
-"AIRCRAFT_U_WIND_COMPONENT": 12,
-"AIRCRAFT_V_WIND_COMPONENT": 13,
-"AIRCRAFT_TEMPERATURE": 14,
-"AIRCRAFT_SPECIFIC_HUMIDITY": 15,
-"ACARS_U_WIND_COMPONENT": 16,
-"ACARS_V_WIND_COMPONENT": 17,
-"ACARS_TEMPERATURE": 18,
-"ACARS_SPECIFIC_HUMIDITY": 19,
-"MARINE_SFC_U_WIND_COMPONENT": 20,
-"MARINE_SFC_V_WIND_COMPONENT": 21,
-"MARINE_SFC_TEMPERATURE": 22,
-"MARINE_SFC_SPECIFIC_HUMIDITY": 23,
-"MARINE_SFC_PRESSURE": 24,
-"LAND_SFC_U_WIND_COMPONENT": 25,
-"LAND_SFC_V_WIND_COMPONENT": 26,
-"LAND_SFC_TEMPERATURE": 27,
-"LAND_SFC_SPECIFIC_HUMIDITY": 28,
-"LAND_SFC_PRESSURE": 29,
-"SAT_U_WIND_COMPONENT": 30,
-"SAT_V_WIND_COMPONENT": 31,
-"ATOV_TEMPERATURE": 32,
-"AIRS_TEMPERATURE": 33,
-"AIRS_SPECIFIC_HUMIDITY": 34,
-"GPS_PRECIPITABLE_WATER": 35,
-"VADWND_U_WIND_COMPONENT": 36,
-"VADWND_V_WIND_COMPONENT": 37,
-"CIMMS_AMV_U_WIND_COMPONENT": 38,
-"CIMMS_AMV_V_WIND_COMPONENT": 39,
-"DOPPLER_RADIAL_VELOCITY": 40,
-"RADAR_REFLECTIVITY": 41,
-"RADAR_CLEARAIR_REFLECTIVITY": 42,
-"PRECIPITATION_FALL_SPEED": 43,
-"METAR_U_10_METER_WIND": 44,
-"METAR_V_10_METER_WIND": 45,
-"METAR_TEMPERATURE_2_METER": 46,
-"METAR_SPECIFIC_HUMIDITY_2_METER": 47,
-"METAR_SURFACE_PRESSURE": 48,
-"METAR_POT_TEMP_2_METER": 49,
-"DEWPOINT": 50,
-"DEWPOINT_2_METER": 51,
-"BUOY_DEWPOINT": 52,
-"SHIP_DEWPOINT": 53,
-"SYNOP_DEWPOINT": 54,
-"AIREP_DEWPOINT": 55,
-"AMDAR_DEWPOINT": 56,
-"PILOT_DEWPOINT": 57,
-"BOGUS_DEWPOINT": 58,
-"AIRS_DEWPOINT": 59,
-"METAR_DEWPOINT_2_METER": 60,
-"RADIOSONDE_DEWPOINT": 61,
-"DROPSONDE_DEWPOINT": 62,
-"AIRCRAFT_DEWPOINT": 63,
-"ACARS_DEWPOINT": 64,
-"MARINE_SFC_DEWPOINT": 65,
-"LAND_SFC_DEWPOINT": 66,
-"RADIOSONDE_RELATIVE_HUMIDITY": 67,
-"DROPSONDE_RELATIVE_HUMIDITY": 68,
-"AIRCRAFT_RELATIVE_HUMIDITY": 69,
-"ACARS_RELATIVE_HUMIDITY": 70,
-"MARINE_SFC_RELATIVE_HUMIDITY": 71,
-"LAND_SFC_RELATIVE_HUMIDITY": 72,
-"METAR_RELATIVE_HUMIDITY_2_METER": 73,
-"AIRS_RELATIVE_HUMIDITY": 74,
-"MESONET_RELATIVE_HUMIDITY": 75,
-"RADIOSONDE_SURFACE_ALTIMETER": 76,
-"DROPSONDE_SURFACE_ALTIMETER": 77,
-"MARINE_SFC_ALTIMETER": 78,
-"LAND_SFC_ALTIMETER": 79,
-"METAR_ALTIMETER": 80,
-"MESONET_SURFACE_ALTIMETER": 81,
-"TEMPERATURE": 82,
-"SPECIFIC_HUMIDITY": 83,
-"PRESSURE": 84,
-"GPSRO_REFRACTIVITY": 85,
-"VORTEX_LAT": 86,
-"VORTEX_LON": 87,
-"VORTEX_PMIN": 88,
-"VORTEX_WMAX": 89,
-"BUOY_U_WIND_COMPONENT": 90,
-"BUOY_V_WIND_COMPONENT": 91,
-"BUOY_SURFACE_PRESSURE": 92,
-"BUOY_TEMPERATURE": 93,
-"SHIP_U_WIND_COMPONENT": 94,
-"SHIP_V_WIND_COMPONENT": 95,
-"SHIP_SURFACE_PRESSURE": 96,
-"SHIP_TEMPERATURE": 97,
-"SYNOP_U_WIND_COMPONENT": 98,
-"SYNOP_V_WIND_COMPONENT": 99,
-"SYNOP_SURFACE_PRESSURE": 100,
-"SYNOP_SPECIFIC_HUMIDITY": 101,
-"SYNOP_TEMPERATURE": 102,
-"AIREP_U_WIND_COMPONENT": 103,
-"AIREP_V_WIND_COMPONENT": 104,
-"AIREP_PRESSURE": 105,
-"AIREP_TEMPERATURE": 106,
-"AMDAR_U_WIND_COMPONENT": 107,
-"AMDAR_V_WIND_COMPONENT": 108,
-"AMDAR_PRESSURE": 109,
-"AMDAR_TEMPERATURE": 110,
-"PILOT_U_WIND_COMPONENT": 111,
-"PILOT_V_WIND_COMPONENT": 112,
-"PILOT_PRESSURE": 113,
-"PILOT_TEMPERATURE": 114,
-"BOGUS_U_WIND_COMPONENT": 115,
-"BOGUS_V_WIND_COMPONENT": 116,
-"BOGUS_PRESSURE": 117,
-"BOGUS_TEMPERATURE": 118,
-"PROFILER_U_WIND_COMPONENT": 119,
-"PROFILER_V_WIND_COMPONENT": 120,
-"PROFILER_PRESSURE": 121,
-"SATEM_THICKNESS": 122,
-"NOAA_1_VTPR1_RADIANCE": 123,
-"NOAA_2_VTPR1_RADIANCE": 124,
-"NOAA_3_VTPR1_RADIANCE": 125,
-"NOAA_4_VTPR1_RADIANCE": 126,
-"NOAA_5_HIRS_RADIANCE": 127,
-"NOAA_5_MSU_TB": 128,
-"NOAA_5_AVHRR_RADIANCE": 129,
-"NOAA_6_HIRS_RADIANCE": 130,
-"NOAA_6_MSU_TB": 131,
-"NOAA_6_AVHRR_RADIANCE": 132,
-"NOAA_7_HIRS_RADIANCE": 133,
-"NOAA_7_MSU_TB": 134,
-"NOAA_7_AVHRR_RADIANCE": 135,
-"NOAA_8_HIRS_RADIANCE": 136,
-"NOAA_8_MSU_TB": 137,
-"NOAA_8_AVHRR_RADIANCE": 138,
-"NOAA_9_HIRS_RADIANCE": 139,
-"NOAA_9_MSU_TB": 140,
-"NOAA_9_AVHRR_RADIANCE": 141,
-"NOAA_10_HIRS_RADIANCE": 142,
-"NOAA_10_MSU_TB": 143,
-"NOAA_10_AVHRR_RADIANCE": 144,
-"NOAA_11_HIRS_RADIANCE": 145,
-"NOAA_11_MSU_TB": 146,
-"NOAA_11_AVHRR_RADIANCE": 147,
-"NOAA_12_HIRS_RADIANCE": 148,
-"NOAA_12_MSU_TB": 149,
-"NOAA_12_AVHRR_RADIANCE": 150,
-"NOAA_13_AVHRR_RADIANCE": 151,
-"NOAA_14_HIRS_RADIANCE": 152,
-"NOAA_14_MSU_TB": 153,
-"NOAA_14_AVHRR_RADIANCE": 154,
-"NOAA_15_HIRS_RADIANCE": 155,
-"NOAA_15_AMSUA_TB": 156,
-"NOAA_15_AMSUB_TB": 157,
-"NOAA_15_AVHRR_RADIANCE": 158,
-"NOAA_16_HIRS_RADIANCE": 159,
-"NOAA_16_AMSUA_TB": 160,
-"NOAA_16_AMSUB_TB": 161,
-"NOAA_16_AVHRR_RADIANCE": 162,
-"NOAA_17_HIRS_RADIANCE": 163,
-"NOAA_17_AMSUA_TB": 164,
-"NOAA_17_AMSUB_TB": 165,
-"NOAA_17_AVHRR_RADIANCE": 166,
-"NOAA_18_HIRS_RADIANCE": 167,
-"NOAA_18_AMSUA_TB": 168,
-"NOAA_18_AVHRR_RADIANCE": 169,
-"NOAA_18_MHS_TB": 170,
-"NOAA_19_HIRS_RADIANCE": 171,
-"NOAA_19_AMSUA_TB": 172,
-"NOAA_19_AVHRR_RADIANCE": 173,
-"NOAA_19_MHS_TB": 174,
-"NOAA_20_ATMS_TB": 175,
-"NOAA_20_VIIRS_RADIANCE": 176,
-"DMSP_8_SSMI_TB": 177,
-"DMSP_9_SSMI_TB": 178,
-"DMSP_10_SSMI_TB": 179,
-"DMSP_11_SSMI_TB": 180,
-"DMSP_11_SSMT2_TB": 181,
-"DMSP_12_SSMI_TB": 182,
-"DMSP_12_SSMT2_TB": 183,
-"DMSP_13_SSMI_TB": 184,
-"DMSP_14_SSMI_TB": 185,
-"DMSP_14_SSMT2_TB": 186,
-"DMSP_15_SSMI_TB": 187,
-"DMSP_15_SSMT2_TB": 188,
-"DMSP_16_SSMIS_TB": 189,
-"DMSP_17_SSMIS_TB": 190,
-"DMSP_18_SSMIS_TB": 191,
-"DMSP_19_SSMIS_TB": 192,
-"METEOSAT_1_MVIRI_RADIANCE": 193,
-"METEOSAT_2_MVIRI_RADIANCE": 194,
-"METEOSAT_3_MVIRI_RADIANCE": 195,
-"METEOSAT_4_MVIRI_RADIANCE": 196,
-"METEOSAT_5_MVIRI_RADIANCE": 197,
-"METEOSAT_6_MVIRI_RADIANCE": 198,
-"METEOSAT_7_MVIRI_RADIANCE": 199,
-"GOES_4_SOUNDER_RADIANCE": 200,
-"GOES_5_SOUNDER_RADIANCE": 201,
-"GOES_6_SOUNDER_RADIANCE": 202,
-"GOES_7_SOUNDER_RADIANCE": 203,
-"GOES_8_IMAGER_RADIANCE": 204,
-"GOES_8_SOUNDER_RADIANCE": 205,
-"GOES_9_IMAGER_RADIANCE": 206,
-"GOES_9_SOUNDER_RADIANCE": 207,
-"GOES_10_IMAGER_RADIANCE": 208,
-"GOES_10_SOUNDER_RADIANCE": 209,
-"GOES_11_IMAGER_RADIANCE": 210,
-"GOES_11_SOUNDER_RADIANCE": 211,
-"GOES_12_IMAGER_RADIANCE": 212,
-"GOES_12_SOUNDER_RADIANCE": 213,
-"GOES_13_IMAGER_RADIANCE": 214,
-"GOES_13_SOUNDER_RADIANCE": 215,
-"GOES_14_IMAGER_RADIANCE": 216,
-"GOES_14_SOUNDER_RADIANCE": 217,
-"GOES_15_IMAGER_RADIANCE": 218,
-"GOES_15_SOUNDER_RADIANCE": 219,
-"GOES_16_ABI_RADIANCE": 220,
-"GOES_17_ABI_RADIANCE": 221,
-"GOES_18_ABI_RADIANCE": 222,
-"GOES_19_ABI_RADIANCE": 223,
-"GMS_1_IMAGER_RADIANCE": 224,
-"GMS_2_IMAGER_RADIANCE": 225,
-"GMS_3_IMAGER_RADIANCE": 226,
-"GMS_4_IMAGER_RADIANCE": 227,
-"GMS_5_IMAGER_RADIANCE": 228,
-"FY2_2_VISSR_RADIANCE": 229,
-"FY2_3_VISSR_RADIANCE": 230,
-"FY2_4_VISSR_RADIANCE": 231,
-"FY2_5_VISSR_RADIANCE": 232,
-"FY2_7_VISSR_RADIANCE": 233,
-"TRMM_1_TMI_TB": 234,
-"ERS_1_ATSR_RADIANCE": 235,
-"ERS_1_MWR_TB": 236,
-"ERS_2_ATSR_RADIANCE": 237,
-"ERS_2_MWR_TB": 238,
-"EOS_1_MODIS_RADIANCE": 239,
-"EOS_1_ASTER_RADIANCE": 240,
-"EOS_2_AMSUA_TB": 241,
-"EOS_2_AIRS_RADIANCE": 242,
-"EOS_2_HSB_TB": 243,
-"EOS_2_MODIS_RADIANCE": 244,
-"EOS_2_AMSRE_TB": 245,
-"METOP_1_HIRS_RADIANCE": 246,
-"METOP_1_AMSUA_TB": 247,
-"METOP_1_AVHRR_RADIANCE": 248,
-"METOP_1_MHS_TB": 249,
-"METOP_2_HIRS_RADIANCE": 250,
-"METOP_2_AMSUA_TB": 251,
-"METOP_2_AVHRR_RADIANCE": 252,
-"METOP_2_MHS_TB": 253,
-"METOP_3_AVHRR_RADIANCE": 254,
-"ENVISAT_1_ATSR_RADIANCE": 255,
-"ENVISAT_1_MWR_TB": 256,
-"MSG_1_SEVIRI_RADIANCE": 257,
-"MSG_2_SEVIRI_RADIANCE": 258,
-"MSG_3_SEVIRI_RADIANCE": 259,
-"MSG_4_SEVIRI_RADIANCE": 260,
-"MSG_4_SEVIRI_TB": 261,
-"MSG_4_SEVIRI_BDRF": 262,
-"FY1_3_MVISR_RADIANCE": 263,
-"FY1_4_MVISR_RADIANCE": 264,
-"MTSAT_1_IMAGER_RADIANCE": 265,
-"MTSAT_2_IMAGER_RADIANCE": 266,
-"CORIOLIS_1_WINDSAT_TB": 267,
-"JPSS_0_ATMS_TB": 268,
-"JPSS_0_VIIRS_RADIANCE": 269,
-"SENTINEL3_1_SLSTR_RADIANCE": 270,
-"SENTINEL3_2_SLSTR_RADIANCE": 271,
-"MEGHATR_1_SAPHIR_TB": 272,
-"MEGHATR_1_MADRAS_TB": 273,
-"FY3_1_MWTS_TB": 274,
-"FY3_1_MWHS_TB": 275,
-"FY3_1_IRAS_RADIANCE": 276,
-"FY3_1_MWRI_TB": 277,
-"FY3_2_MWTS_TB": 278,
-"FY3_2_MWHS_TB": 279,
-"FY3_2_MWRI_TB": 280,
-"FY3_3_MWRI_TB": 281,
-"FY3_3_MWTS2_TB": 282,
-"FY3_3_MWHS2_TB": 283,
-"FY3_3_MERSI1_RADIANCE": 284,
-"FY3_4_MWRI_TB": 285,
-"FY3_4_MWTS2_TB": 286,
-"FY3_4_MWHS2_TB": 287,
-"FY3_4_MERSI2_RADIANCE": 288,
-"COMS_1_MI_RADIANCE": 289,
-"METEOR_M_1_MSUMR_RADIANCE": 290,
-"METEOR_M_2_MSUMR_RADIANCE": 291,
-"METEOR_M_2_MTVZAGY_TB": 292,
-"CALIPSO_1_IIR_RADIANCE": 293,
-"GCOM_W_1_AMSR2_TB": 294,
-"NIMBUS_3_MRIR_RADIANCE": 295,
-"NIMBUS_4_THIR_RADIANCE": 296,
-"NIMBUS_5_THIR_RADIANCE": 297,
-"NIMBUS_6_HIRS_RADIANCE": 298,
-"NIMBUS_6_SCAMS_TB": 299,
-"NIMBUS_6_THIR_RADIANCE": 300,
-"NIMBUS_7_SMMR_TB": 301,
-"NIMBUS_7_THIR_RADIANCE": 302,
-"HIMAWARI_8_AHI_RADIANCE": 303,
-"HIMAWARI_9_AHI_RADIANCE": 304,
-"MTG_1_FCI_RADIANCE": 305,
-"SARAL_1_ALTIKA_TB": 306,
-"METOPSG_1_ICI_TB": 307,
-"METOPSG_1_METIMAGE_RADIANCE": 308,
-"METOPSG_1_MWS_TB": 309,
-"METOPSG_1_MWI_TB": 310,
-"LANDSAT_4_TM_RADIANCE": 311,
-"LANDSAT_5_TM_RADIANCE": 312,
-"LANDSAT_7_TM_RADIANCE": 313,
-"LANDSAT_8_TIRS_RADIANCE": 314,
-"JASON_2_AMR_TB": 315,
-"GPM_1_GMI_TB": 316,
-"GPM_1_DPR_TB": 317,
-"INSAT3_4_IMAGER_RADIANCE": 318,
-"INSAT3_4_SOUNDER_RADIANCE": 319,
-"INSAT3_5_IMAGER_RADIANCE": 320,
-"INSAT3_5_SOUNDER_RADIANCE": 321,
-"TICFIRE_1_MBFIRI_RADIANCE": 322,
-"ISS_1_ECOSTRES_RADIANCE": 323,
-"HJ1_2_IRMSS_RADIANCE": 324,
-"GKOMPSAT2_1_AMI_RADIANCE": 325,
-"GCOM_C_1_SGLI_RADIANCE": 326,
-"SMOS_1_MIRAS_TB": 327,
-"ORS_6_COWVR_TB": 328,
-"FY4_1_AGRI_RADIANCE": 329,
-"TROPICS_0_TROPICS_TB": 330,
-"GF5_1_VIMS_RADIANCE": 331,
-"HY2_1_MWRI_TB": 332,
-"CLOUDSAT_1_CPR_TB": 333,
-}
\ No newline at end of file
diff --git a/dartwrf/obs/obsseq.py b/dartwrf/obs/obsseq.py
index 7f0541f..7f35bd7 100755
--- a/dartwrf/obs/obsseq.py
+++ b/dartwrf/obs/obsseq.py
@@ -87,7 +87,7 @@ class ObsRecord(pd.DataFrame):
"""Retrieve H(x_prior) for all ensemble members
Returns:
- np.array (n_obs, n_ens)
+ np.array (n_obs, ensemble_size)
"""
return self._get_model_Hx('prior')
@@ -95,7 +95,7 @@ class ObsRecord(pd.DataFrame):
"""Retrieve H(x_posterior) for all ensemble members
Returns:
- np.array (n_obs, n_ens)
+ np.array (n_obs, ensemble_size)
"""
return self._get_model_Hx('posterior')
diff --git a/dartwrf/obs/run_obs_diag.py b/dartwrf/obs/run_obs_diag.py
index fa1e78e..2546438 100644
--- a/dartwrf/obs/run_obs_diag.py
+++ b/dartwrf/obs/run_obs_diag.py
@@ -9,7 +9,7 @@ rundir_program = '/home/fs71386/lkugler/data/run_DART/'
def prepare(obserr_iszero='.true.'):
copy(cluster.scriptsdir+'/../templates/input.eval.nml',
rundir_program+'/input.nml')
- sed_inplace(rundir_program+'/input.nml', '<n_ens>', str(int(exp.n_ens)))
+ sed_inplace(rundir_program+'/input.nml', '<n_ens>', str(int(exp.ensemble_size)))
sed_inplace(rundir_program+'/input.nml', '<zero_error_obs>', obserr_iszero)
sed_inplace(rundir_program+'/input.nml', '<horiz_dist_only>', '.false.') # dummy
sed_inplace(rundir_program+'/input.nml', '<vert_norm_hgt>', '5000.0') # dummy
diff --git a/dartwrf/prep_IC_prior.py b/dartwrf/prep_IC_prior.py
index b91c823..5c84794 100755
--- a/dartwrf/prep_IC_prior.py
+++ b/dartwrf/prep_IC_prior.py
@@ -23,7 +23,7 @@ def create_wrfrst_in_WRF_rundir(time: dt.datetime, prior_init_time: dt.datetime,
"""Copy WRF restart files to run_WRF directory
These files will be used as initial conditions for the next WRF run
"""
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
clean_wrfdir(cluster.wrf_rundir(iens))
prior_wrfrst = prior_path_exp + prior_init_time.strftime('/%Y-%m-%d_%H:%M/') \
@@ -48,7 +48,7 @@ def create_updated_wrfinput_from_wrfout(time: dt.datetime, prior_init_time: dt.d
"""
print('writing updated wrfout to WRF run directory as wrfinput')
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
prior_wrfout = prior_path_exp + prior_init_time.strftime('/%Y-%m-%d_%H:%M/') \
+str(iens)+time.strftime('/wrfout_d01_%Y-%m-%d_%H:%M:%S')
new_start_wrfinput = cluster.wrf_rundir(iens) + '/wrfinput_d01'
diff --git a/dartwrf/prepare_namelist.py b/dartwrf/prepare_namelist.py
index cffbef8..025a9b5 100755
--- a/dartwrf/prepare_namelist.py
+++ b/dartwrf/prepare_namelist.py
@@ -1,85 +1,34 @@
"""Create namelist.input files
Usage:
-prepare_namelist.py <begin> <end> <intv> [--radt=<minutes>] [--restart=<flag>] [--restart_interval=<minutes>]
+prepare_namelist.py <config> <begin> <end> <intv> [--radt=<minutes>] [--restart=<flag>] [--restart_interval=<minutes>]
Options:
--radt=<minutes> Radiation interval [default: 5]
--restart=<flag> Restart flag (.true., .false.) [default: .false.]
--restart_interval=<minutes> Restart frequency [default: 720]
"""
-import os, sys, shutil, warnings
+import os, sys
import datetime as dt
from docopt import docopt
-from dartwrf.exp_config import exp
+from dartwrf.namelist_handler import WRF_namelist
from dartwrf.server_config import cluster
-from dartwrf.utils import sed_inplace, copy, symlink, mkdir
-
-def run(iens, begin, end, hist_interval_s=5*60, radt=5, archive=True,
- restart=False, restart_interval=720):
- """Create a namelist.input file for each ensemble member
-
- Args:
- archive (bool): if True, write to archivedir of experiment
- if False, write to WRF run directory
- restart (str): fortran bool whether to use wrfinput or wrfrst
- restart_interval (int): output frequency of wrfrst (minutes)
-
- Returns
- None
- """
- rundir = cluster.wrf_rundir(iens)
- copy(cluster.namelist, rundir+'/namelist.input')
-
- sed_inplace(rundir+'/namelist.input', '<dx>', str(int(exp.model_dx)))
- #sed_inplace(rundir+'/namelist.input', '<timestep>', str(int(exp.timestep)))
- sed_inplace(rundir+'/namelist.input', '<hist_interval_s>', str(int(hist_interval_s)))
-
- sed_inplace(rundir+'/namelist.input', '<radt>', str(int(radt)))
-
- rst_flag = '.true.' if restart else '.false.'
- sed_inplace(rundir+'/namelist.input', '<restart>', rst_flag)
- sed_inplace(rundir+'/namelist.input', '<restart_interval>', str(int(float(restart_interval))))
-
- if archive:
- archdir = cluster.archivedir+begin.strftime('/%Y-%m-%d_%H:%M/'+str(iens)+'/')
- os.makedirs(archdir, exist_ok=True)
- else:
- archdir = './'
- sed_inplace(rundir+'/namelist.input', '<archivedir>', archdir)
-
- # set times
- for k, v in {'<y1>': '%Y', '<m1>': '%m', '<d1>': '%d',
- '<HH1>': '%H', '<MM1>': '%M', '<SS1>': '%S'}.items():
- sed_inplace(rundir+'/namelist.input', k, begin.strftime(v))
- for k, v in {'<y2>': '%Y', '<m2>': '%m', '<d2>': '%d',
- '<HH2>': '%H', '<MM2>': '%M', '<SS2>': '%S'}.items():
- sed_inplace(rundir+'/namelist.input', k, end.strftime(v))
-
- print('saved', rundir+'/namelist.input')
-
-
- if archive:
- init_dir = cluster.archivedir+begin.strftime('/%Y-%m-%d_%H:%M/')+str(iens)
- os.makedirs(init_dir, exist_ok=True)
- copy(rundir+'/namelist.input', init_dir+'/namelist.input')
- print('archived', init_dir+'/namelist.input')
-
- if not restart:
- wrfin_old = rundir+'/wrfinput_d01'
- wrfin_arch = init_dir+'/wrfinput_d01'
- copy(wrfin_old, wrfin_arch)
- print('archived', wrfin_arch)
+from dartwrf.utils import sed_inplace, copy, read_dict_from_pyfile
if __name__ == '__main__':
-
+
args = docopt(__doc__)
+ archive = True
+
+ f_config = args['<config>']
+ exp = read_dict_from_pyfile(f_config)
+
begin = dt.datetime.strptime(args['<begin>'], '%Y-%m-%d_%H:%M:%S')
end = dt.datetime.strptime(args['<end>'], '%Y-%m-%d_%H:%M:%S')
- intv = int(args['<intv>'])
-
+ hist_interval_s = int(args['<intv>'])
+
radt = int(args['--radt'])
if not radt:
radt = '5'
@@ -87,13 +36,58 @@ if __name__ == '__main__':
restart = False
if args['--restart'] == '.true.':
restart = True
+ rst_flag = '.true.' if restart else '.false.'
restart_interval = args['--restart_interval']
if not restart_interval:
restart_interval = 720
+
+ # replace these variables in the namelist
+ replace_dict = {
+ 'time_control': {
+ # time start
+ '<y1>': begin.strftime('%Y'),
+ '<m1>': begin.strftime('%m'),
+ '<d1>': begin.strftime('%d'),
+ '<HH1>': begin.strftime('%H'),
+ '<MM1>': begin.strftime('%M'),
+ '<SS1>': begin.strftime('%S'),
+
+ # time end
+ '<y2>': end.strftime('%Y'),
+ '<m2>': end.strftime('%m'),
+ '<d2>': end.strftime('%d'),
+ '<HH2>': end.strftime('%H'),
+ '<MM2>': end.strftime('%M'),
+ '<SS2>': end.strftime('%S'),
+
+ # other variables
+ '<dx>': str(int(exp.model_dx)),
+ '<hist_interval_s>': str(int(hist_interval_s)),
+ '<radt>': str(int(radt)),
+ '<restart>': rst_flag,
+ '<restart_interval>': str(int(float(restart_interval))),
+ },
+ }
+
+ print('prepare namelists for all ens members',radt,restart,restart_interval)
+ for iens in range(1, exp.ensemble_size+1):
- print('prepare namelists for all ens members',intv,radt,restart,restart_interval)
- for iens in range(1, exp.n_ens+1):
+ nml = WRF_namelist()
+ nml.read(cluster.WRF_namelist_template)
+
+ # replace parameters
+ for section, section_dict in replace_dict.items():
+ for key, value in section_dict.items():
+ nml.namelist[section][key] = value
+
+ f_out = cluster.wrf_rundir_base +'/'+ exp.name + '/'+str(iens)+'/namelist.input'
+ nml.write(f_out)
+ print('saved', f_out)
- run(iens, begin, end, hist_interval_s=intv, radt=radt,
- restart=restart, restart_interval=restart_interval)
+
+ if archive:
+ archdir = cluster.archive_base+'/'+exp.name+begin.strftime('/%Y-%m-%d_%H:%M/'+str(iens)+'/')
+ os.makedirs(archdir, exist_ok=True)
+ else:
+ archdir = './'
\ No newline at end of file
diff --git a/dartwrf/prepare_wrfrundir.py b/dartwrf/prepare_wrfrundir.py
index cd85e64..081cec2 100755
--- a/dartwrf/prepare_wrfrundir.py
+++ b/dartwrf/prepare_wrfrundir.py
@@ -12,16 +12,18 @@ Returns:
"""
import os, sys, shutil
import datetime as dt
-
-from dartwrf.exp_config import exp
from dartwrf.server_config import cluster
-from dartwrf.utils import symlink, link_contents, try_remove
+from dartwrf.utils import symlink, link_contents, try_remove, read_dict_from_pyfile
from dartwrf import prepare_namelist
if __name__ == '__main__':
+
+ time = sys.argv[1]
+ exp = read_dict_from_pyfile(sys.argv[2])
+
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
rundir = cluster.wrf_rundir(iens)
os.makedirs(rundir, exist_ok=True)
link_contents(cluster.srcdir, rundir)
@@ -32,8 +34,9 @@ if __name__ == '__main__':
# prepare input profiles
try_remove(rundir+'/input_sounding') # remove existing file
+
if hasattr(exp, 'input_profile'):
- init_time = dt.datetime.strptime(sys.argv[1], '%Y-%m-%d_%H:%M')
+ init_time = dt.datetime.strptime(time, '%Y-%m-%d_%H:%M')
# prep namelist for ./ideal.exe
prepare_namelist.run(iens, begin=init_time, end=init_time, archive=False) # time not important
diff --git a/dartwrf/run_ens.jet.sh b/dartwrf/run_ens.jet.sh
deleted file mode 100755
index de21234..0000000
--- a/dartwrf/run_ens.jet.sh
+++ /dev/null
@@ -1,25 +0,0 @@
-<cluster.wrf_modules>
-export SLURM_STEP_GRES=none
-
-echo "SLURM_ARRAY_TASK_ID:"$SLURM_ARRAY_TASK_ID
-EXPNAME=<exp.expname>
-MAINDIR=<cluster.wrf_rundir_base>
-
-IENS=$SLURM_ARRAY_TASK_ID
-RUNDIR=$MAINDIR/$EXPNAME/$IENS
-echo "ENSEMBLE NR: "$IENS" in "$RUNDIR
-cd $RUNDIR
-rm -rf rsl.out.0*
-echo "mpirun -np <exp.np_WRF> ./wrf.exe"
-mpirun -np <exp.np_WRF> ./wrf.exe
-
-
-# error checking
-line=`tail -n 2 rsl.out.0000`
-if [[ $line == *"SUCCESS COMPLETE WRF"* ]];
-then
- echo $RUNDIR 'SUCCESS COMPLETE WRF'
-else
- echo $RUNDIR $line
- exit 1
-fi
diff --git a/dartwrf/run_ens.vsc.gfortran.sh b/dartwrf/run_ens.vsc.gfortran.sh
deleted file mode 100755
index 48dab4c..0000000
--- a/dartwrf/run_ens.vsc.gfortran.sh
+++ /dev/null
@@ -1,42 +0,0 @@
-
-module purge
-#module load intel-mpi/2019.6 intel/19.1.0 netcdf-fortran/4.4.5-intel-19.0.5.281-qye4cqn zlib/1.2.11-intel-19.1.0.166-hs6m2qh hdf5/1.10.5-intel-19.0.5.281-qyzojtm netcdf/4.7.0-intel-19.0.5.281-75t52g6
-module load gcc/12.2.0-gcc-8.5.0-aal4zp2 openmpi/4.1.4-gcc-12.2.0-khtxitv
-export SLURM_STEP_GRES=none
-
-## $SLURM_ARRAY_TASK_ID
-echo "SLURM_ARRAY_TASK_ID:"$SLURM_ARRAY_TASK_ID
-EXPNAME=<exp.expname>
-
-MAINDIR=<cluster.wrf_rundir_base>
-#pinning=(0-11 12-23 24-35 36-47)
-pinning=(0,1,2,3,4,5,6,7,8,9,10,11 12,13,14,15,16,17,18,19,20,21,22,23 24,25,26,27,28,29,30,31,32,33,34,35 36,37,38,39,40,41,42,43,44,45,46,47)
-
-for ((n=1; n<=4; n++))
-do
- IENS="$(((($SLURM_ARRAY_TASK_ID - 1)* 4) + $n))"
- RUNDIR=$MAINDIR/$EXPNAME/$IENS
- echo "ENSEMBLE NR: "$IENS" in "$RUNDIR
- cd $RUNDIR
- rm -rf rsl.out.0*
- echo "mpirun -np 12 --cpu-set ${pinning[$n-1]} /home/fs71386/lkugler/run_container.sh python.gcc9.5.0.vsc4.sif ./wrf.exe"
- mpirun -np 12 --cpu-set ${pinning[$n-1]} /home/fs71386/lkugler/run_container.sh python.gcc9.5.0.vsc4.sif ./wrf.exe &
- cd ../
-done
-wait
-
-# error checking
-for ((n=1; n<=4; n++))
-do
- IENS="$(((($SLURM_ARRAY_TASK_ID - 1)* 4) + $n))"
- RUNDIR=$MAINDIR/$EXPNAME/$IENS
- cd $RUNDIR
- line=`tail -n 1 rsl.out.0000`
- if [[ $line == *"SUCCESS COMPLETE WRF"* ]];
- then
- echo $RUNDIR 'SUCCESS COMPLETE WRF'
- else
- echo $RUNDIR $line
- exit 1
- fi
-done
diff --git a/dartwrf/run_ens.vsc.mem_0096.sh b/dartwrf/run_ens.vsc.mem_0096.sh
deleted file mode 100755
index 40904ff..0000000
--- a/dartwrf/run_ens.vsc.mem_0096.sh
+++ /dev/null
@@ -1,43 +0,0 @@
-
-module purge
-module load intel-mpi/2019.6 intel/19.1.0 netcdf-fortran/4.4.5-intel-19.0.5.281-qye4cqn zlib/1.2.11-intel-19.1.0.166-hs6m2qh hdf5/1.10.5-intel-19.0.5.281-qyzojtm netcdf/4.7.0-intel-19.0.5.281-75t52g6
-export I_MPI_DEBUG=4
-export OMP_NUM_THREADS=1
-mem_per_task=10G
-cpu_per_task=12
-
-echo "SLURM_ARRAY_TASK_ID:"$SLURM_ARRAY_TASK_ID
-EXPNAME=<exp.expname>
-
-MAINDIR=<cluster.wrf_rundir_base>
-pinning=(0-11 12-23 24-35 36-47)
-
-mytasks=4
-for n in `seq 1 $mytasks`
-do
- IENS="$(((($SLURM_ARRAY_TASK_ID - 1)* 4) + $n))" # ensemble number (5,6,7,8 for job array element 2)
- RUNDIR=$MAINDIR/$EXPNAME/$IENS
- echo "ENSEMBLE NR: "$IENS" in "$RUNDIR
- cd $RUNDIR
- rm -rf rsl.out.0*
- echo "srun --mem=$mem_per_task --cpus-per-task=$cpu_per_task --cpu_bind=map_cpu:${pinning[$n-1]} --ntasks=1 ./wrf.exe &"
- srun --mem=$mem_per_task --cpus-per-task=$cpu_per_task --cpu_bind=map_cpu:${pinning[$n-1]} --ntasks=1 ./wrf.exe &
- cd ../
-done
-wait
-
-# error checking
-for ((n=1; n<=4; n++))
-do
- IENS="$(((($SLURM_ARRAY_TASK_ID - 1)* 4) + $n))"
- RUNDIR=$MAINDIR/$EXPNAME/$IENS
- cd $RUNDIR
- line=`tail -n 1 rsl.out.0000`
- if [[ $line == *"SUCCESS COMPLETE WRF"* ]];
- then
- echo $RUNDIR 'SUCCESS COMPLETE WRF'
- else
- echo $RUNDIR $line
- exit 1
- fi
-done
diff --git a/dartwrf/run_ens.vsc.sh b/dartwrf/run_ens.vsc.sh
deleted file mode 100755
index 8c2dc1e..0000000
--- a/dartwrf/run_ens.vsc.sh
+++ /dev/null
@@ -1,40 +0,0 @@
-
-module purge
-module load intel-mpi/2019.6 intel/19.1.0 netcdf-fortran/4.4.5-intel-19.0.5.281-qye4cqn zlib/1.2.11-intel-19.1.0.166-hs6m2qh hdf5/1.10.5-intel-19.0.5.281-qyzojtm netcdf/4.7.0-intel-19.0.5.281-75t52g6
-export SLURM_STEP_GRES=none
-
-## $SLURM_ARRAY_TASK_ID
-echo "SLURM_ARRAY_TASK_ID:"$SLURM_ARRAY_TASK_ID
-EXPNAME=<exp.expname>
-
-MAINDIR=<cluster.wrf_rundir_base>
-pinning=(0-11 12-23 24-35 36-47)
-
-for ((n=1; n<=4; n++))
-do
- IENS="$(((($SLURM_ARRAY_TASK_ID - 1)* 4) + $n))"
- RUNDIR=$MAINDIR/$EXPNAME/$IENS
- echo "ENSEMBLE NR: "$IENS" in "$RUNDIR
- cd $RUNDIR
- rm -rf rsl.out.0*
- echo "mpirun -genv I_MPI_PIN_PROCESSOR_LIST="${pinning[$n-1]}" -np 12 ./wrf.exe"
- mpirun -genv I_MPI_PIN_PROCESSOR_LIST=${pinning[$n-1]} -np 12 ./wrf.exe &
- cd ../
-done
-wait
-
-# error checking
-for ((n=1; n<=4; n++))
-do
- IENS="$(((($SLURM_ARRAY_TASK_ID - 1)* 4) + $n))"
- RUNDIR=$MAINDIR/$EXPNAME/$IENS
- cd $RUNDIR
- line=`tail -n 1 rsl.out.0000`
- if [[ $line == *"SUCCESS COMPLETE WRF"* ]];
- then
- echo $RUNDIR 'SUCCESS COMPLETE WRF'
- else
- echo $RUNDIR $line
- exit 1
- fi
-done
diff --git a/dartwrf/run_ens.vsc.single.sh b/dartwrf/run_ens.vsc.single.sh
deleted file mode 100755
index 6997f07..0000000
--- a/dartwrf/run_ens.vsc.single.sh
+++ /dev/null
@@ -1,15 +0,0 @@
-
-module purge
-module load intel-mpi/2019.6 intel/19.1.0 netcdf-fortran/4.4.5-intel-19.0.5.281-qye4cqn zlib/1.2.11-intel-19.1.0.166-hs6m2qh hdf5/1.10.5-intel-19.0.5.281-qyzojtm netcdf/4.7.0-intel-19.0.5.281-75t52g6
-
-## $SLURM_ARRAY_TASK_ID
-echo "SLURM_ARRAY_TASK_ID:"$SLURM_ARRAY_TASK_ID
-EXPNAME=<exp.expname>
-
-MAINDIR=<cluster.wrf_rundir_base>
-IENS=$SLURM_ARRAY_TASK_ID
-RUNDIR=$MAINDIR/$EXPNAME/$IENS
-echo "ENSEMBLE NR: "$IENS" in "$RUNDIR
-cd $RUNDIR
-rm -rf rsl.out.0*
-mpirun -genv I_MPI_PIN_PROCESSOR_LIST=0-19 -np 20 ./wrf.exe
diff --git a/dartwrf/server_config.py b/dartwrf/server_config.py
index ebb475d..c787988 100755
--- a/dartwrf/server_config.py
+++ b/dartwrf/server_config.py
@@ -1,55 +1,56 @@
"""Cluster configuration file, see docstring of ClusterConfig class in dartwrf/utils.py for details"""
-from dartwrf import utils
-from dartwrf.exp_config import exp
-
-cluster = utils.ClusterConfig(exp)
-cluster.name = 'jet'
-cluster.max_nproc = 20
-cluster.use_slurm = True
-cluster.size_WRF_jobarray = exp.n_ens
-cluster.np_WRF = 16
-
-# binaries
-cluster.python = '/jetfs/home/lkugler/miniforge3/envs/verif/bin/python'
-cluster.ncks = '/jetfs/spack/opt/spack/linux-rhel8-skylake_avx512/intel-2021.7.1/nco-5.1.0-izrhxv24jqco5epjhf5ledsqwanojc5m/bin/ncks'
-cluster.ideal = '/jetfs/home/lkugler/data/compile/bin/ideal-v4.6.0_20250210_StS.exe'
-cluster.wrfexe = '/jetfs/home/lkugler/data/compile/bin/wrf-v4.6.0_20250210_StS.exe'
-cluster.dart_modules = 'module purge; module load rttov/v13.2-gcc-8.5.0'
-cluster.wrf_modules = """module purge; module load netcdf-fortran/4.5.3-intel-2021.7.1-27ldrnt"""
-
-# paths for data output
-cluster.wrf_rundir_base = '/jetfs/home/lkugler/data/run_WRF/' # path for temporary files
-cluster.dart_rundir_base = '/jetfs/home/lkugler/data/run_DART/' # path for temporary files
-cluster.archive_base = '/jetfs/home/lkugler/data/sim_archive/'
-
-# paths used as input
-cluster.srcdir = '/jetfs/home/lkugler/data/compile/WRF-4.3/run'
-cluster.dart_srcdir = '/jetfs/home/lkugler/data/compile/DART/DART-10.8.3_10pct/models/wrf/work/'
-cluster.rttov_srcdir = '/jetfs/home/lkugler/data/compile/RTTOV13/rtcoef_rttov13/'
-cluster.dartwrf_dir = '/jetfs/home/lkugler/DART-WRF/'
-
-# other inputs
-cluster.geo_em_nature = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.250m_1600x1600'
-cluster.geo_em_forecast = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.2km_200x200'
-#cluster.obs_impact_filename = cluster.dartwrf_dir+'/templates/impactfactor_T.txt'
-cluster.namelist = cluster.dartwrf_dir+'/config/templates/namelist.input_nat_exact'
-cluster.rttov_nml = cluster.dartwrf_dir + "/config/templates/obs_def_rttov.VIS+WV.nml"
-cluster.run_WRF = '/jetfs/home/lkugler/DART-WRF/dartwrf/run_ens.jet.sh'
-
-cluster.slurm_cfg = {"account": "lkugler", "partition": "all",
- "ntasks": "1", "ntasks-per-core": "1", "mem": "30G",
- "mail-type": "FAIL", "mail-user": "lukas.kugler@univie.ac.at"}
-
-# WRF file format, will only change if WRF changes
-cluster.wrfout_format = '/wrfout_d01_%Y-%m-%d_%H:%M:%S'
-
-# pattern for the init_time folder in sim_archive
-cluster.pattern_init_time = "/%Y-%m-%d_%H:%M/"
-
-# how an obs_seq.out file is archived
-cluster.pattern_obs_seq_out = cluster.archivedir + \
- "/diagnostics/%Y-%m-%d_%H:%M_obs_seq.out"
+from re import A
+from dartwrf.utils import ClusterConfig
+
+cluster = ClusterConfig(
+ max_nproc = 20,
+ max_nproc_for_each_ensemble_member = 16,
+ use_slurm = True,
+
+ # binaries
+ python = '/jetfs/home/lkugler/miniforge3/envs/verif/bin/python',
+ ncks = '/jetfs/spack/opt/spack/linux-rhel8-skylake_avx512/intel-2021.7.1/nco-5.1.0-izrhxv24jqco5epjhf5ledsqwanojc5m/bin/ncks',
+ ideal = '/jetfs/home/lkugler/data/compile/bin/ideal-v4.6.0_20250210_StS.exe',
+ wrfexe = '/jetfs/home/lkugler/data/compile/bin/wrf-v4.6.0_20250210_StS.exe',
+ dart_modules = 'module purge; module load rttov/v13.2-gcc-8.5.0',
+ wrf_modules = """module purge; module load netcdf-fortran/4.5.3-intel-2021.7.1-27ldrnt""",
+
+ # paths for data output
+ wrf_rundir_base = '/jetfs/home/lkugler/data/run_WRF/', # path for temporary files
+ dart_rundir_base = '/jetfs/home/lkugler/data/run_DART/', # path for temporary files
+ archive_base = '/jetfs/home/lkugler/data/sim_archive/',
+
+ # paths used as input
+ srcdir = '/jetfs/home/lkugler/data/compile/WRF-4.3/run',
+ dart_srcdir = '/jetfs/home/lkugler/data/compile/DART/DART-10.8.3_10pct/models/wrf/work/',
+ rttov_srcdir = '/jetfs/home/lkugler/data/compile/RTTOV13/rtcoef_rttov13/',
-# how an obs_seq.final file is archived
-cluster.pattern_obs_seq_final = cluster.archivedir + \
- "/diagnostics/%Y-%m-%d_%H:%M_obs_seq.final"
+ dartwrf_dir_dev = '/jetfs/home/lkugler/DART-WRF/',
+ namelist = '/jetfs/home/lkugler/DART-WRF/config/templates/namelist.input_nat_exact',
+ rttov_nml = "/jetfs/home/lkugler/DART-WRF/config/templates/obs_def_rttov.VIS+WV.nml",
+
+ # other inputs
+ geo_em_nature = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.250m_1600x1600',
+ geo_em_forecast = '/jetfs/home/lkugler/data/sim_archive/geo_em.d01.nc.2km_200x200',
+ #obs_impact_filename = dartwrf_dir_dev+'/templates/impactfactor_T.txt'
+
+ WRF_exe_template = '/jetfs/home/lkugler/DART-WRF/config/templates/run_WRF.jet.sh',
+ WRF_ideal_template = '/jetfs/home/lkugler/DART-WRF/config/templates/run_WRF_ideal.sh',
+
+ slurm_cfg = {"account": "lkugler", "partition": "all",
+ "ntasks": "1", "ntasks-per-core": "1", "mem": "30G",
+ "mail-type": "FAIL", "mail-user": "lukas.kugler@univie.ac.at"},
+
+ # WRF file format, will only change if WRF changes
+ wrfout_format = '/wrfout_d01_%Y-%m-%d_%H:%M:%S',
+
+ # pattern for the init_time folder in sim_archive
+ pattern_init_time = "/%Y-%m-%d_%H:%M/",
+
+ # how an obs_seq.out file is archived
+ pattern_obs_seq_out = "<archivedir>/diagnostics/%Y-%m-%d_%H:%M_obs_seq.out",
+
+ # how an obs_seq.final file is archived
+ pattern_obs_seq_final = "<archivedir>/diagnostics/%Y-%m-%d_%H:%M_obs_seq.final",
+
+ )
\ No newline at end of file
diff --git a/dartwrf/update_IC.py b/dartwrf/update_IC.py
index 4c66a47..883f76a 100755
--- a/dartwrf/update_IC.py
+++ b/dartwrf/update_IC.py
@@ -24,7 +24,7 @@ def update_initials_in_WRF_rundir(time):
update_vars = ['Times',]
update_vars.extend(exp.update_vars)
- for iens in range(1, exp.n_ens+1):
+ for iens in range(1, exp.ensemble_size+1):
ic_file = cluster.wrf_rundir(iens) + time.strftime(initials_fmt)
if not os.path.isfile(ic_file):
raise IOError(ic_file+' does not exist, updating impossible!')
diff --git a/dartwrf/utils.py b/dartwrf/utils.py
index 1205441..fad027b 100755
--- a/dartwrf/utils.py
+++ b/dartwrf/utils.py
@@ -15,56 +15,20 @@ import datetime as dt
import re
import tempfile
import pickle
+import importlib.util
+userhome = os.path.expanduser('~')
-class Experiment(object):
- """Collection of variables which define the experiment
-
- Attributes:
- expname (str): Name of the experiment
- model_dx (int): WRF grid spacing in meters
- n_ens (int): Ensemble size
-
- nature_wrfout_pattern (str): Path to the nature run, where we take observations from;
- the path can contain wildcards (*,?), e.g. '/jetfs/exp1/*/1/wrfout_d01_%Y-%m-%d_%H:%M:%S'
- input_profile (str): Path to WRF idealized input profiles;
- e.g. '/data/initial_profiles/wrf/ens/raso.fc.<iens>.wrfprof';
- <iens> is replaced by 001-040 for a 40-member ensemble
-
- update_vars (list of str): Variables which will be updated after assimilation (update_IC.py)
- e.g. ['U', 'V', 'W', 'THM', 'PH', 'MU', 'QVAPOR',]
-
- observations (list of dict): Dictionaries which define an observation;
- keys:
- `error_generate`: measurement error standard-deviation;
- `error_assimilate`: assigned observation error std-dev;
- `heights`: list of integers at which observations are taken;
- `loc_horiz_km`: float of horizontal localization half-width in km;
- `loc_vert_km`: float of vertical localization half-width in km;
-
- use_existing_obsseq (str, False): Path to existing obs_seq.out file (False: generate new one);
- time string is replaced by actual time: /path/%Y-%m-%d_%H:%M_obs_seq.out
-
- dart_nml (dict): updates to the default input.nml of DART (in dart_srcdir)
- keys are namelist section headers (e.g. &filter_nml)
- values are dictionaries of parameters and values (e.g. dict(ens_size=exp.n_ens,))
-
- """
-
- def __init__(self):
- pass
-
+def import_from_path(module_name, file_path):
+ spec = importlib.util.spec_from_file_location(module_name, file_path)
+ module = importlib.util.module_from_spec(spec) # type: ignore
+ sys.modules[module_name] = module
+ spec.loader.exec_module(module) # type: ignore
+ return module
class ClusterConfig(object):
"""Collection of variables regarding the cluster configuration
- Configuration name docs
-
- When coding, use configuration settings like this:
- $ from exp_config import exp
- $ from cluster_config import cluster
- $ path = cluster.archivedir
-
Attributes:
name (str): Name of the cluster
max_nproc (int): Maximum number of processors that can be used
@@ -81,10 +45,6 @@ class ClusterConfig(object):
dart_modules (str): Command to load modules before running DART
wrf_modules (str): Command to load modules before running WRF
- wrf_rundir_base (str): Path to temporary files for WRF
- dart_rundir_base (str): Path to temporary files for DART
- archive_base (str): Path to long-time output storage
-
srcdir (str): Path to where WRF has been compiled, including the 'run' folder of WRF, e.g. /home/WRF-4.3/run
dart_srcdir (str): Path to DART compile directory, e.g. /home/DART-9.11.9/models/wrf/work
rttov_srcdir (str): Path to RTTOV compile directory, e.g. /home/RTTOV13/rtcoef_rttov13/
@@ -103,48 +63,46 @@ class ClusterConfig(object):
"""
- def __init__(self, exp):
- self.exp = exp # to access the experiment config in here
-
+ def __init__(self,
+ max_nproc: int,
+ max_nproc_for_each_ensemble_member: int,
+ WRF_ideal_template: str,
+ WRF_exe_template: str,
+ archive_base: str,
+ wrf_rundir_base: str,
+ dart_rundir_base: str,
+ dartwrf_dir_dev: str,
+ WRF_namelist_template: str,
+ **kwargs):
# defaults
+ # these are overwritten with choices in **kwargs
self.dart_modules = ''
self.wrf_modules = ''
self.size_jobarray = '1'
-
- @property
- def archivedir(self):
- """Path to the directory where data for the experiment is stored
-
- Example:
- `/users/abcd/data/sim_archive/experiment1/`
- """
- return self.archive_base+'/'+self.exp.expname+'/'
-
- @property
- def scripts_rundir(self):
- """Path to the directory where the DART-WRF scripts are executed
-
- Note:
- If you want to execute scripts from the folder where you develop code, use `self.dartwrf_dir` (not sure if this works)
- If you want to execute the code from a different place ('research'), then use `self.archivedir+'/DART-WRF/'`
-
- Example:
- `/user/data/sim_archive/DART-WRF/dartwrf/`
- """
- return self.archivedir+'/DART-WRF/dartwrf/'
-
- @property
- def dart_rundir(self):
- """Path to the directory where DART programs will run
- Includes the experiment name
- """
- return self.dart_rundir_base+'/'+self.exp.expname+'/'
-
- def wrf_rundir(self, iens):
- """Path to the directory where an ensemble member will run WRF
- Includes the experiment name and the ensemble member index
- """
- return self.wrf_rundir_base+'/'+self.exp.expname+'/'+str(iens)+'/'
+ self.use_slurm = False
+ self.slurm_cfg = {}
+ self.log_dir = './'
+ self.slurm_scripts_dir = './'
+ self.archive_base = archive_base
+ self.wrf_rundir_base = wrf_rundir_base
+ self.dart_rundir_base = dart_rundir_base
+ self.dartwrf_dir_dev = dartwrf_dir_dev
+ self.WRF_namelist_template = WRF_namelist_template
+ self.python = 'python'
+ self.pattern_obs_seq_out = '<archivedir>/diagnostics/%Y-%m-%d_%H:%M_obs_seq.out'
+ self.pattern_obs_seq_final = '<archivedir>/diagnostics/%Y-%m-%d_%H:%M_obs_seq.final'
+
+ self.max_nproc = max_nproc
+ self.max_nproc_for_each_ensemble_member = max_nproc_for_each_ensemble_member
+ self.WRF_ideal_template = WRF_ideal_template
+ self.WRF_exe_template = WRF_exe_template
+
+ # user defined
+ for key, value in kwargs.items():
+ setattr(self, key, value)
+
+ def __str__(self):
+ return f'ClusterConfig: {self.__dict__}'
def run_job(self, cmd, jobname='', cfg_update=dict(), depends_on=None):
"""Run scripts in a shell
@@ -174,7 +132,100 @@ class ClusterConfig(object):
raise Exception('Error running command >>> '+cmd)
-userhome = os.path.expanduser('~')
+class Config(object):
+ """Collection of variables which define the experiment
+
+ Attributes:
+ expname (str): Name of the experiment
+ model_dx (int): WRF grid spacing in meters
+ ensemble_size (int): Ensemble size
+
+ nature_wrfout_pattern (str): Path to the nature run WRF files, where can be generated from;
+ the path can contain wildcards (*,?), e.g. '/jetfs/exp1/*/1/wrfout_d01_%Y-%m-%d_%H:%M:%S'
+
+ input_profile (str): Path to sounding profiles as initial condition (see WRF ideal guide)
+ e.g. '/data/initial_profiles/wrf/ens/raso.fc.<iens>.wrfprof';
+ <iens> is replaced by 001-040 for a 40-member ensemble
+
+ update_vars (list of str): Variables which will be updated after assimilation (update_IC.py)
+ e.g. ['U', 'V', 'W', 'THM', 'PH', 'MU', 'QVAPOR',]
+
+ observations (list of dict): Dictionaries which define an observation;
+ keys:
+ `error_generate`: measurement error standard-deviation;
+ `error_assimilate`: assigned observation error std-dev;
+ `heights`: list of integers at which observations are taken;
+ `loc_horiz_km`: float of horizontal localization half-width in km;
+ `loc_vert_km`: float of vertical localization half-width in km;
+
+ use_existing_obsseq (str, False): Path to existing obs_seq.out file (False: generate new one);
+ time string is replaced by actual time: /path/%Y-%m-%d_%H:%M_obs_seq.out
+
+ dart_nml (dict): updates to the default input.nml of DART (in dart_srcdir)
+ keys are namelist section headers (e.g. &filter_nml)
+ values are dictionaries of parameters and values (e.g. dict(ens_size=exp.ensemble_size,))
+
+ wrf_rundir_base (str): Path to temporary files for WRF
+ dart_rundir_base (str): Path to temporary files for DART
+ archive_base (str): Path to long-time output storage
+
+ """
+
+ def __init__(self, name: str, model_dx: int, ensemble_size: int,
+ update_vars: list=[], dart_nml: dict={},
+ use_existing_obsseq: bool | str = False,
+ input_profile: bool | str = False,
+ nature_wrfout_pattern: bool | str = False,
+ **kwargs):
+
+ # defining the compulsory variables
+ self.name = name
+ self.model_dx = model_dx
+ self.ensemble_size = ensemble_size
+ self.update_vars = update_vars
+ self.dart_nml = dart_nml
+
+ # optional
+ self.use_existing_obsseq = use_existing_obsseq
+ self.input_profile = input_profile
+ self.nature_wrfout_pattern = nature_wrfout_pattern
+
+ if not update_vars:
+ warnings.warn('No `update_vars` defined, not updating any variables after assimilation!')
+
+ if not dart_nml:
+ warnings.warn('No `dart_nml` defined, using default DART namelist!')
+
+ if not isinstance(use_existing_obsseq, str):
+ if use_existing_obsseq != False:
+ raise ValueError('`use_existing_obsseq` must be a string or False, but is', use_existing_obsseq)
+
+ if isinstance(use_existing_obsseq, str):
+ print('Using existing observation sequence', use_existing_obsseq)
+
+ # user defined
+ for key, value in kwargs.items():
+ setattr(self, key, value)
+
+
+
+def write_dict_to_pyfile(d: dict, filename: str):
+ """Write a dictionary to a python file"""
+ os.makedirs(os.path.dirname(filename), exist_ok=True)
+ with open(filename, 'w') as f:
+ for key, value in d.items():
+ f.write(key+' = '+str(value)+'\n')
+
+def read_dict_from_pyfile(filename: str) -> Config:
+ """Read a dictionary from a python file,
+ return as Config object
+ """
+ with open(filename, 'r') as f:
+ d = {}
+ for line in f:
+ key, value = line.split('=')
+ d[key.strip()] = value.strip()
+ return Config(**d)
def shell(args, pythonpath=None):
diff --git a/dartwrf/workflows.py b/dartwrf/workflows.py
index a90dc61..f4f1b97 100644
--- a/dartwrf/workflows.py
+++ b/dartwrf/workflows.py
@@ -10,12 +10,15 @@ import sys
import shutil
import warnings
import datetime as dt
+import random
+import string
-from dartwrf.utils import script_to_str, shell
+from dartwrf.utils import script_to_str, shell, import_from_path, read_dict_from_pyfile, write_dict_to_pyfile
+from dartwrf.utils import Config
class WorkFlows(object):
- def __init__(self, exp_config='cfg.py', server_config='server.py'):
+ def __init__(self, expname: str, server_config: str='server_config.py'):
"""Set up the experiment folder in `archivedir`.
1. Copy the selected config files
@@ -34,91 +37,105 @@ class WorkFlows(object):
exp (obj): experiment configuration as defined in exp_config file
"""
- def _copy_dartwrf_to_archive(dirs_exist_ok=False):
- # Copy DART-WRF/dartwrf/ to self.cluster.archivedir folder
- # copy the dartwrf python package
- shutil.copytree(self.cluster.dartwrf_dir+'/dartwrf/',
- self.cluster.archivedir+'/DART-WRF/dartwrf/',
- ignore=shutil.ignore_patterns('*.git',),
- dirs_exist_ok=dirs_exist_ok)
- print('>>> DART-WRF scripts: "' +
- self.cluster.archivedir+'/DART-WRF/"')
-
- # copy this script for reproducibility
- script_executing_this_process = sys.argv[0]
- shutil.copy(script_executing_this_process,
- self.cluster.archivedir+'/DART-WRF/')
-
- def _save_config_to(config_fname, destination):
+ def _save_config_to(src: str, dst: str):
try:
- shutil.copyfile('config/'+config_fname, destination)
+ shutil.copyfile(src, dst)
except shutil.SameFileError:
pass
print('------------------------------------------------------')
- print('>>> Starting experiment ... ')
- print('>>> Experiment configuration: "./config/'+exp_config+'" ')
- print('>>> Server configuration: "./config/'+server_config+'"')
-
- # 1
- # copy the selected config files (arguments to Workflows(...)) to the scripts directory
- # ./DART-WRF/dartwrf/server_config.py and ./DART-WRF/dartwrf/exp_config.py
- # these config files will be used later, and no others!
- # usually /home/DART-WRF/dartwrf/
- original_scripts_dir = '/'.join(__file__.split('/')[:-1])
- _save_config_to(server_config, original_scripts_dir +
- '/server_config.py')
- _save_config_to(exp_config, original_scripts_dir+'/exp_config.py')
-
- # 2
- # import the configuration files from where we copied them just before
- sys.path.append(original_scripts_dir)
- from server_config import cluster
- self.cluster = cluster
- from exp_config import exp
- self.exp = exp
-
- print(" ")
- print('>>> Main data folder: "'+self.cluster.archivedir+'"')
- print('>>> Temporary DART folder: "'+self.cluster.dart_rundir+'"')
- print('>>> Temporary WRF folder: "' +
- self.cluster.wrf_rundir_base+'"')
-
- # 3
- # Set paths and backup scripts
- self.cluster.log_dir = self.cluster.archivedir+'/logs/'
+ self.expname = expname
+ print('>>> Experiment name: "'+self.expname+'"')
+ print('>>> Server configuration: "'+server_config+'"')
+
+ self.cluster = import_from_path('config', server_config).cluster
+
+ # some helpful variables
+ self.archivedir = self.cluster.archive_base+'/'+self.expname+'/'
+ dart_rundir = self.cluster.dart_rundir_base+'/'+self.expname+'/'
+ self.scripts_rundir = self.archivedir+'/DART-WRF/dartwrf/'
+ pattern_obs_seq_out = self.cluster.pattern_obs_seq_out.replace(
+ '<archivedir>', self.archivedir)
+ pattern_obs_seq_final = self.cluster.pattern_obs_seq_final.replace(
+ '<archivedir>', self.archivedir)
+
+ # collect all variables to put into the config file
+ self.configure_defaults = {
+ 'expname': self.expname,
+ 'archivedir': self.archivedir,
+ 'dart_rundir': dart_rundir,
+ 'scripts_rundir': self.scripts_rundir,
+ 'pattern_obs_seq_out': pattern_obs_seq_out,
+ 'pattern_obs_seq_final': pattern_obs_seq_final,
+ }
+
+ self.f_cfg_base = self.archivedir + '/DART-WRF/configs/'
+ self.f_cfg_current = None
+
+ ############### ARCHIVE SCRIPTS AND CONFIGS
+ # Copy scripts and config files to `self.archivedir` folder
+ dirs_exist_ok = False
+ if os.path.exists(self.archivedir):
+ if input('The experiment name already exists! Overwrite existing experiment? (Y/n) ') in ['Y', 'y']:
+ dirs_exist_ok = True
+
+ shutil.copytree(self.cluster.dartwrf_dir_dev,
+ self.archivedir+'/DART-WRF/',
+ ignore=shutil.ignore_patterns('*.git','config/','__*','tests/'),
+ dirs_exist_ok=dirs_exist_ok)
+
+ # copy cluster config to /DART-WRF/dartwrf/
+ _save_config_to(server_config, self.scripts_rundir + '/server_config.py')
+
print(" ")
- print('>>> Log-files: "'+self.cluster.log_dir+'"')
+ print('>>> Running experiment in "'+self.archivedir+'"')
+ self.cluster.log_dir = self.archivedir+'/logs/'
+
if self.cluster.use_slurm:
- self.cluster.slurm_scripts_dir = self.cluster.archivedir+'/slurm-scripts/'
- print('>>> SLURM scripts: "'+self.cluster.slurm_scripts_dir+'"')
+ self.cluster.slurm_scripts_dir = self.archivedir+'/slurm-scripts/'
+
print(" ")
-
- # 4
- # to be able to generate obs_seq.in files, we need a dictionary to convert obs kinds to numbers
- # a) we read the obs kind definitions (obs_kind_mod.f90 from DART code)
- # b) we generate a python file with this dictionary
- import create_obskind_table
- create_obskind_table.run(server_config)
-
- # 5
- # Copy scripts and config files to `self.cluster.archivedir` folder
- try:
- _copy_dartwrf_to_archive()
- except FileExistsError as e:
- if input('The experiment name already exists! Overwrite existing experiment? (Y/n) ') in ['Y', 'y']:
- _copy_dartwrf_to_archive(dirs_exist_ok=True)
- else:
- raise e
+ print('>>> DART will run in "'+dart_rundir+'"')
+ print('>>> WRF will run in "'+self.cluster.wrf_rundir_base+'/'+self.expname+'"')
# 6
# we set the path from where python should import dartwrf modules
- # every python command then imports DART-WRF from self.cluster.archivedir+'/DART-WRF/dartwrf/'
- self.cluster.python = 'export PYTHONPATH=' + \
- self.cluster.scripts_rundir+'/../; '+self.cluster.python
+ # every python command then imports DART-WRF from self.archivedir+'/DART-WRF/dartwrf/'
+ self.cluster.python = 'export PYTHONPATH=' + \
+ self.scripts_rundir+'/../; '+self.cluster.python
print('>>> DART-WRF experiment initialized. ')
print('------------------------------------------------------')
+ def configure(self, **kwargs):
+ """Update the config in Experiment.cfg
+ because we cant just forward arguments to command line calls,
+ we write a config file in a specified directory
+ """
+ # is there an already existing config?
+ if self.f_cfg_current is None:
+ # there is no config, we write a new one
+ cfg = Config(name=self.expname,
+ **self.configure_defaults,
+ **kwargs)
+ self.cfg = cfg # this will be accessed by the module functions below
+
+ else:
+ # there is already a config, we update it
+
+ # read existing config
+ cfg = read_dict_from_pyfile(self.f_cfg_current)
+
+ # set attributes in existing object
+ for key, value in kwargs.items():
+ setattr(cfg, key, value)
+
+ # finally, write cfg to file
+ # generate random string for filename
+ random_str = ''.join(random.choices(string.ascii_lowercase + string.digits, k=4))
+ self.f_cfg_current = self.f_cfg_base+'/cfg_'+random_str+'.py'
+ write_dict_to_pyfile(cfg.__dict__, self.f_cfg_current)
+
+
def prepare_WRFrundir(self, init_time):
"""Prepare WRF run directories for all ensemble members
@@ -132,7 +149,7 @@ class WorkFlows(object):
Returns:
None
"""
- cmd = 'python '+self.cluster.scripts_rundir + \
+ cmd = self.cluster.python+' '+self.scripts_rundir + \
'/prepare_wrfrundir.py '+init_time.strftime('%Y-%m-%d_%H:%M')
shell(cmd)
@@ -147,46 +164,13 @@ class WorkFlows(object):
"""
times_str = ','.join([t.strftime('%Y-%m-%d_%H:%M') for t in times])
- cmd = self.cluster.python+' '+self.cluster.scripts_rundir + \
+ cmd = self.cluster.python+' '+self.scripts_rundir + \
'/obs/create_obsseq_out.py '+times_str
- id = self.cluster.run_job(cmd, "obsgen-"+self.exp.expname,
+ id = self.cluster.run_job(cmd, "obsgen-"+self.expname,
cfg_update={"ntasks": "20", "time": "30", "mem": "200G", "ntasks-per-node": "20"}, depends_on=[depends_on])
return id
- def run_ideal(self, depends_on=None):
- """Run WRF's ideal.exe for every ensemble member
-
- Args:
- depends_on (str, optional): job ID of a previous job after which to run this job
-
- Returns:
- str: job ID of the submitted job
- """
- cmd = self.cluster.wrf_modules+"""
- export SLURM_STEP_GRES=none
- # run ideal.exe in parallel
- for ((n=1; n<="""+str(self.exp.n_ens)+"""; n++))
- do
- rundir="""+self.cluster.wrf_rundir_base+'/'+self.exp.expname+"""/$n
- echo $rundir
- cd $rundir
- mpirun -np 1 ./ideal.exe &
- done
- wait
-
- # move log file to sim_archive
- for ((n=1; n<="""+str(self.exp.n_ens)+"""; n++))
- do
- rundir="""+self.cluster.wrf_rundir_base+'/'+self.exp.expname+"""/$n
- touch -a $rundir/rsl.out.0000 # create log file if it doesnt exist, to avoid error in mv if it doesnt exist
- mv $rundir/rsl.out.0000 $rundir/rsl.out.input
- done
- """
- id = self.cluster.run_job(cmd, "ideal-"+self.exp.expname,
- cfg_update={"ntasks": "40", "ntasks-per-node": "40",
- "time": "30", "mem": "200G"}, depends_on=[depends_on])
- return id
def wrfinput_insert_wbubble(self, perturb=True, depends_on=None):
"""Inserts warm-bubble temperature perturbations into wrfinput files
@@ -204,13 +188,31 @@ class WorkFlows(object):
pstr = ' '
if perturb:
pstr = ' perturb'
- cmd = self.cluster.python+' '+self.cluster.scripts_rundir + \
+ cmd = self.cluster.python+' '+self.scripts_rundir + \
'/create_wbubble_wrfinput.py'+pstr
id = self.cluster.run_job(
- cmd, "ins_wbub-"+self.exp.expname, cfg_update={"time": "5"}, depends_on=[depends_on])
+ cmd, "ins_wbub-"+self.expname, cfg_update={"time": "5"}, depends_on=[depends_on])
return id
+ def run_ideal(self, depends_on=None):
+ """Run WRF's ideal.exe for every ensemble member
+
+ Args:
+ depends_on (str, optional): job ID of a previous job after which to run this job
+
+ Returns:
+ str: job ID of the submitted job
+ """
+ cmd = script_to_str(self.cluster.WRF_ideal_template
+ ).replace('<expname>', self.expname
+ ).replace('<wrf_rundir_base>', self.cluster.wrf_rundir_base
+ ).replace('<wrf_modules>', self.cluster.wrf_modules,
+ )
+ id = self.cluster.run_job(cmd, "ideal-"+self.expname,
+ cfg_update={"ntasks": "1", "time": "30", "mem": "200G"}, depends_on=[depends_on])
+ return id
+
def run_ENS(self, begin, end, first_second=False,
input_is_restart=True, output_restart_interval=360, hist_interval_s=300,
depends_on=None):
@@ -228,45 +230,55 @@ class WorkFlows(object):
Returns:
str: job ID of the submitted job
"""
-
- def prepare_WRF_inputfiles(begin, end, hist_interval_s=300, radt=1,
- output_restart_interval=False, depends_on=None):
-
- args = [self.cluster.python, self.cluster.scripts_rundir+'/prepare_namelist.py',
- begin.strftime(
- '%Y-%m-%d_%H:%M:%S'), end.strftime('%Y-%m-%d_%H:%M:%S'),
- str(hist_interval_s), '--radt='+str(radt), '--restart='+restart_flag,]
-
- if output_restart_interval != False:
+ def _prepare_WRF_inputfiles(begin, end, hist_interval_s=300, radt=1,
+ output_restart_interval: int | None = None,
+ depends_on=None):
+
+ args = [self.cluster.python, self.scripts_rundir+'/prepare_namelist.py',
+ self.f_cfg_current,
+ begin.strftime('%Y-%m-%d_%H:%M:%S'),
+ end.strftime('%Y-%m-%d_%H:%M:%S'),
+ str(hist_interval_s),
+ '--radt='+str(radt),
+ '--restart='+restart_flag,]
+
+ if output_restart_interval:
args.append('--restart_interval=' +
str(int(float(output_restart_interval))))
return self.cluster.run_job(' '.join(args), "preWRF",
cfg_update=dict(time="2"), depends_on=[depends_on])
+
+ ###########################################
+ # SLURM configuration for WRF
+ cfg_wrf = {"array": "1-"+str(self.cfg.ensemble_size),
+ "nodes": "1",
+ "ntasks": str(self.cluster.max_nproc_for_each_ensemble_member),
+ "ntasks-per-core": "1", "mem": "90G", }
id = depends_on
restart_flag = '.false.' if not input_is_restart else '.true.'
- wrf_cmd = script_to_str(self.cluster.run_WRF
- ).replace('<exp.expname>', self.exp.expname
- ).replace('<cluster.wrf_rundir_base>', self.cluster.wrf_rundir_base
- ).replace('<cluster.wrf_modules>', self.cluster.wrf_modules,
- ).replace('<exp.np_WRF>', str(self.cluster.np_WRF))
-
+
+ # command from template file
+ wrf_cmd = script_to_str(self.cluster.WRF_exe_template
+ ).replace('<expname>', self.expname
+ ).replace('<wrf_rundir_base>', self.cluster.wrf_rundir_base
+ ).replace('<wrf_modules>', self.cluster.wrf_modules,
+ ).replace('<WRF_number_of_processors>', "16")
+
+ # if 1-second forecast is required
if first_second:
- id = prepare_WRF_inputfiles(begin, begin+dt.timedelta(seconds=1),
+ id = _prepare_WRF_inputfiles(begin, begin+dt.timedelta(seconds=1),
hist_interval_s=1, # to get an output every 1 s
radt=0, # to get a cloud fraction CFRAC after 1 s
output_restart_interval=output_restart_interval,
depends_on=id)
- id = self.cluster.run_job(wrf_cmd, "WRF-"+self.exp.expname,
- cfg_update={"array": "1-"+str(self.cluster.size_WRF_jobarray),
- "nodes": "1", "ntasks": str(self.cluster.np_WRF), "ntasks-per-core": "1",
- "time": "5", "mem": "100G"},
- depends_on=[id])
+ id = self.cluster.run_job(wrf_cmd, "WRF-"+self.expname,
+ cfg_update=cfg_wrf, depends_on=[id])
# forecast for the whole forecast duration
- id = prepare_WRF_inputfiles(begin, end,
+ id = _prepare_WRF_inputfiles(begin, end,
hist_interval_s=hist_interval_s,
output_restart_interval=output_restart_interval,
depends_on=id)
@@ -274,16 +286,15 @@ class WorkFlows(object):
time_in_simulation_hours = (end-begin).total_seconds()/3600
runtime_wallclock_mins_expected = int(
time_in_simulation_hours*30 + 10) # usually <15 min/hour
- cfg_update = {"array": "1-"+str(self.cluster.size_WRF_jobarray),
- "nodes": "1", "ntasks": str(self.cluster.np_WRF), "ntasks-per-core": "1",
- "time": str(runtime_wallclock_mins_expected), "mem": "90G", }
+ cfg_wrf.update({"time": str(runtime_wallclock_mins_expected)})
+
if runtime_wallclock_mins_expected > 25:
- cfg_update.update({"partition": "amd"})
+ cfg_wrf.update({"partition": "amd"})
# #cfg_update.update({"exclude": "jet03"})
id = self.cluster.run_job(
- wrf_cmd, "WRF-"+self.exp.expname, cfg_update=cfg_update, depends_on=[id])
+ wrf_cmd, "WRF-"+self.expname, cfg_update=cfg_wrf, depends_on=[id])
return id
def assimilate(self, assim_time, prior_init_time, prior_valid_time, prior_path_exp,
@@ -293,7 +304,7 @@ class WorkFlows(object):
Args:
assim_time (dt.datetime): timestamp of prior wrfout files
prior_init_time (dt.datetime): timestamp to find the directory where the prior wrfout files are
- prior_path_exp (str): use this directory to get prior state (i.e. self.cluster.archivedir)
+ prior_path_exp (str): use this directory to get prior state (i.e. self.archivedir)
Returns:
str: job ID of the submitted job
@@ -301,13 +312,13 @@ class WorkFlows(object):
if not os.path.exists(prior_path_exp):
raise IOError('prior_path_exp does not exist: '+prior_path_exp)
- cmd = (self.cluster.python+' '+self.cluster.scripts_rundir+'/assimilate.py '
+ cmd = (self.cluster.python+' '+self.scripts_rundir+'/assimilate.py '
+ assim_time.strftime('%Y-%m-%d_%H:%M ')
+ prior_init_time.strftime('%Y-%m-%d_%H:%M ')
+ prior_valid_time.strftime('%Y-%m-%d_%H:%M ')
+ prior_path_exp)
- id = self.cluster.run_job(cmd, "Assim-"+self.exp.expname,
+ id = self.cluster.run_job(cmd, "Assim-"+self.expname,
cfg_update={"ntasks": "20", "time": "30", "mem": "110G",
"ntasks-per-node": "20", "ntasks-per-core": "1"}, depends_on=[depends_on])
return id
@@ -332,13 +343,13 @@ class WorkFlows(object):
else:
tnew = ''
- cmd = (self.cluster.python+' '+self.cluster.scripts_rundir+'/prep_IC_prior.py '
+ cmd = (self.cluster.python+' '+self.scripts_rundir+'/prep_IC_prior.py '
+ prior_path_exp
+ prior_init_time.strftime(' %Y-%m-%d_%H:%M')
+ prior_valid_time.strftime(' %Y-%m-%d_%H:%M')
+ tnew)
id = self.cluster.run_job(
- cmd, "IC-prior-"+self.exp.expname, cfg_update=dict(time="18"), depends_on=[depends_on])
+ cmd, "IC-prior-"+self.expname, cfg_update=dict(time="18"), depends_on=[depends_on])
return id
def update_IC_from_DA(self, assim_time, depends_on=None):
@@ -351,40 +362,42 @@ class WorkFlows(object):
Returns:
str: job ID of the submitted job
"""
- cmd = self.cluster.python+' '+self.cluster.scripts_rundir + \
+ cmd = self.cluster.python+' '+self.scripts_rundir + \
'/update_IC.py '+assim_time.strftime('%Y-%m-%d_%H:%M')
- id = self.cluster.run_job(cmd, "IC-update-"+self.exp.expname,
+ id = self.cluster.run_job(cmd, "IC-update-"+self.expname,
cfg_update=dict(time="18"), depends_on=[depends_on])
return id
def create_satimages(self, init_time, depends_on=None):
"""Run a job array, one job per ensemble member, to create satellite images"""
cmd = 'module purge; module load rttov/v13.2-gcc-8.5.0; ' \
- + 'python ~/RTTOV-WRF/run_init.py '+self.cluster.archivedir+init_time.strftime('/%Y-%m-%d_%H:%M/ ') \
+ + 'python ~/RTTOV-WRF/run_init.py '+self.archivedir+init_time.strftime('/%Y-%m-%d_%H:%M/ ') \
+ '$SLURM_ARRAY_TASK_ID'
- id = self.cluster.run_job(cmd, "RTTOV-"+self.exp.expname,
- cfg_update={"ntasks": "1", "time": "60", "mem": "10G", "array": "1-"+str(self.exp.n_ens)}, depends_on=[depends_on])
+ id = self.cluster.run_job(cmd, "RTTOV-"+self.expname,
+ cfg_update={"ntasks": "1", "time": "60", "mem": "10G",
+ "array": "1-"+str(self.cfg.ensemble_size)},
+ depends_on=[depends_on])
return id
def gen_obsseq(self, depends_on=None):
"""(not included in DART-WRF)"""
- cmd = self.cluster.python+' '+self.cluster.scripts_rundir+'/obsseq_to_netcdf.py'
+ cmd = self.cluster.python+' '+self.scripts_rundir+'/obsseq_to_netcdf.py'
id = self.cluster.run_job("obsseq_netcdf", cfg_update={"time": "10", "mail-type": "FAIL,END"},
depends_on=[depends_on])
return id
def evaluate_obs_posterior_after_analysis(self, init, valid, depends_on=None):
- cmd = self.cluster.python+' '+self.cluster.scripts_rundir+'/evaluate_obs_space.py ' + \
+ cmd = self.cluster.python+' '+self.scripts_rundir+'/evaluate_obs_space.py ' + \
init.strftime('%Y-%m-%d_%H:%M,') + \
valid.strftime('%Y-%m-%d_%H:%M:%S')
- id = self.cluster.run_job(cmd, 'eval+1'+self.exp.expname, cfg_update={"ntasks": "16", "mem": "80G", "ntasks-per-node": "16", "ntasks-per-core": "2",
+ id = self.cluster.run_job(cmd, 'eval+1'+self.expname, cfg_update={"ntasks": "16", "mem": "80G", "ntasks-per-node": "16", "ntasks-per-core": "2",
"time": "9", "mail-type": "FAIL"},
depends_on=[depends_on])
- # cmd = self.cluster.python+' '+self.cluster.scripts_rundir + \
+ # cmd = self.cluster.python+' '+self.scripts_rundir + \
# '/calc_linear_posterior.py '+init.strftime('%Y-%m-%d_%H:%M')
- # id = self.cluster.run_job(cmd, 'linpost'+self.exp.expname, cfg_update={"ntasks": "16", "mem": "80G", "ntasks-per-node": "16", "ntasks-per-core": "2",
+ # id = self.cluster.run_job(cmd, 'linpost'+self.expname, cfg_update={"ntasks": "16", "mem": "80G", "ntasks-per-node": "16", "ntasks-per-core": "2",
# "time": "15", "mail-type": "FAIL"},
# depends_on=[id])
return id
@@ -392,17 +405,17 @@ class WorkFlows(object):
def verify_sat(self, depends_on=None):
"""(not included in DART-WRF)"""
cmd = self.cluster.python+' /jetfs/home/lkugler/osse_analysis/plot_from_raw/analyze_fc.py ' + \
- self.exp.expname+' '+self.exp.nature_exp + ' sat has_node np=2 mem=110G'
+ self.expname+' '+self.cfg.nature_exp + ' sat has_node np=2 mem=110G'
- self.cluster.run_job(cmd, "verif-SAT-"+self.exp.expname,
+ self.cluster.run_job(cmd, "verif-SAT-"+self.expname,
cfg_update={"time": "60", "mail-type": "FAIL,END", "ntasks": "2",
"ntasks-per-node": "1", "ntasks-per-core": "2", "mem": "110G", }, depends_on=[depends_on])
def verify_wrf(self, depends_on=None):
"""(not included in DART-WRF)"""
cmd = self.cluster.python+' /jetfs/home/lkugler/osse_analysis/plot_from_raw/analyze_fc.py ' + \
- self.exp.expname+' '+self.exp.nature_exp + ' wrf has_node np=10 mem=250G'
+ self.expname+' '+self.cfg.nature_exp + ' wrf has_node np=10 mem=250G'
- self.cluster.run_job(cmd, "verif-WRF-"+self.exp.expname,
+ self.cluster.run_job(cmd, "verif-WRF-"+self.expname,
cfg_update={"time": "210", "mail-type": "FAIL,END", "ntasks": "10",
"ntasks-per-node": "10", "ntasks-per-core": "1", "mem": "250G"}, depends_on=[depends_on])
diff --git a/docs/source/tutorial1.rst b/docs/source/tutorial1.rst
index 4817bc4..ed73fa4 100644
--- a/docs/source/tutorial1.rst
+++ b/docs/source/tutorial1.rst
@@ -39,14 +39,14 @@ Then set ``exp_config='exp1.py`` in the call to :class:`dartwrf.workflows.WorkFl
Customize the settings in ``config/exp1.py``:
* `expname` should be a unique experiment name and will be used as folder name
-* `n_ens` is the ensemble size
+* `ensemble_size` is the ensemble size
* `update_vars` are the WRF variables which shall be updated by the assimilation
.. code-block:: python
exp = Experiment()
exp.expname = "exp1"
- exp.n_ens = 40
+ exp.ensemble_size = 40
exp.update_vars = ['THM', 'PH', 'MU', 'QVAPOR',]
@@ -95,9 +95,9 @@ If you want to modify any parameters, specify your changes in a python dictionar
sampling_error_correction='.true.',
),
'&filter_nml':
- dict(ens_size=exp.n_ens,
- num_output_state_members=exp.n_ens,
- num_output_obs_members=exp.n_ens,
+ dict(ens_size=exp.ensemble_size,
+ num_output_state_members=exp.ensemble_size,
+ num_output_obs_members=exp.ensemble_size,
inf_flavor=['0', '4'],
output_members='.true.',
output_mean='.true.',
diff --git a/multiple_exps.py b/multiple_exps.py
new file mode 100644
index 0000000..94567bf
--- /dev/null
+++ b/multiple_exps.py
@@ -0,0 +1,79 @@
+import datetime as dt
+import sys
+import pandas as pd
+from dartwrf.workflows import WorkFlows
+from config import defaults
+
+
+# test multiple assimilation windows (11-12, 12-13, 13-14, )
+timedelta_btw_assim = dt.timedelta(minutes=15)
+assim_times_start = pd.date_range('2000-01-01 11:00', '2000-01-01 13:00', freq='h')
+
+for t0 in assim_times_start:
+
+ # set constants
+ w = WorkFlows(server_config='/jetfs/home/lkugler/DART-WRF/config/jet.py',
+ expname=t0.strftime('test_%H%M'))
+
+ # set potentially varying parameters
+ ens_size = 3
+ dart_nml = defaults.dart_nml
+ dart_nml.update(ens_size=ens_size)
+
+ w.configure(model_dx=2000,
+ ensemble_size=3,
+ dart_nml = dart_nml,
+ use_existing_obsseq=False,
+ update_vars = ['U', 'V', 'W', 'THM', 'PH', 'MU', 'QVAPOR', 'QCLOUD', 'QICE', 'QSNOW', 'PSFC'],
+ input_profile = '/mnt/jetfs/home/lkugler/data/initial_profiles/wrf/ens/2022-03-31/raso.fc.<iens>.wrfprof',
+ nature_wrfout_pattern = '/jetfs/home/lkugler/data/sim_archive/exp_v1.18_P1_nature+1/*/1/wrfout_d01_%Y-%m-%d_%H:%M:%S',
+ nature_exp = 'nat_250m_blockavg2km',)
+
+
+ time = t0
+ id = None
+ w.prepare_WRFrundir(time)
+ id = w.run_ideal(depends_on=id)
+ sys.exit()
+
+ prior_init_time = time - dt.timedelta(hours=1)
+ prior_valid_time = time
+ prior_path_exp = '/jetfs/home/lkugler/data/sim_archive/exp_v1.19_P2_noDA+1/'
+
+ # assimilate at these times
+ assim_times = pd.date_range(time, time + dt.timedelta(hours=1), freq=timedelta_btw_assim)
+ last_assim_time = assim_times[-1]
+
+ # loop over assimilations
+ for i, t in enumerate(assim_times):
+
+ id = w.assimilate(time, prior_init_time, prior_valid_time, prior_path_exp, depends_on=id)
+
+ # 1) Set posterior = prior
+ id = w.prepare_IC_from_prior(prior_path_exp, prior_init_time, prior_valid_time, depends_on=id)
+
+ # 2) Update posterior += updates from assimilation
+ id = w.update_IC_from_DA(time, depends_on=id)
+
+ # How long shall we integrate?
+ timedelta_integrate = timedelta_btw_assim
+ output_restart_interval = timedelta_btw_assim.total_seconds()/60
+ if time == last_assim_time:
+ timedelta_integrate = dt.timedelta(hours=4)
+ output_restart_interval = 9999 # no restart file after last assim
+
+ # 3) Run WRF ensemble
+ id = w.run_ENS(begin=time, # start integration from here
+ end=time + timedelta_integrate, # integrate until here
+ output_restart_interval=output_restart_interval,
+ depends_on=id)
+
+ if t < last_assim_time:
+ # continue the next cycle
+ prior_init_time = assim_times[i]
+ prior_valid_time = assim_times[i+1]
+ prior_path_exp = w.archivedir # use own exp path as prior
+ else:
+ # exit the cycles
+ break
+
\ No newline at end of file
--
GitLab