Skip to main content

coupled model configuration generation

Project description

CoupledModelDriver

tests codecov build version license style

coupledmodeldriver generates an overlying job submission framework and configuration directories for NEMS-coupled coastal ocean model ensembles.

It utilizes nemspy to generate NEMS configuration files, shares common configurations between runs, and organizes spinup and mesh partition into separate jobs for dependant submission.

Supported models and platforms

  • models
    • circulation models
    • forcings
      • ATMESH
      • WW3DATA
      • HURDAT best track
      • OWI
  • platforms
    • local
    • Slurm
      • Hera
      • Stampede2

Usage

Example scripts can be found at examples/<platform>

1. generate JSON configuration files

The following command creates a configuration for coupling (ATMESH + WW3DATA) -> ADCIRC over a small Shinnecock Inlet mesh:

initialize_adcirc \
    --platform HERA \
    --mesh-directory /scratch2/COASTAL/coastal/save/shared/models/meshes/shinnecock/v1.0 \
    --output-directory hera_shinnecock_ike_spinup_tidal_atmesh_ww3data \
    --modeled-start-time 20080823 \
    --modeled-duration 14:06:00:00 \
    --modeled-timestep 00:00:02 \
    --nems-interval 01:00:00 \
    --tidal-spinup-duration 12:06:00:00 \
    --adcirc-executable /scratch2/COASTAL/coastal/save/shared/repositories/ADC-WW3-NWM-NEMS/NEMS/exe/NEMS.x \
    --adcprep-executable /scratch2/COASTAL/coastal/save/shared/repositories/ADC-WW3-NWM-NEMS/ADCIRC/work/adcprep \
    --modulefile /scratch2/COASTAL/coastal/save/shared/repositories/ADC-WW3-NWM-NEMS/modulefiles/envmodules_intel.hera \
    --generate-script \
    --forcings tidal,atmesh,ww3data \
    --tidal-source TPXO \
    --tidal-path /scratch2/COASTAL/coastal/save/shared/models/forcings/tides/h_tpxo9.v1.nc \
    --atmesh-path /scratch2/COASTAL/coastal/save/shared/models/forcings/shinnecock/ike/wind_atm_fin_ch_time_vec.nc \
    --ww3data-path /scratch2/COASTAL/coastal/save/shared/models/forcings/shinnecock/ike/ww3.Constant.20151214_sxy_ike_date.nc

Alternatively, the following Python code creates the same configuration:

from datetime import datetime, timedelta
from pathlib import Path

from adcircpy.forcing.tides import Tides
from adcircpy.forcing.tides.tides import TidalSource
from adcircpy.forcing.waves.ww3 import WaveWatch3DataForcing
from adcircpy.forcing.winds.atmesh import AtmosphericMeshForcing

from coupledmodeldriver import Platform
from coupledmodeldriver.generate import (
    NEMSADCIRCRunConfiguration,
)

# directory to which to write configuration
OUTPUT_DIRECTORY = Path(__file__).parent / Path(__file__).stem

# start and end times for model
MODELED_START_TIME = datetime(year=2012, month=10, day=22, hour=6)
MODELED_DURATION = timedelta(days=4, hours=5)
MODELED_TIMESTEP = timedelta(seconds=2)
TIDAL_SPINUP_DURATION = timedelta(days=12.5)
NEMS_INTERVAL = timedelta(hours=1)

# directories containing forcings and mesh
MESH_DIRECTORY = '/scratch2/COASTAL/coastal/save/shared/models/meshes/hsofs/120m/v3.0_20210401'
FORCINGS_DIRECTORY = '/scratch2/COASTAL/coastal/save/shared/models/forcings/hsofs/120m/sandy'
HAMTIDE_DIRECTORY = '/scratch2/COASTAL/coastal/save/shared/models/forcings/tides/hamtide'
TPXO_FILENAME = '/scratch2/COASTAL/coastal/save/shared/models/forcings/tides/h_tpxo9.v1.nc'

# connections between coupled components
NEMS_CONNECTIONS = ['ATM -> OCN', 'WAV -> OCN']
NEMS_SEQUENCE = [
    'ATM -> OCN',
    'WAV -> OCN',
    'ATM',
    'WAV',
    'OCN',
]

# platform-specific parameters
PLATFORM = Platform.HERA
ADCIRC_PROCESSORS = 15 * PLATFORM.value['processors_per_node']
NEMS_EXECUTABLE = (
    '/scratch2/COASTAL/coastal/save/shared/repositories/ADC-WW3-NWM-NEMS/NEMS/exe/NEMS.x'
)
ADCPREP_EXECUTABLE = (
    '/scratch2/COASTAL/coastal/save/shared/repositories/ADC-WW3-NWM-NEMS/ADCIRC/work/adcprep'
)
MODULEFILE = '/scratch2/COASTAL/coastal/save/shared/repositories/ADC-WW3-NWM-NEMS/modulefiles/envmodules_intel.hera'
SLURM_JOB_DURATION = timedelta(hours=6)

if __name__ == '__main__':
    # initialize `adcircpy` forcing objects
    FORCINGS_DIRECTORY = Path(FORCINGS_DIRECTORY)
    tidal_forcing = Tides(tidal_source=TidalSource.TPXO, resource=TPXO_FILENAME)
    tidal_forcing.use_all()
    wind_forcing = AtmosphericMeshForcing(
        filename=FORCINGS_DIRECTORY / 'Wind_HWRF_SANDY_Nov2018_ExtendedSmoothT.nc',
        nws=17,
        interval_seconds=3600,
    )
    wave_forcing = WaveWatch3DataForcing(
        filename=FORCINGS_DIRECTORY / 'ww3.HWRF.NOV2018.2012_sxy.nc',
        nrs=5,
        interval_seconds=3600,
    )
    forcings = [tidal_forcing, wind_forcing, wave_forcing]

    configuration = NEMSADCIRCRunConfiguration(
        mesh_directory=MESH_DIRECTORY,
        modeled_start_time=MODELED_START_TIME,
        modeled_end_time=MODELED_START_TIME + MODELED_DURATION,
        modeled_timestep=MODELED_TIMESTEP,
        nems_interval=NEMS_INTERVAL,
        nems_connections=NEMS_CONNECTIONS,
        nems_mediations=None,
        nems_sequence=NEMS_SEQUENCE,
        tidal_spinup_duration=TIDAL_SPINUP_DURATION,
        platform=PLATFORM,
        perturbations=None,
        forcings=forcings,
        adcirc_processors=ADCIRC_PROCESSORS,
        slurm_partition=None,
        slurm_job_duration=SLURM_JOB_DURATION,
        slurm_email_address=None,
        nems_executable=NEMS_EXECUTABLE,
        adcprep_executable=ADCPREP_EXECUTABLE,
        source_filename=MODULEFILE,
    )
    configuration.write_directory(OUTPUT_DIRECTORY, overwrite=False)

Either method will create the directory hera_shinnecock_ike_spinup_tidal_atmesh_ww3data/ with the following JSON configuration files:

๐Ÿ“ฆ hera_shinnecock_ike_spinup_tidal_atmesh_ww3data/
โ”ฃ โœŽ configure_modeldriver.json
โ”ฃ โœŽ configure_adcirc.json
โ”ฃ โœŽ configure_nems.json
โ”ฃ โœŽ configure_slurm.json
โ”ฃ โœŽ configure_tidal_forcing.json
โ”ฃ โœŽ configure_atmesh.json
โ”ฃ โœŽ configure_ww3data.json

These files contain relevant configuration values for an ADCIRC run. You will likely wish to change these values to alter the resulting run, before generating the actual model configuration.

2. generate model configuration files

Run the following command to read the JSON configuration and generate the ADCIRC run configuration:

generate_adcirc

The resulting configuration will have the following structure:

๐Ÿ“ฆ hera_shinnecock_ike_spinup_tidal_atmesh_ww3data/
โ”ฃ โœŽ configure_modeldriver.json
โ”ฃ โœŽ configure_adcirc.json
โ”ฃ โœŽ configure_nems.json
โ”ฃ โœŽ configure_slurm.json
โ”ฃ โœŽ configure_tidal_forcing.json
โ”ฃ โœŽ configure_atmesh.json
โ”ฃ โœŽ configure_ww3data.json
โ”ฃ ๐Ÿ“‚ coldstart/
โ”ƒ  โ”ฃ ๐Ÿ“œ fort.13
โ”ƒ  โ”ฃ ๐Ÿ”— fort.14 -> ../fort.14
โ”ƒ  โ”ฃ ๐Ÿ“œ fort.15
โ”ƒ  โ”ฃ ๐Ÿ”— nems.configure -> ../nems.configure.coldstart
โ”ƒ  โ”ฃ ๐Ÿ”— config.rc -> ../config.rc.coldstart
โ”ƒ  โ”ฃ ๐Ÿ”— model_configure -> ../model_configure.coldstart
โ”ƒ  โ”ฃ ๐Ÿ”— adcprep.job -> ../job_adcprep_hera.job
โ”ƒ  โ”ฃ ๐Ÿ”— adcirc.job -> ../job_adcirc_hera.job.coldstart
โ”ƒ  โ”— ๐Ÿ”— setup.sh -> ../setup.sh.coldstart
โ”ฃ ๐Ÿ“‚ runs/
โ”ƒ  โ”— ๐Ÿ“‚ run_1/
โ”ƒ    โ”ฃ ๐Ÿ“œ fort.13
โ”ƒ    โ”ฃ ๐Ÿ”— fort.14 -> ../../fort.14
โ”ƒ    โ”ฃ ๐Ÿ“œ fort.15
โ”ƒ    โ”ฃ ๐Ÿ”— fort.67.nc -> ../../coldstart/fort.67.nc
โ”ƒ    โ”ฃ ๐Ÿ”— nems.configure -> ../../nems.configure.hotstart
โ”ƒ    โ”ฃ ๐Ÿ”— config.rc -> ../../config.rc.hotstart
โ”ƒ    โ”ฃ ๐Ÿ”— model_configure -> ../../model_configure.hotstart
โ”ƒ    โ”ฃ ๐Ÿ”— adcprep.job -> ../../job_adcprep_hera.job
โ”ƒ    โ”ฃ ๐Ÿ”— adcirc.job -> ../../job_adcirc_hera.job.hotstart
โ”ƒ    โ”— ๐Ÿ”— setup.sh -> ../../setup.sh.hotstart
โ”ฃ ๐Ÿ“œ fort.14
โ”ฃ ๐Ÿ“œ nems.configure.coldstart
โ”ฃ ๐Ÿ“œ nems.configure.hotstart
โ”ฃ ๐Ÿ“œ config.rc.coldstart
โ”ฃ ๐Ÿ“œ config.rc.hotstart
โ”ฃ ๐Ÿ“œ model_configure.coldstart
โ”ฃ ๐Ÿ“œ model_configure.hotstart
โ”ฃ ๐Ÿ“œ job_adcprep_hera.job
โ”ฃ ๐Ÿ“œ job_adcirc_hera.job.coldstart
โ”ฃ ๐Ÿ“œ job_adcirc_hera.job.hotstart
โ”ฃ ๐Ÿ“œ setup.sh.coldstart
โ”ฃ ๐Ÿ“œ setup.sh.hotstart
โ”ฃ ๐Ÿ“œ cleanup.sh
โ”—  โ–ถ run_hera.sh

3. run the model

Run the following to submit the model run to the Slurm job queue:

sh run_hera.sh

The queue will have the following jobs added:

   JOBID                  NAME CPUS NODE                      DEPENDENCY          SUBMIT_TIME           START_TIME             END_TIME
16368044 ADCIRC_MESH_PARTITION    1    1                          (null)  2021-02-18T19:29:17                  N/A                  N/A
16368045      ADCIRC_COLDSTART   11    1  afterany:16368044(unfulfilled)  2021-02-18T19:29:17                  N/A                  N/A
16368046 ADCIRC_MESH_PARTITION    1    1  afterany:16368045(unfulfilled)  2021-02-18T19:29:17                  N/A                  N/A
16368047       ADCIRC_HOTSTART   13    1  afterany:16368046(unfulfilled)  2021-02-18T19:29:17                  N/A                  N/A

Command-line interface

coupledmodeldriver exposes the following CLI commands:

  • initialize_adcirc
  • generate_adcirc

Initialize ADCIRC configuration (initialize_adcirc)

initialize_adcirc creates JSON configuration files according to the given parameters.

usage: initialize_adcirc [-h] --platform PLATFORM --mesh-directory MESH_DIRECTORY --modeled-start-time MODELED_START_TIME
                         --modeled-duration MODELED_DURATION --modeled-timestep MODELED_TIMESTEP [--nems-interval NEMS_INTERVAL]
                         [--modulefile MODULEFILE] [--tidal-spinup-duration TIDAL_SPINUP_DURATION] [--forcings FORCINGS]
                         [--adcirc-executable ADCIRC_EXECUTABLE] [--adcprep-executable ADCPREP_EXECUTABLE]
                         [--adcirc-processors ADCIRC_PROCESSORS] [--job-duration JOB_DURATION] [--output-directory OUTPUT_DIRECTORY]
                         [--generate-script] [--skip-existing]

optional arguments:
  -h, --help            show this help message and exit
  --platform PLATFORM   HPC platform for which to configure
  --mesh-directory MESH_DIRECTORY
                        path to input mesh (`fort.13`, `fort.14`)
  --modeled-start-time MODELED_START_TIME
                        start time within the modeled system
  --modeled-duration MODELED_DURATION
                        end time within the modeled system
  --modeled-timestep MODELED_TIMESTEP
                        time interval within the modeled system
  --nems-interval NEMS_INTERVAL
                        main loop interval of NEMS run
  --modulefile MODULEFILE
                        path to module file to `source`
  --tidal-spinup-duration TIDAL_SPINUP_DURATION
                        spinup time for ADCIRC tidal coldstart
  --forcings FORCINGS   comma-separated list of forcings to configure, from ['tidal', 'atmesh', 'besttrack', 'owi', 'ww3data']
  --adcirc-executable ADCIRC_EXECUTABLE
                        filename of compiled `adcirc` or `NEMS.x`
  --adcprep-executable ADCPREP_EXECUTABLE
                        filename of compiled `adcprep`
  --adcirc-processors ADCIRC_PROCESSORS
                        numbers of processors to assign for ADCIRC
  --job-duration JOB_DURATION
                        wall clock time for job
  --output-directory OUTPUT_DIRECTORY
                        directory to which to write configuration files
  --generate-script     write shell script to load configuration
  --skip-existing       skip existing files

ADCIRC run options that are not exposed by this command, such as runs or gwce_solution_scheme, can be specified by directly modifying the JSON files.

Generate ADCIRC configuration (generate_adcirc)

generate_adcirc reads a set of JSON configuration files and generates an ADCIRC run configuration from the options read from these files.

usage: generate_adcirc [-h] [--configuration-directory CONFIGURATION_DIRECTORY] [--output-directory OUTPUT_DIRECTORY]
                       [--skip-existing] [--verbose]

optional arguments:
  -h, --help            show this help message and exit
  --configuration-directory CONFIGURATION_DIRECTORY
                        path containing JSON configuration files
  --output-directory OUTPUT_DIRECTORY
                        path to store generated configuration files
  --skip-existing       skip existing files
  --verbose             show more verbose log messages

After this configuration is generated, the model can be started by executing the run_<platform>.sh script.

Project details


Release history Release notifications | RSS feed

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

coupledmodeldriver-1.2.21.tar.gz (34.4 kB view hashes)

Uploaded Source

Built Distribution

coupledmodeldriver-1.2.21-py3-none-any.whl (39.5 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page