ufs-da-workflow

UFS DA (Data Assimilation) Workflow

• Available coupling configurations:

1. S2SWA: ATM (FV3+CCPP) + OCN (MOM6) + ICE (CICE) + WAV (WW3)
2. NG-GODAS: ATM (DATM) + OCN (MOM6) + ICE (CICE)

Quick Start Guide

1. Clone the develop branch of the authoritative repository:

git clone -b develop --recursive https://github.com/ufs-community/ufs-da-workflow

2. Move to the sorc directory:

cd ufs-da-workflow/sorc

3. Run the build script:

• Workflow components: YES, JEDI-bundle: NO

./app_build.sh -a=[APP]

where [APP] is S2SWA or NG-GODAS.

• Workflow components: YES, JEDI-bundle: YES

./app_build.sh -a=[APP] --jedi=on

• Workflow components: NO, JEDI-bundle: YES

./app_build.sh --jedi=only

4. Load the python environment to set up the workflow:

cd ..
module use modulefiles
module load wflow_[machine] 

where [machine] is gaeac6, hera, hercules, orion, or ursa.

5. Copy a sample configuration and modify it as needed:

cd parm
cp config_samples/config.[sample_case].yaml config.yaml
vim config.yaml

Change the parameter values such as ACCOUNT as needed.

6. Set up the workflow in the case-specific experiment directory:

./setup_wflow_env.py

7. Move to the experimental case directory:

cd ../../exp_case/[EXP_CASE_NAME]

where [EXP_CASE_NAME] is specified in the configuration file config.yaml.

8. Launch the workflow tasks:

./automate_launch_script.py -i [time interval in seconds]

where the default value of [time interval in seconds] is 30. This means that the launch script launch_rocoto_wflow.sh is submitted every 30 seconds.