Getting Started

This page shows you how to run regression test cases using ufs-s2s-model on Hera (NOAA's R&D HPCS) or WCOSS Phase 2. If you have any problems, please check the issue pages.

1. Downloading the source code
2. Compiling the models
3. Running test cases

Downloading the source code

% git clone https://github.com/ufs-community/ufs-s2s-model
% cd ufs-s2s-model
% git checkout develop (or tags)
% git submodule update --init --recursive

Compiling the models

In your .cshrc, you need to set "limit stacksize unlimited" or in your .bashrc, you need to set "ulimit -s unlimited". Otherwise your jobs might hang.

To build UFS-S2S-model using the NEMSAppbuilder:

% ./NEMS/NEMSAppBuilder app=coupledFV3_MOM6_CICE 

To build UFS-S2S-model using make:

cd NEMS
make app=coupledFV3_MOM6_CICE build

To see what other options are available when using make (e.g. cleaning components separately):

make app=coupledFV3_MOM6_CICE

This will echo the available options.

Check the Advanced Topics for information about how build and run the model in Debug mode.

Running test cases

To run all 4 test cases (or compsets):

% ./NEMS/NEMSCompsetRun -f --project=USER_CPU_ACCOUNT 

where USER_CPU_ACCOUNT is a cpu_account the user belongs to, e.g. marine_cpu. This step will also compile the code automatically if the code has not been compiled yet.

To specify the location where the compsets will be run:

% ./NEMS/NEMSCompsetRun -f --project=USER_CPU_ACCOUNT --temp-dir=path_to_scratch_directory

Note: if you have an environment variable set for a specific scratch directory, you can use that here:

% ./NEMS/NEMSCompsetRun -f --project=USER_CPU_ACCOUNT --temp-dir=$S1/test_name 

where $S1 for example would point to the users' stmp1 space and test_name is any name chosen by the user. The directory test_name will be created if it does not exist.