Add Flexible Execution Options for swash_runmodel (Serial, Parallel, Batch, and Array Submissions)

[fernanqv]

It would be great if the method run_model allowed the user to run the model in the way that best suits their computational infrastructure. For example:

For a serial execution:

cd testcase; swashrun -input testcase

For a parallel execution:

export NP=8
source source.sh
cd testcase; mpirun -np $NP swashrun $INPUT

To submit each job in batch mode to a cluster:

sbatch launchSwash.sh

If possible, it would also be interesting to enable the option for the user to submit all test cases at once. The following example demonstrates the launch of a parametric job on Lustre:

sbatch --array=0-15%4 launchArraySwash.sh

Another example running the model from a container:

apptainer exec --bind `pwd`/../testcases:/tmp --pwd /tmp/a11stwav  hySwash.sif  swashrun -input a11stw01