This code is a standalone test rig around the YOGconvection parameterisation routines for debugging. It uses a series of input profiles from CAM to pass to and run the parameterisation before returning predicted profiles.
The test-rig is contained in test_cam_profile.f90.
Profiles are output at relevant points in the code for analysis.
This code has dependencies on the NetCDF and NetCDF-Fortran libraries. It requires a Fortran compiler and python.
Set up a virtual python environment and install any dependencies required for the code:
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
Before running the code generate the SAM sounding profiles in the YOG_convection/resources/
directory as described in the main README in the Contents section about YOG_convection/.
Then build and run the fortran code using CMake.\
mkdir build
cd build
cmake ..
cmake --build .
./CAM_profile
This will generate NN_test_output.nc in the profile_test/ directory containing
column profiles for a various timesteps and intermetiate variables.
Finally, run the plot_nn_profiles.py python code from within the virtual environment
to produce comparison plots of the profiles at different stages in the parameterisation.
These will be output to screen using matplotlib.
python plot_nn_profiles.py