This github is the code implementation corresponding to the article "WRF-ML v1.0: A Bridge between WRF v4.3 and Machine Learning Parameterizations and its Application to Atmospheric Radiative Transferr"
Please contact x7zhong@gmail.com if you have any questions about this code.
Contains the libraries necessary for building WRF.
Contains the library necessary for building and running WRF_DL
Contains the files used for building the library in dl-inference-plugin, and the python script to run ML models within WRF_DL
ML-based radiation emulators are saved.
Contains the files that are different from the original WRF v4.3 files and used for implementing the WRF-ML coupler. To run the WRF coupled with ML-based radiation schemes, you need to add files that did not exist or overwrite the existing WRF files.
Contains the examples about how to call Python script from Fortran code.
add the path to the WRF ML inference library for example: ##WRF DL INFERENCE LIB export WRF_DL_INFER_DIR=$DIR/dl-inference-plugin export LD_LIBRARY_PATH=$WRF_DL_INFER_DIR/lib:$LD_LIBRARY_PATH
After running the ./configure when compiling WRF, modify the WRF configure.wrf file as:
add -L${path_to_dl-inference-plugin/lib} and "-lDL_inference_plugin" following LIB_EXTERNAL
Add ML-based parameterization schemes' related modules in the main/depend.common file, e.g.:
This part is used for profiling