Optimization of Neural Network Radiation Emulator using Sherpa

These files contain main source codes for developing neural network radiation emulator and weight/bias coefficients for implementing to the WRF model. 

Reference: Kim, P. S. & Song, H.-J. (2022), Usefulness of automatic hyperparameter optimization in developing radiation emulator in a numerical weather prediction model, Atmosphere, 13, 721.

Source : Main codes for optimizing neural network radiation emulator using Sherpa

weight and bias coefficients from neural network training for two control simulations (l1, l2) and Sherpa

l1_full : learning rate 0.0001 and batch size 3000000

l1_50000 : learning rate 0.0001 and batch size 50000

l1_5000 : learning rate 0.0001 and batch size 5000

l1_500 : learning rate 0.0001 and batch size 500

l2_full : learning rate 1 and batch size 3000000

l2_50000 : learning rate 1 and batch size 50000

l2_5000 : learning rate 1 and batch size 5000

l2_500 : learning rate 1 and batch size 500

sp_full : Sherpa and batch size 3000000

sp_50000 : Sherpa and batch size 50000

sp_5000 : Sherpa and batch size 5000

sp_500 : Sherpa and batch size 500

Training and validation sets are given in the following link:
July: https://doi.org/10.5281/zenodo.5636127

Source codes for control simulations (l1, l2) and neural network radiation emulator for simulating the WRF model are avaliable in the following link:
https://doi.org/10.5281/zenodo.5638436

Reference: Song, H.-J., Roh, S., Lee, J., Nam, G., Yun, E., Yoon, J., & Kim, P. S., (2022). Benefits of stochastic weight averaging in developing neural network radiation scheme for numerical weather prediction. Journal of Advances in Modeling Earth Systems, https://doi.org/10.1002/essoar.10508964.2