The Data and Codes for Training, Testing, and Prognostic Validation of A ResNet Ensemble for Moist Physics (ResCu-en)
Authors/Creators
- 1. Department of Earth System Science, Tsinghua University, Beijing, China
- 2. Scripps Institution of Oceanography, La Jolla, CA, USA
Description
The data and codes for Training, Testing, and Prognostic Validation of A ResNet Ensemble for Moist Physics (ResCu-en) are stored in this repositary.
This project is built on python3.7 and tensorflow-gpu2.3.0, and the scripts for analysis and plots are on jupyter-notebook.
Please be sure to install all considered python packages in an environment.
Please read the ReadME-2.txt.
For the entire training and testing datasets in both the baseline and +4K SST climates. Please download them from Dryad (https://doi.org/10.6075/J0CZ35PP and https://doi.org/10.6075/J03J3BGF), Onedrive (https://1drv.ms/u/s!ArKTPPs6U_9DjxPJeSReKlbsLzyh?e=PDlWYJ), and Dropbox (https://www.dropbox.com/s/yc4fx35laqwt0fu/SPCAM_ML_4K.tar.gz?dl=0 and https://www.dropbox.com/s/4pxahzwt9v55u2m/SPCAM_ML_RAD.tar.gz?dl=0).