This repository contains the code, in mostly chronological order, to download, train, and analyze the BoltCast model.

The data acquisition folder has the wget/s3fs scripts to acquire the GLM/GFS data.

The data cleaning folder has the code, in step order, how the GFS data was downselected and matched to the GLM observations.

The model training folder has the code to develop the two model architectures.

The model analysis folder has the code to conduct the metric analysis and explainable AI techniques.

Python packages are included in the text file.

Sample model output is included in the BC_output.mov file, recognizing the file size limitations in GitHub.

Line by line diagrams are included from the corresponding paper.

Please reach out to bmac7167@gmail.com if you need assistance in reviewing this repository.