Skip to content

CME Arrival Time Prediction Using Convolutional Neural Network

Notifications You must be signed in to change notification settings

MohamedNedal/CME-CNN

 
 

Repository files navigation

CME-CNN

This is an implementation of the CNN regression model for Coronal Mass Ejections (CMEs) arrival time prediction using Python 3, Keras and TensorFlow. The model is trained to learn the mapping between features of CMEs white-light observations (from SOHO LASCO C2) and its corresponding transit time from the Sun to the Earth. For example, the features extracted by the first convolutional layer for a CME is shown below:

CME Arrival Time Prediction Using Convolutional Neural Network
Yimin Wang, Jiajia Liu, Ye Jiang, Robert Erdélyi
The Astrophysical Journal, 2019

Model structure

The architecture of the proposed CNN regression model is:

Datasets

A catalogue of all observed geo-effective CMEs since the beginning of the SOHO era, i.e. from 1996 to early 2018 was established by combining the following four CME databases: the Richardson and Cane list (http://www.srl.caltech.edu/ACE/ASC/DATA/level3/icmetable2.htm), the full halo CME list provided by the University of Science and Technology of China (http://space.ustc.edu.cn/dreams/fhcmes/index.php), the George Mason University CME/ICME list (http://solar.gmu.edu/heliophysics/index.php/GMU_CME/ICME_List), and the CME Scoreboard by NASA (https://kauai.ccmc.gsfc.nasa.gov/CMEscoreboard/). After removing duplicates, 276 geo-effective events were obtained. Deatails on the dataset construction could be found in the paper (see citation below).

Training

To train a new model:
python3 train.py

Prediction analyses

To visualize the model:
python3 visualize_model.py

To analyse prediction performance:
python3 results_analysis.py

Citation

If you use this code for your research, please cite our paper CME Arrival Time Prediction Using Convolutional Neural Network:

@article{cme2019,
  title={CME Arrival Time Prediction Using Convolutional Neural Network},
  author={Yimin Wang, Jiajia Liu, Ye Jiang, Robert Erdélyi},
  journal={The Astrophysical Journal},
  year={2019}
}

About

CME Arrival Time Prediction Using Convolutional Neural Network

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%