The code in this repository is associated with the paper "Comparative Analysis of TEC Prediction Models During Geomagnetic Storm and Quiet Conditions Using Diffusion, Transformer, and SARIMA".
We would like to express our gratitude to the following open-source projects for their contributions to our work:
Generative Diffusion model: We thank the "diffusion" project for open-sourcing their model code, which provided inspiration and references for our project, please refer to the link for more information.
For the paper "Generative Time Series Forecasting with Diffusion, Denoise, and Disentanglement", please refer to the arXiv link for more information.
Data : The datasets utilized in this manuscript, specifically the Global Ionosphere Maps (GIMs) Total Electron Content (TEC) data, were sourced from global navigation satellite systems (GNSS). These data are publicly available on the website GIMs TEC Data. Additionally, the Disturbance Storm Time (Dst) index data, crucial for our analysis, can be accessed via Dst Index Data. The C1PG prediction data, referenced in our study, originates from a global TEC prediction model developed by the Center for Orbit Determination in Europe (CODE). This data can be found at C1PG Prediction Data.