Abstract
We present a suite of VAMPIRE (Van Allen belt Multi-day Predictions by Implementing a Random Forest for Electrons) models capable of predicting if the outer radiation belt crosses set percentile thresholds. We use Random Forest classification models to predict if the daily ∼2 MeV electron flux level across the outer radiation belt exceeds thresholds from the 60th to the 95th percentiles. Most models show a balanced accuracy of >0.7 (>70%) at nowcasting and ∼0.6 (60%) at forecasting up to 6 days in advance, a longer forecast than current operational models. Using feature importance (mean decrease in impurity), we determine the key inputs that are important in driving increasing flux levels and over what timescales they have an impact. Crucially, we find that only the average AL index from various days beforehand is required to be able to forecast radiation belt fluxes with good skill, meaning that models such as these could be operationally viable for space weather stakeholders. We call this suite a Collection Of VAMPIRE models for Enhanced Notification.
| Original language | English |
|---|---|
| Article number | e2026SW004989 |
| Number of pages | 21 |
| Journal | Space Weather |
| Volume | 24 |
| Issue number | 6 |
| Early online date | 4 Jun 2026 |
| DOIs | |
| Publication status | Published - Jun 2026 |
Keywords
- radiation belt
- forecasting
- machine learning
- random forest
Fingerprint
Dive into the research topics of 'COVEN: Providing a Variety of Threshold-Based Forecasts for the Outer Radiation Belt'. Together they form a unique fingerprint.Projects
- 1 Finished
-
STFC Consolidated Grant for the Solar and Space Physics Group at Northumbria University
McLaughlin, J. (PI), Bentley, S. (CoI), Rae, J. (CoI), Jeffrey, N. (CoI), Coxon, J. (CoI) & Watt, C. (CoI)
Science and Technologies Facilities Council
1/04/23 → 31/03/26
Project: Research
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver