Solar Wind Prediction Using Deep Learning
Abstract Emanating from the base of the Sun's corona, the solar wind fills the interplanetary medium with a magnetized stream of charged particles whose interaction with the Earth's magnetosphere has space weather consequences such as geomagnetic storms. Accurately predicting the solar win...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-09-01
|
| Series: | Space Weather |
| Subjects: | |
| Online Access: | https://doi.org/10.1029/2020SW002478 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1841536447016861696 |
|---|---|
| author | Vishal Upendran Mark C. M. Cheung Shravan Hanasoge Ganapathy Krishnamurthi |
| author_facet | Vishal Upendran Mark C. M. Cheung Shravan Hanasoge Ganapathy Krishnamurthi |
| author_sort | Vishal Upendran |
| collection | DOAJ |
| description | Abstract Emanating from the base of the Sun's corona, the solar wind fills the interplanetary medium with a magnetized stream of charged particles whose interaction with the Earth's magnetosphere has space weather consequences such as geomagnetic storms. Accurately predicting the solar wind through measurements of the spatiotemporally evolving conditions in the solar atmosphere is important but remains an unsolved problem in heliophysics and space weather research. In this work, we use deep learning for prediction of solar wind (SW) properties. We use extreme ultraviolet images of the solar corona from space‐based observations to predict the SW speed from the National Aeronautics and Space Administration (NASA) OMNIWEB data set, measured at Lagragian Point 1. We evaluate our model against autoregressive and naive models and find that our model outperforms the benchmark models, obtaining a best fit correlation of 0.55 ± 0.03 with the observed data. Upon visualization and investigation of how the model uses data to make predictions, we find higher activation at the coronal holes for fast wind prediction (≈3 to 4 days prior to prediction), and at the active regions for slow wind prediction. These trends bear an uncanny similarity to the influence of regions potentially being the sources of fast and slow wind, as reported in literature. This suggests that our model was able to learn some of the salient associations between coronal and solar wind structure without built‐in physics knowledge. Such an approach may help us discover hitherto unknown relationships in heliophysics data sets. |
| format | Article |
| id | doaj-art-e46f2bda717c428b8ca470b91d34acf2 |
| institution | Kabale University |
| issn | 1542-7390 |
| language | English |
| publishDate | 2020-09-01 |
| publisher | Wiley |
| record_format | Article |
| series | Space Weather |
| spelling | doaj-art-e46f2bda717c428b8ca470b91d34acf22025-01-14T16:30:54ZengWileySpace Weather1542-73902020-09-01189n/an/a10.1029/2020SW002478Solar Wind Prediction Using Deep LearningVishal Upendran0Mark C. M. Cheung1Shravan Hanasoge2Ganapathy Krishnamurthi3Inter‐University Centre for Astronomy and Astrophysics Pune IndiaLockheed Martin Solar and Astrophysics Laboratory Palo Alto CA USADepartment of Astronomy and Astrophysics Tata Institute of Fundamental Research Mumbai IndiaDepartment of Engineering Design Indian Institute of Technology‐Madras Chennai IndiaAbstract Emanating from the base of the Sun's corona, the solar wind fills the interplanetary medium with a magnetized stream of charged particles whose interaction with the Earth's magnetosphere has space weather consequences such as geomagnetic storms. Accurately predicting the solar wind through measurements of the spatiotemporally evolving conditions in the solar atmosphere is important but remains an unsolved problem in heliophysics and space weather research. In this work, we use deep learning for prediction of solar wind (SW) properties. We use extreme ultraviolet images of the solar corona from space‐based observations to predict the SW speed from the National Aeronautics and Space Administration (NASA) OMNIWEB data set, measured at Lagragian Point 1. We evaluate our model against autoregressive and naive models and find that our model outperforms the benchmark models, obtaining a best fit correlation of 0.55 ± 0.03 with the observed data. Upon visualization and investigation of how the model uses data to make predictions, we find higher activation at the coronal holes for fast wind prediction (≈3 to 4 days prior to prediction), and at the active regions for slow wind prediction. These trends bear an uncanny similarity to the influence of regions potentially being the sources of fast and slow wind, as reported in literature. This suggests that our model was able to learn some of the salient associations between coronal and solar wind structure without built‐in physics knowledge. Such an approach may help us discover hitherto unknown relationships in heliophysics data sets.https://doi.org/10.1029/2020SW002478solar winddeep learningAIACNNLSTMGrad‐CAM |
| spellingShingle | Vishal Upendran Mark C. M. Cheung Shravan Hanasoge Ganapathy Krishnamurthi Solar Wind Prediction Using Deep Learning Space Weather solar wind deep learning AIA CNN LSTM Grad‐CAM |
| title | Solar Wind Prediction Using Deep Learning |
| title_full | Solar Wind Prediction Using Deep Learning |
| title_fullStr | Solar Wind Prediction Using Deep Learning |
| title_full_unstemmed | Solar Wind Prediction Using Deep Learning |
| title_short | Solar Wind Prediction Using Deep Learning |
| title_sort | solar wind prediction using deep learning |
| topic | solar wind deep learning AIA CNN LSTM Grad‐CAM |
| url | https://doi.org/10.1029/2020SW002478 |
| work_keys_str_mv | AT vishalupendran solarwindpredictionusingdeeplearning AT markcmcheung solarwindpredictionusingdeeplearning AT shravanhanasoge solarwindpredictionusingdeeplearning AT ganapathykrishnamurthi solarwindpredictionusingdeeplearning |