Multi-Parameter Solar Wind Prediction Based on Deep Learning
-
摘要: 太阳风中高速等离子体流与地球磁层相互作用可引发地磁暴等空间天气事件，因此，准确预测太阳风参数对于空间天气预警和现代技术系统的稳定运行至关重要。本文利用具有片段嵌入和交叉注意力机制的深度学习模型TimeXer来深入挖掘太阳风速度、动压、质子密度、质子温度间的复杂依赖关系，该模型仅利用历史太阳风数据和时间信息，即可实现未来72小时太阳风参数的高准确率预测，并且具有可解释性。在太阳活动水平低年（2021年）和太阳活动水平高年（2024年）的测试结果表明：（1）TimeXer对太阳风速度、动压、质子密度、质子温度的预测均方根误差分别为68.39 km/s、2.12 nPa、5.02 N/cm3、8.83×104 K，绝对误差分别为47.65 km/s、1.00 nPa、3.13 N/cm3、4.49×104 K，与几类传统及先进的深度学习方法相比，本模型性能更优，即使在磁暴期间，也能较准确地捕获各太阳风参数的整体变化趋势；（2）当历史数据输入长度为336时（即太阳风约14天的准周期），模型的预测性能最佳；（3）基于太阳风参数依赖关系的联合建模预测明显优于单参数预测；（4）交叉注意力权重分析显示，四个太阳风参数对质子温度和太阳风速度预测的重要性基本一致，太阳风速度和质子温度对质子密度的预测贡献较大，而质子温度、太阳风速度与年时间信息对太阳风动压的预测影响较强，且随着时间信息尺度的增大，时间信息的重要性进一步提升。
Abstract: When interacting with the Earth's magnetosphere, high-speed plasma flows in the solar wind can trigger space weather events such as geomagnetic storms. Therefore, accurately forecasting solar wind parameters is critical for early warnings of space weather and the stable operation of modern technological systems. This study employs TimeXer, a deep learning model incorporating patch embedding and cross-attention mechanism, to explore the complex dependencies among solar wind speed, dynamic pressure, proton density, and proton temperature. This model can accurately predict solar wind parameters for the next 72 hours by only using historical solar wind data and time information, and it is also interpretable. Test results during low solar activity level (2021) and high solar activity level (2024) periods demonstrate: (1) TimeXer's root mean square errors (RMSE) for solar wind speed, dynamic pressure, proton density, and proton temperature are 68.39 km/s, 2.12 nPa, 5.02 N/cm³, and 8.83×10⁴ K, respectively, while the mean absolute errors (MAE) are 47.65 km/s, 1.00 nPa, 3.13 N/cm³, and 4.49×10⁴ K. Compared with traditional and advanced deep learning methods, TimeXer exhibits superior performance, even can accurately capture the overall variation trends of solar wind parameters during geomagnetic storm. (2) Optimal prediction performance is achieved with a historical input length of 336 hours (corresponding to the solar wind's ~14-day quasi-period). (3) The joint modeling prediction based on the inter-parameter dependencies of solar wind parameters is significantly better than the single-parameter prediction. (4) Cross-attention weight analysis reveals that the four solar wind parameters contribute similarly to proton temperature and solar wind speed predictions. The solar wind speed and proton temperature contribute more to the prediction of proton density, while the proton temperature, solar wind speed, and annual time information have a more substantial influence on the prediction of solar wind dynamic pressure. Moreover, the importance of time information grows with increasing scales of time information.-
Key words:
- Solar wind parameters /
- Joint prediction /
- Attention mechanism /
- Interpretability
-
-
计量
- 文章访问数: 109
- HTML全文浏览量: 9
- PDF下载量: 7
-
被引次数:
0(来源:Crossref)
0(来源:其他)
下载:
