A statistical technique for space weather forecasting
- Paper number
IAC-08.D5.3.4
- Author
Dr. Aleksei Parnowski, Space Research Institute of NAS and NSA of Ukraine, Ukraine
- Year
2008
- Abstract
A statistical technique for space weather forecasting is proposed. It is based on the linear regression method, which, when applied to a significantly large sample of satellite (e.g. ACE) data, provides an accurate and reliable forecast of space weather parameters several hours ahead. Numerical routines for real-time prediction of space weather parameters have been created and the corresponding software for the Dst index 9-hour forecast have been extensively tested. The linear correlation between the predicted and the measured Dst values ranges from 0.798 for 9-hour forecast to 0.987 for 1-hour forecast. The prediction efficiency ranges from 0.652 for 9-hour forecast to 0.975 for 1-hour forecast. Note that not only Dst but virtually any space weather parameter including high-energy electron flux can be predicted in this way. This technique can also be used for determination of geoeffective parameters and construction of phenomenological models of solar wind - magnetosphere interaction. These models can provide substantial help for understanding of physical mechanisms that drive geomagnetic storms.
- Abstract document
- Manuscript document
(absent)