How good is the prediction of space weather based on solar and interplanetary properties?
Abstract
Space weather prediction involves advance forecasting of the magnitude and the on-set time of major geomagnetic storms at the earth. In a previous attempt, a logistic regression model based on solar and interplanetary variables was developed which was based on an exhaustive study of the solar origins of major geomagnetic storms recorded during 1996- 2002. In this paper, the logistic regression model developed earlier is refined using a new database of the solar and interplanetary characteristics of the major geomagnetic storms recorded during 2003-2004, which leads to better prediction results. The model is also used to estimate the relative importance of each solar and interplanetary variable in predicting major geomagnetic storms. In an attempt aimed at an early prediction of the occurrence of geomagnetic storms, the interplanetary variables are excluded and a new model based only on the solar variables is developed. The new model did not perform well, which indicates that the solar variables responsible for geomagnetic activity at the Earth are not well-understood.
- Publication:
-
Proceedings of the ILWS Workshop
- Pub Date:
- 2006
- Bibcode:
- 2006ilws.conf..102S
- Keywords:
-
- Coronal mass ejections;
- geomagnetic storms;
- interplanetary properties;
- space weather prediction.