Application in predicting the geomagnetic storm with variation characteristics of cosmic ray

An algorithm is introduced to use the cosmic ray neutron detector data to predict great geomagnetic storms. The connection between cosmic ray and great geomagnetic storms and the relationship between cosmic ray stations are qualitatively analyzed. The neutron detector data from Irkutsk and Oulu stations are employed and wavelet denoising technology is used to improve the stability of the data. It is found that under the same conditions of universal time the correlation of the two stations is higher than quiet days during geomagnetic disturbances, while it is contrary under the same local time. Thus the variation in disturbed time and quiet time can be used to predict geomagnetic storms. The algorithm is specifically used to analyze the geomagnetic storm events in July 2004. It is found that before geomagnetic storm, relative fluxes of cosmic rays of Irkutsk and Oulu station became different. It can be used as a precursor of strong geomagnetic storms. Statistics with all relative events during March 2001 to May 2005 support an appropriate threshold related to the relative difference of cosmic ray about strong geomagnetic storms. It is tested with all events during September 2005 to December 2011 with the threshold. The result turned out to be encouraging with the accuracy rate reaching 87.5% (7 out of 8) and false forecast rate reaching 35.7% (5 out of 14).