MIT Haystack 
Statistical Analysis of Solar Geomagnetic Storm Occurrences
Seth Sivak
About Us





     Several questions were directed to me before the start of this project.  Some of the questions were:
  • Are there maxima found near the Equinoxes?
  • Do geomagnetic storms occur the most during March and April?
  • Do very large storms happen at random or do they follow a pattern?
  • When is the best time to plan radar operations to collect data on these storms?
     The data we used was taken from the Madrigal Database created by the Millstone Hill Observatory.  This database contains geomagnetic and solar flux data from 1950 to 2002.  We used the Kp (planetary) index to analyze geomagnetic storms and we used F10.7 cm Solar X-Ray Flux to analyze the solar activity cycle.

A broad view of the data was taken and this graph was produced.  This graph shows the solar cycle (yearly average of F10.7 cm X-Ray Solar Flux in Green) and the number of geomagnetic storm elements that reach above Kp > 6 (histogram of Kp > 6 in Blue).  Clearly there is a relationship here between the sun and the earth.  Every 11 years the solar cycle reaches a maximum, however geomagnetic storms seem to have two maximums.  One geomagnetic maximum is found during the years prior to the absolute solar maximum, and the other is found during the years after the absolute solar maximum.  Several studies have mentioned this interesting phenomena and there are two basic results.  One states that the second geomagnetic maximum will occur three to four years after solar maximum [Gonzalez et al., 1990], the other states that the second geomagnetic maximum will occur only 18-24 months after solar maximum [K. Schlegel, 2001].  Our data shows that the second geomagnetic maximum is closer to the second theory which is; the second geomagnetic maximum happens within two years of the absolute solar maximum.
A main goal of the project was to verify the maxima at the equinoxes.  To verify this we chose to create histograms, because they count the number of elements that fit into the given parameters.  Essentially this graph shows that in September there were nearly 300 Kp elements that were equal to or greater than 6.  Geomagnetic storms with Kp levels 6 and above are considered fairly large, but they occur fairly often yielding a large data set.  This graph is run over the entire 52 years, which is about five solar cycles.  The conclusion we came to from this graph was that the most geomagnetic storms with Kp 6 and above occur in September followed by March and April.  By doing some plots that zoomed in on September and March we found that there were clear maximums within a few days of the actual equinoxes.
This graph shows a histogram of Kp elements greater than 8.  These storms are very large and fairly rare.  As you can see by the graph there is a peak in between the maximums in September and March/April.  This maximum is in July.  This was a very interesting discovery because there was no mention of this before we started the research.  We decided to find some papers on this topic.  We felt it was vital to do some research into what is causing this peak and what people have discovered about this July maximum.  After completing some research we realized that very few people even noticed the maximum in July and no one has explained it yet.  The maximum in July was an incredibly interesting discovery and turned into the main focus of our work.  This graph also shows a very interesting minimum.  This minimum is seen during December and January.  In most previous research very little is done about minima, that is one reason this minimum is so interesting. 
This graph to the left shows the entire year in three day interval averages.  It is a histogram of geomagnetic storms with Kp levels greater than 8.  The red lines show the equinoxes, and it shows how close the maximums are to exact equinox.  Also this graph shows the clear maximum in July.  Equinoctial maximums were discovered nearly a century ago [Cortie, 1912].  The exact dates of these maximums and the exact reason that this occurs is a large debate.  Several theoretical models [Cortie, 1912; Russell and McPherron, 1973; McIntosh, 1959; Lyatsky, 2001; Berthelier, 1976; Green, 1984; Orlando, 1995; Clua de Gonzalez et al., 1993; Cliver et al., 2000] have attempted to explain them, however all of them differ in their predictions of which days should be the maxima. 
This graph was created over BHM or universal time.  It is a histogram for geomagnetic  storms with Kp greater than 6.  This graph clearly shows a maximum between 18 UT (1800 BHM on the graph) and 3 UT (300 BHM on the graph).  It also shows a minimum around 9 UT.  This variation is most interesting because Kp is a planetary index, which means that the local time should make no difference.  It may be night time for some of the magnetometers but it may be day time for others. We believe there must be a physical mechanism at work here. 


   Berthelier, A., Influence of the polarity of the interplanetary magnetic field on the annual and diurnal variations of magnetic activity, J. Geophys. Res., 81, 4546, 1976
   Cliver, E. W., Y. Kamide, and A. G. Ling, Mountains versus valleys: Semiannual variation of geomagnetic activity, J. Geophys. Res., 105, 2413, 2000
   Clua de Gonzalez, A. L., W. D. Gonzalez, S. L. G. Dutra, and B.T. Tsurutani, Periodic variation in the geomagnetic activity: A study based on the Ap index, J. Geophys. Res., 98, 9215, 1993
   Cortie, A. L., Sunspots and terrestrial magnetic phenomena, 1898-1911, Mon. Notic. Roy. Astonom. Soc., 73, 52, 1912
   Gonzalez, W. D., A. L. C. Gonzalez, and B.T. Tsurutani, Dual-peak distribution of intense geomagnetic storms, Planet. Space Sci., 38, 181-187, 1990
   Lyatsky, W., P. T. Newell, and A. Hamza, Solar illumination as cause of the equinoctial preference for geomagnetic activity, Geophys. Res. Lett., 28, 12, 2353-2356, 2001
   McIntosh, D. H., On the annual variation of magnetic disturbances, Phil. Trans. Roy. Soc. London, Ser. A, 251, 525, 1959
   Russell, C. T., and R. L. McPherron, Semi-annual variation of geomagnetic activity, J. Geophys. Res., 78, 92, 1973
   Orlando, M., G. Moreno, M. Parisi, and M. Storini, Diurnal modulation of the geomagnetic activity induced by the southward component of the interplanetary magnetic field, J. Geophys. Res., 100, 19, 565, 1995
   Schlegel K., The strongest geomagnetic storms of the last century, Radio Science Bulletin, 298, September, 2001

Download Seth's Power Point Presentations Here:
Haystack Presentation
School Presentation
Haystack Home Page     Contact Us     Links     Project Home Page     Acknowledgments
Web Page Created 06-03-02