Model and algorithms for geomagnetic data analysis in the tasks of detection of geomagnetic disturbances and calculation of geomagnetic activity index

• download

   This work is dedicated to the problem of processing and analysis of the
   Earth magnetic field variations, detection of quiet daily variations and
   local peculiarities formed during high geomagnetic activity. The
   non-stationary character of the process and the presence of
   different-scale local peculiarities of different form and duration, make
   the traditional methods for data modeling and analysis not effective and
   do not allow to:

   - identify short-period fluctuations in the data during high geomagnetic
   activity;

   - detect "quiet" (characteristic) changes of field variation and to
   estimate its change during storms;

   - develop an automatic algorithm to calculate Sq-curve and K-index by J.
   Bartels method.

   The paper presents a complex multi-component model of geomagnetic signal
   (on the example of H component), based on wavelets, and allowing us to
   describe "quiet" changes of field variation and the different-scale local
   peculierities, formed before and during magnetic storms. Also, a method of
   detection of a characteristic component of the model and short-period
   fluctuations is proposed in the paper. The algorithms and software based
   on this method allow us to automatically detect the diurnal variation, to
   plot Sq-curve and K-index and to identify the peculiarities occurring
   during the increase of geomagnetic activity. Efficiency of the developed
   tools is proved statistically ("Paratunka" station (Kamchatka) data, for
   the period 2002-2008).