Search of predictors of geoeffective heliospheric events by means of muon hodoscope URAGAN

The flux of charged primary cosmic rays (PCR) is modulated by changes of interplanetary magnetic field which depend on various heliospheric processes. After interaction of PCR with nuclei of atoms of the Earth's atmosphere, these modulations are transferred to a flux of secondary muons. Muon hodoscope URAGAN allows to trace changes not only of intensity of the muon flux, but also its angular distribution in a wide range of zenith angles (0-80°). Some results of searching of predictors of geoeffective heliospheric events by means of muon hodoscope URAGAN are presented.


Introduction
The flux of charged primary cosmic rays (PCR) is modulated by changes of interplanetary magnetic field which depend on various heliospheric processes. The Earth magnetosphere is substantially subject to influence of dynamic processes in the interplanetary environment: strong disturbances of the interplanetary magnetic field result in reorganization of the magnetic field inside magnetosphere, to development of geomagnetic storms and substorms. Disturbances of the interplanetary magnetic field are propagated from the Sun to the Earth with a speed of the solar wind 300-1500 km•s -1 , but the velocity of propagation of PCR is close to the light velocity. The PCR flux modulated by the disturbance reaches the near-Earth space considerably earlier than this disturbance. After interaction of PCR with nuclei of atoms of the Earth's atmosphere, these modulations are transferred to the flux of secondary muons. Muon hodoscope URAGAN (MH URAGAN) [1] allows to trace changes not only of the intensity of the muon flux, but also its angular distribution in a wide range of zenith angles (0-80º). The results of a search of predictors of geoeffective heliospheric processes with the help of the analysis of variations of the muon flux are considered.

Initial data
As the initial data for searching the predictors, the URAGAN data and OMNI database were used.
At preparation of the data of MH URAGAN in real time, which are presented at the site of the Unique Scientific Facility 'Experimental complex NEVOD' [2], time series of hourly average values and their statistical errors for various characteristics of a registered muon flux are formed:  Isum -count rate with barometric correction, s -1 ;  A Z -vertical projection of the vector of the local anisotropy with barometric correction;  A S -horizontal projection of the vector of the local anisotropy to the north-south axis;  A E -horizontal projection of the vector of the local anisotropy to the west-east axis.
The secondary characteristics are formed concerning the average values obtained during 24 hours:  r h -horizontal projection of the vector of the relative anisotropy;  r S -horizontal projection of the vector of the relative anisotropy to the north-south axis;  r E -horizontal projection of the vector of the relative anisotropy to the west-east axis.
Among the characteristics of the muon flux, the most convenient are those which have no long-term trends. But variations of characteristics because of the daily rotation of the Earth, in our case cannot be neglected. Therefore the following characteristics were examined: r h , r S , r E . From the time series of hour values r h . r S , r E , the following secondary quantities (here symbol X designates r h or r S , or r E ) were calculated: 1) δX 1 -the ratio of the difference of the current and the preceding hour values of the characteristic X to the statistical error of this difference; 2) δX 24 -the ratio of the difference of the current and 24 hours prior to it hour values of the characteristic X to the statistical error of this difference; 3) σX 24 -root-mean-square deviation of the hour values of the characteristic X over the period 24 hours (including the current hour).
As indicators of the condition of the heliosphere and magnetosphere of the Earth, the following parameters of the interplanetary magnetic field (IMF) and geomagnetic activity from database OMNI were used:  Vsw -hourly average value of the speed of the solar wind, km•s -1 ;  BzGSE -hourly average value of z-projection of IMF in GSE coordinate system, nT;  Dst -hourly average equatorial index of the geomagnetic activity, nT;  Kp -the planetary geomagnetic activity index.
For the analysis, the URAGAN data and OMNI data from February, 2007 till December, 2013 were used. The following criteria of the perturbed state in OMNI data were used: Vsw ≥ 500 km•s -1 , BzGSE ≤ -7 nT, Dst ≤ -30 nT, Kp ≥ 4.

Searching predictors
Joint processing of time series of the URAGAN and OMNI data in a moving time window of 5 day was carried out. First for consideration the intervals in which in the first 4 days there were no disturbances, and on the last day Kp ≥ 4 have been taken. As there were found only 65 such intervals, for the increase in statistics the intervals in which in preceding four days there could be disturbances on Vsw, BzGSE, Dst have been added, too. Thus, for the visual analysis, 130 5-day intervals in which the last day value Kp ≥ 4 was observed, have been selected.
In the selected intervals the behaviour of horizontal projections of the vector of the relative anisotropy (r h , r S , r E ) and their derivative quantities were analyzed (δr h1 , δr S1 , δr E1 , δr h24 , δr S24 , δr E24 , σr h24 , σr S24 , σr E24 ). During the analysis, those projections for which in the first two days of the examined interval their values were appreciably smaller, than in the subsequent two, or qualitative by different, were marked. In figure 1, examples of the images of time series prepared for the visual analysis are shown.