Physics of auroral phenomena : proceedings of the 34th Annual seminar, Apatity, 01 - 04 March, 2011 / [ed.: A. G. Yahnin, A. A. Mochalov]. - Апатиты : Издательство Кольского научного центра РАН, 2011. - 231 с. : ил.

“P hysics o f Auroral Phenomena", Proc. XXXIV Annual Seminar, A patity, pp. 3 9 - 42, 2011 © Kola Science Centre, Russian Academy of Science, 2011 POST-PROCESSING OF WAVELET TRANSFORMS FOR MAGNETOGRAMS RECEIVED ALONG THE NORTHERN PART OF THE GEOMAGNETIC MERIDIAN AS A NEW TOOL FOR CLASSIFICATION AND PREDICTION OF MAGNETOSPHERIC STORMS N.A. Barkhatov1’2, S.E. Revunov1, D.V. Shadrukov1 1. Nizhniy Novgorod State Pedagogical University, Nizhniy Novgorod, Russia 2. Nizhny Novgorod State University o fArchitecture and Civil Engineering, Nizhniy Novgorod, Russia Abstract. Research to analyzing the results of the post-processing of the wavelet components for horizontal component of the geomagnetic field perturbation in range Pc4-5 recorded along the meridian chain of stations during different strength of geomagnetic storms is devoted. Concludes that the intensity of the expected or rising geomagnetic storm. Analysis of spatial and temporal spectrum distribution of geomagnetic variations to the existing network of geomagnetic stations allows us to study the dynamics of displacement o f polar cap and auroral oval boundary during magnetospheric disturbances. The proposed method for classification and forecast geomagnetic storms can be used. Introduction Spectral processing of satellite and ground observations for study Solar-Terrestrial relationships with the accumulation of information about the state of the interplanetary medium and geomagnetic field has been advanced. Spectral and wavelet analysis of satellite and Earth-surface data at different stages of magnetospheric storms have allowed groups of the Institute of Earth Physics and Space Research Institute a number of features of Solar- Terrestrial relations [Pilipenko et al., 1999; Kozyreva et al, 2006] to establish. However, there is difficulty interpreting the results obtained using these techniques. Of particular difficulty the analysis of the wavelet patterns is represented. It is connected by that the technics of wavelet transformation as a result of calculation gives a superfluous information picture of a spectrum. The essence of our proposed post-processing is focused to reporting the results as graphs or local maxima of wavelet skeletons, allowing you the key features of the spectra dynamics present. The data used and processing techniques In study minutes data for geomagnetic field components from a network of stations IMAGE (http://www.geo.fmi.fi/image) is used. Intervals of station data with intervals of dayside (for a network of stations), magnetospheric storms with shape of a classic «bay» and with varying intensity on the indication of the Dst-index (http://spidr.ngdc.noaa.gov/spidr/) recorded between 2000 to 2003 were collected. A total of 12 intervals with 72 hours duration were found. In this intervals weak and moderate storms (Dst> -100 nT), severe storms (Dst> -200 nT) and extreme storms (Dst <-200 nT) were collected. All intervals contain the key stages in the expansion of magnetospheric storms - sudden commencement, main phase and the beginning of the recovery phase. For the analysis of geomagnetic disturbances 7 stations (Homsund, Bear Island, Kevo, Masi, Sodankyla, Oulujarvi, Uppsala) located at approximately the same geomagnetic meridian (106-109) were selected. Data processing by the wavelet transform and the post-processing results of the wavelet patterns are performed. The basic wavelet in numerical simulations Doubechies fourth-order function has been chosen. Scaling coefficients of the wavelet transform in the range of 4 to 15, which corresponds to the range studied oscillation periods of 120 seconds (Pc4) to 450 seconds (Pc5) are used. Example of wavelet processing technique for fragment of the magnetogram corresponding to the geomagnetic storm main phase recorded at the station Kevo 18.04.2001 in Fig. la are shown. Here, the bright areas correspond to high correlation of the signal with the base wavelet, dark areas correspond to low correlation. It is known that in a calculation technique of wavelet transforms is superfluous spectral information. The redundancy of this representation can be intensely reduced when only the maxima of the spectrum on the graph are shown. This forms the «wavelet skeleton» or «skeletons» are called, which consists of a set of curved lines that track the position of local maxima in the spectrum. Thus, in place of high values for coefficients in the wavelet pattern «skeleton» points are located. This technique for presentation results like «skeleton» of the wavelet pattern lets you focus on the key features of the dynamics of the spectra. On Fig. lb an example of the spectrum shown in Fig. la in the «wavelet skeleton» form so-called local maxima graph is shows. 39 Polar Geophysical Institute