O.V. Mandrikova, Yu.A. Polozov, A.S. Perezhogin

The subject of the present paper and related research are ionospheric parameters and, in particular, discovering local abnormal features connected with solar activity or arising during the periods of the increased seismic activity. These abnormal features contain important information for a researcher and during signal processing there is a problem of their identification. In the given paper we have suggested a new technique aimed at studying time features of parameters of the ionosphere and numerical decisions based on wavelet-transform have been introduced. They make it possible to extract local structures in data in an automatic mode. In this paper we have suggested numerical solutions to the automatic extraction of abnormal features and their further analysis, which includes transforming data into the wavelet space, extracting informative time-frequency intervals in data and analyzing local structures forming the process.