Wavelet Transform And Its Applications To The Analysis Of Solar Microwave Spikes

The wavelet analysis is a kind of efficient multi-scale methods. It has been wildly used in signal processing and image manipulation for its many good properties. This thesis mainly studies the applications of wavelet transform to the analysis of solar microwave spikes observed. The major works can be summarized as follows:The history of wavelet transform and basic theory of wavelet analysis are indicated. Also the solar radio research field as well as the radio spectrometers is introduced. Radio spike is an important kind of the fine structures in solar radio bursts. The most salient feature of spikes is their very narrow bandwidth. Bandwidths of spikes set the most severe requirements on the emission process and limit the source size. In the violent solar bursts at the end of October 2003, two groups of spikes were recorded by the 1.08-2.04GHz spectrometers of National Astronomical Observatories of China (Beijing) on 27 October 2003. The mean and minimum bandwidths of spikes are calculated with two different methods based on the wavelet analysis. The first method is a kind of feature detection algorithms based on the combination of wavelet filter and fitting an assumed spike profile-gauss curve into the denoised data. The second method determines spike bandwidth scale based on their scalegram. This allows measuring the scale bandwidth independently of manual spike selection criteria. The smallest relative FWHM bandwidth of spikes was found at 0.38%, 0.34%. This result is consistent with the conclusion of Messmer et al. At last, mean power spectra and mean spatial power spectra of two events are computed by use of Fourier transform. The mean peak fluxes of spikes are also calculated and the relationships between mean power spectra/mean peak flux and their bandwidth are analyzed.