Wavelet Analysis And Its Applications To The Processing Of Solar Radio Data Observed

This thesis mainly studies the applications of wavelet transform to the processing of solar radio data observed. The major works can be summarized as follows:Based on a method with the combination of channel-modification and wavelet filter, we recognize the frequency band of fine structures and delete all the interference that is shown as grid chart superposed on the dynamic spectra. These girding is often considered that result from the apparatus function of receiving device and the noise harmonics.In a layered model of definite solar atmosphere distance, eight events of spike emission, observed by the Beijing radio spectrograph in the 2.6-3.8GHz frequency range, are analyzed. Firstly, let us transform the frequency scale of the radio spectra to the scale of distances in the solar atmosphere. After we gain the 2-D data of time versus distance, the spatial power spectra at a fixed moment can be computed by use of wavelet transform and Fourier transform respectively. Then we averaged the power spectra over the corresponding time intervals, the resulting time-averaged 1-D spatial power spectra were drawn and then performed a least-square linear fit. We find the power-law indices are scattered in a broad interval of values from –0.66 up to –2.85. In some cases they considerably deviate from the value of –5/3 ,resembling that of Kolmogrov spectra of turbulent cascades,but were consistent with the result of Karlicky in the 1-2GHz frequency range.A technique for calculating solar radio polarization is put forward. And the outlines of polarization of these eight events above have been shown. The important step of this new method is how to estimate the burst background logically. Here we introduce two typical ways, one is the wavelet transform, the other is the extended interpolation method .Our last results indicate that this method is simple and reliable.