Time-frequency analysis of sunspot numbers using synchrosqueezing transform
Lan Yu, Cao Yu, Feng Song · 2017
An accurate time-frequency representation can provide important physical information about solar oscillation and quasi-periodic signal. The traditional methods, such as fast Fourier transform, short-time Fourier transform and continuous wavelet transform, have limitations when distinguishing the instantaneous frequencies of the solar oscillation and quasi-periodic signal. The Synchrosqueezing transform (SST) is a promising tool that provide a more accurate time-frequency representation. We used two synthetic signals to illustrate that the SST is more efficient and accurate than those traditional methods in time-frequency representation. Subsequently, the period of the sunspot numbers were analyzed using the SST. The result demonstrates that the SST differentiates more accurate frequency components to the continuous wavelet transform.