Singular spectrum analysis as a tool for plasma fluctuations analysis

We describe the application of the singular spectrum analysis (SSA), an advanced adaptive statistical technique, to denoise experimental signals and to remove trends in order to isolate the relevant fluctuating components. We illustrate a fast denoising algorithm and show its performances relatively to synthetic and experimental signals of the soft x-ray (SXR) spectrometer and the polarimeter installed in the reversed field experiment (RFX) device. As a further application, we report a first estimate of the electron temperature fluctuations in the core of the RFX experiment. They have been performed with a multifilter SXR spectrometer and applying SSA for the first time in plasma physics the singular spectrum analysis. We find that temperature fluctuations are typically not larger than a few percent and are well correlated with magnetic fluctuations. (C) 2001 American Institute of Physics.