Wavelet analysis of CME, X-ray flare, and sunspot series

Context. Coronal mass ejections (CMEs) and solar flares are the most energetic transient phenomena taking place at the Sun. Together they are principally responsible for disturbances in outer geospace. Coronal mass ejections and solar flares are believed to be correlated with the solar cycle, which is mainly characterized by sunspot numbers. Aims. Here, we search for pattern identification in CMEs, X-ray solar flares, and sunspot number time series using a new data mining process and a quantitative procedure to correlate these series. Methods. This new process consists of the combination of a decomposition method with the wavelet transform technique applied to the series ranging from 2000 until 2012. A simple moving average is used for the time-series decomposition as a high-pass filter. A continuous wavelet transform is applied to the series in sequence, which permits us to uncover signals previously masked by the original time series. We made use of the wavelet coherence to find some correlation between the data. Results. The results have shown the existence of periodic and intermittent signals in the CMEs, flares, and sunspot time series. For the CME and flare series, few and relatively short time intervals without any signal were observed. Signals with an intermittent character take place during some epochs of the maximum and descending phases of the solar cycle 23 and rising phase of solar cycle 24. A comparison among X-ray flares, sunspots, and CME time series shows a stronger relation between flare and CMEs, although during some short intervals (four-eight months) and in a relatively narrow band. Yet, in contrast we have obtained a fainter or even absent relation between the X-ray flares and sunspot number series as well as between the CMEs and sunspot number series.