We study the clustering properties of fast Coronal Mass Ejections (CMEs) that occurred during solar cycles 23 and 24. We apply two methods: the Max spectrum method can detect the predominant clusters and the de-clustering threshold time method provides details on the typical clustering properties and time scales. Our analysis shows that during the different phases of solar cycles 23 and 24, CMEs with speed $\geq 1000\ km/s$ preferentially occur as isolated events and in clusters with on average two members. However, clusters with more members appear particularly during the maximum phases of the solar cycles. Over the total period and in the maximum phases of solar cycles 23 and 24, about 50% are isolated events, 18% (12%) occur in clusters with 2 (3) members, and another 20% in larger clusters $\geq 4$, whereas in solar minimum fast CMEs tend to occur more frequently as isolated events (62%). During different solar cycle phases, the typical de-clustering time scales of fast CMEs are $\tau_c=28-32\ hrs$, irrespective of the very different occurrence frequencies of CMEs during solar minimum and maximum. These findings suggest that $\tau_c$ for extreme events may reflect the characteristic energy build-up time for large flare and CME-prolific active ARs. Associating statistically the clustering properties of fast CMEs with the Disturbance storm index Dst at Earth suggests that fast CMEs occuring in clusters tend to produce larger geomagnetic storms than isolated fast CMEs. This may be related to CME-CME interaction producing a more complex and stronger interaction with the Earth magnetosphere.

We apply a nonlinear mean-field dynamo model which includes a budget equation for the dynamics of Wolf numbers to predict solar activity. This dynamo model takes into account the algebraic and dynamic nonlinearities of the alpha effect, where the equation for the dynamic nonlinearity is derived from the conservation law for the magnetic helicity. The budget equation for the evolution of the Wolf number is based on a formation mechanism of sunspots related to the negative effective magnetic pressure instability. This instability redistributes the magnetic flux produced by the mean-field dynamo. To predict solar activity on the time scale of one month we use a method based on a combination of the numerical solution of the nonlinear mean-field dynamo equations and the artificial neural network. A comparison of the results of the prediction of the solar activity with the observed Wolf numbers demonstrates a good agreement between the forecast and observations.

Mutual quasi-periodicities near the solar-rotation period appear in time series based on the Earth's magnetic field, the interplanetary magnetic field, and signed solar-magnetic fields. Dominant among these is one at 27.03 +/- 0.02 days that has been highlighted by Neugebauer, et al. 2000, J. Geophys. Res., 105, 2315. Extension of their study in time and to different data reveals decadal epochs during which the ~ 27.0 day, a ~ 28.3 day, or other quasi-periods dominate the signal. Space-time eigenvalue analyses of time series in 30 solar latitude bands, based on synoptic maps of unsigned photospheric fields, lead to two maximally independent modes that account for almost 30% of the data variance. One mode spans 45 degrees of latitude in the northern hemisphere and the other one in the southern. The modes rotate around the Sun rigidly, not differentially, suggesting connection with the subsurface dynamo. Spectral analyses yield familiar dominant quasi periods 27.04 +/- 0.03 days in the North and at 28.24 +/- 0.03 days in the South. These are replaced during cycle 23 by one at 26.45 +/- 0.03 days in the North. The modes show no tendency for preferred longitudes separated by ~ 180 degrees.

Principal Components Analysis (PCA) and Independent Component Analysis (ICA) are used to identify global patterns in solar and space data. PCA seeks orthogonal modes of the two-point correlation matrix constructed from a data set. It permits the identification of structures that remain coherent and correlated or which recur throughout a time series. ICA seeks for maximally independent modes and takes into account all order correlations of the data. We apply PCA to the interplanetary-magnetic-field polarity near one AU and to the 3.25R source-surface fields in the solar corona. The rotations of the two-sector structures of these systems vary together to high accuracy during the active interval of solar cycle 23. We then use PCA and ICA to hunt for preferred longitudes in northern hemisphere, Carrington maps of magnetic fields.

Observations show that solar activity is distributed non-axisymmetrically, concentrating at "preferred longitudes". This indicates the important role of non-axisymmetric magnetic fields in the origin of solar activity. We investigate the generation of the non-axisymmetric fields and their coupling with axisymmetric solar magnetic field. Our kinematic generation (dynamo) model operating in a sphere includes solar differential rotation, which approximates the differential rotation obtained by inversion of helioseismic data, modelled distributions of the turbulent resistivity, non-axisymmetric mean helicity, and meridional circulation in the convection zone. We find that (1) the non-axisymmetric modes are localised near the base of the convection zone, where the formation of active regions starts, and at latitudes around $30^{\circ}$; (2) the coupling of non-axisymmetric and axisymmetric modes causes the non-axisymmetric mode to follow the solar cycle; the phase relations between the modes are found. (3) The rate of rotation of the first non-axisymmetric mode is close to that determined in the interplanetary space.

One of the key questions in solar physics that remains to be answered concerns the strength and the distribution of the magnetic fields at the base of the convection zone. The flux tube dynamics requires that toroidal fields of strength as large as 100 kilogauss be present at the base of the convection zone. The kinetic-magnetic equipartition argument leads to smaller field strengths. For possible detection of these relatively small (compared to pressure effects) fields by helioseismic methods it is important to know the range of the field strengths and their distribution. We estimate a range for the toroidal magnetic field strengths at the base of the convection zone using dynamo simulations in a spherical shell. These simulations involve the distribution of rotation provided by helioseismic inversions of the GONG and MDI data. Combining the simulations with the observed line-of-sight surface poloidal field we extract the spatial pattern and magnitude of the mean toroidal magnetic field at the base of the convection zone.

Below the scale of supergranules we find that cellular flows are present in the solar photosphere at two distinct size scales, approximately 2 Mm and 4 Mm, with distinct characteristic times. Simultaneously present in the flow is a non-cellular component, with turbulent scaling properties and containing 30% of the flow energy. These results are obtained by means of wavelet spectral analysis and modeling of vertical photospheric motions in a 2-hour sequence of 120 SOHO/MDI, high resolution, Doppler images near disk center. The wavelets permit detection of specific local flow patterns corresponding to convection cells.

The Sun provides an excellent natural laboratory for nonlinear phenomena. We use motions of magnetic bright points on the solar surface, at the smallest scales yet observed, to study the small scale dynamics of the photospheric plasma. The paths of the bright points are analyzed within a continuous time random walk framework. Their spatial and temporal scaling suggest that the observed motions are the walks of imperfectly correlated tracers on a turbulent fluid flow in the lanes between granular convection cells.