In this talk we review recent developments in our understanding of the solar sources of space weather, i.e. solar flares, coronal mass ejections (CMEs) and associated phenomena. In particular, we will focus on on-disk signatures of the impulsive dynamics of CMEs, flares and associated global waves. Recent results and the new possibilities offered by STEREO (Solar Terrestrial Relations Observatory) will be discussed.

Launched in October 2006, the NASA STEREO mission has now completed more than 18 months of science operations. The twin spacecraft are moving away from the Earth along Earth-like orbits, one ahead of the planet, the other lagging behind. Mounted on the side of each spacecraft, the UK-built Heliospheric Imagers (HI) contain wide- field, visible light cameras that are capable of continuously tracking these events from just off the solar limb to distances of over 1 AU. From their unique positions the HI instruments are able to look back along the Sun-Earth line and image Earth-impacting mass ejections as well as the formation and propagation of co- rotating interaction regions. The UK team have been developing techniques to determine the speed and position of these solar-wind transients from HI data, proving the worth of such measurements to space-weather monitoring.

Recent advances in the understanding of the activity level (incorporating flares and coronal mass ejections) of active regions are presented. A combination of turbulence and complexity studies, using advanced multiscale and multifractal methods) is used to studying the flaring rate of active regions over the last solar cycle. This is combined with a general overview of the success rate of a number of existing flare prediction tools.

We have developed a software tool that automatically detects coronal mass ejections (CMEs) in coronagraphic images. The CACTus tool uses feature recognition methods to identify the starting time, radial velocity and angular extend of the CMEs. It is written in IDL and available through the open source software library "Solar Soft".

CACTus has been applied on all SOHO/LASCO data throughout solar cycle 23, which resulted in an online catalog of all detected CMEs with their characteristics. By coding the 'definition of a CME' in the program, we can confirm that the CME detection criteria are stable over a time span as long as a solar cycle. We will show that this is a considerable advantage over catalogs assembled by human operators. CACTus has been equally applied on the STEREO/SECCHI data available since the beginning of 2007.

In the present paper we want to stress the use of CACTus in the context of near-real-time space weather monitoring. Earth-directed CMEs are the prime cause of major geomagnetic storms and their immediate detection is thus essential for early warnings. CACTus has become part of the IT-infrastructure of the SIDC in Brussels as Regional Warning Center.

We have set-up a version of CACTus detecting in near real time CMEs on incoming LASCO data. The real-time CACTus puts its output on the internet where it is available to all Major events (CMEs with an angular span > 90 degrees) are reported 24h/24h through e-mail to a list of registered SIDC users.

For more information, please visit: http://sidc.be/cactus

Solar radio weather is characterized by a complex variety of radio emission features. Under specific physical conditions, high radio flux density levels and circular polarization characteristics are observed in the decimetric band, capable to interfere Wireless Communication Systems and Global Navigation Satellite Systems.

To properly characterize such solar radio weather events in order to emphasize the aspects relevant to the interfering effectivity an adequate analysis is mandatory.

In this work, we consider selected solar radio outbursts observed in the past solar cycle by the Trieste Solar Radio System and associated with communication jamming to point out the nature and relevance of parametric descriptors for a full characterization of their radioeffectivity.

The comparative analysis of radio descriptors shows that the total time of flux density threshold excess and the time distribution of threshold excess as well as the circular polarization mode are key issues in this framework.

High energy particles from the Sun are clearly an important element of solar effects on spacecraft and aircraft. The ability to predict transient enhancements of particle fluxes is therefore of major interest. The question when and how particles are accelerated to high, sometimes relativistic, energies at the Sun is also of fundamental astrophysial interest, because the proximity of the Sun allows us to observe the partciles by remote sensing and in situ techniques, to resolve the time evolution of the particle fluxes, and also to map the dynamic environment where they are accelerated in the solar corona.

In the 1990s the idea emerged that all solar energetic particle events with space weather relevance are accelerated at shocks driven by coronal mass ejetions (CME). It made its way into recent documents on the assessment of radiation risks for spacecraft. In this review the statistical evidence for a relationship of solar energetic particle (SEP) events with CMEs and flares will be discussed, and it will be argued that both CMEs and flares - i.e. large-scale magnetic restructuring and particle acceleration in active regions - are necessary ingredients for SEP events at the Earth. Statistical studies do not allow us to identify one or the other of the two coronal phenomena as the nt accelerator of the escaping particles. Timing studies of the release of SEP will then be addressed in an attempt to find clearer criteria. The conclusion is that as of today we cannot clearly identify the accelerator over the whole energy range observed in major SEP events, and that new diagnostic tools, including especially SEP measurements from a vantage point well within 1 AU, are necessary to get more insight into the intricate relationship between high energy particles, small scale processes in flares, and large scale coronal restructuing in CMEs.

Coronal mass ejections (CMEs) are large-scale solar eruptions during which the magnetic flux of some 1023 Wb is launched into interplanetary space at velocities in the order of 1000 km s-1, carrying along 1011-1013 kg of coronal plasma. Since the Earth-directed CMEs, and the shocks they drive, are the main source of major geomagnetic storms, the prediction of their arrival times is one of central issues of the Space Weather forecasting. From the empirical studies we know that the Sun-Earth transit times of CMEs depend on their take-off (coronal) speed, size, and mass, as well as on the ambient solar wind speed. These relationships provide the so-called empirical (statistical) forecasting. On the other hand, observations reveal that after the CME take-off, which is governed by the Lorentz force, the CME dynamics becomes strongly affected by the interaction of the erupting structure with the ambient magnetoplasma -- eruptions that are faster than solar wind transfer the momentum and energy to the wind and generally decelerate, whereas slower ones gain the momentum and accelerate. Such a behavior can be expressed in terms of "aerodynamic" drag, which provides the so-called kinematical forecasting, based on solving a relatively simple equation of motion. Finally, a rapid progress in numerical MHD modeling opened a possibility for the most sophisticated forecasting, which employs the complete set of MHD equations to describe the global state of the solar wind and its coupling with the propagating CME. Advantages and drawbacks of these three methods are reviewed and future prospects are discussed.

A brief review is made of user needs for space weather data, products and services, and relationships with solar observation and physics. Some examples of current practice will be given. Several studies have been made of user requirements over the last few years and these indicate ways in which application-oriented measurements may differ from those made for scientific reasons. These include questions of resolution, timeliness and continuity. On the other hand, measurements made to do science can have important uses in applications. Work towards understanding of solar and heliospheric processes also leads to development of modelling and data treatment techniques that can address important user needs and so can be the subject of "transitioning" to space weather products and services. This review will try to identify areas where collaboration between the solar physics research and applications communities in Europe could bring improvements to space weather products and services.

Solar weather is driven by the solar magnetic activity. Differential topological studies of vector fields, using Poincaré-Hopf theorem for manifolds, may describe the solar magnetic state. Helicity, which has been applied to many coronal mass ejection and loop studies, is related to the Gauss linking number. Solar climate may be studied topologically with dynamic systems. A more precise description of the solar weather state is foreseen not only to give better understanding but also better predictions. Neural networks are planned to be trained with these topological descriptions. We will give a progress report of our topological studies.

Like all natural hazards, space weather exhibits occasional extreme events over timescales of decades to centuries. These events are highly significant because they have the potential to disrupt many technological infrastructures that underpin modern civilisation. However, like all extreme hazards, such events are rare, so we have limited data on which to build our understanding of the events. This limitation is even more serious for space weather since it is a global phenomenon. Most other natural hazards (e.g. flash floods) are highly localised, so statistically significant datasets can be assembled by combining data from independent instances of the hazard recorded over a few decades. Such datasets are the foundation on which reliable risk assessment methodologies are built. But we have a single instance of space weather, so we would have to make observations for many centuries in order to build a statistically significant dataset. Thus it is not practicable to assess the risk from extreme events using simple statistical methods. Instead we must exploit our knowledge of solar-terrestrial physics to better understand the factors that can cause extreme space weather events. We can then study how each factor varies and thus assess the risk that they come together in a dangerous combination.