It is now apparent that there are at least two heating mechanisms in the Sun's outer atmosphere, or corona. Wave heating may be the prevalent mechanism in quiet solar periods and may contribute to heating the corona to 1,500,000 K (refs 1-3). The active corona needs additional heating to reach 2,000,000-4,000,000 K; this heat has been theoretically proposed to come from the reconnection and unravelling of magnetic 'braids'. Evidence favouring that process has been inferred, but has not been generally accepted because observations are sparse and, in general, the braided magnetic strands that are thought to have an angular width of about 0.2 arc seconds have not been resolved. Fine-scale braiding has been seen in the chromosphere but not, until now, in the corona. Here we report observations, at a resolution of 0.2 arc seconds, of magnetic braids in a coronal active region that are reconnecting, relaxing and dissipating sufficient energy to heat the structures to about 4,000,000 K. Although our 5-minute observations cannot unambiguously identify the field reconnection and subsequent relaxation as the dominant heating mechanism throughout active regions, the energy available from the observed field relaxation in our example is ample for the observed heating.

Solar flares, coronal mass ejections, and indeed phenomena on all scales observed on the Sun, are inextricably linked with the Sun's magnetic field. The solar surface is covered with magnetic features observed on many spatial scales, which evolve on differing timescales: the largest features, sunspots, follow an 11-year cycle; the smallest seem to follow no cycle. Here, we analyze magnetograms from Solar and Heliospheric Observatory (SOHO)/Michelson Doppler Imager (full disk and high resolution) and Hinode/Solar Optical Telescope to determine the fluxes of all currently observable surface magnetic features. We show that by using a "clumping" algorithm, which counts a single "flux massif" as one feature, all feature fluxes, regardless of flux strength, follow the same distribution-a power law with slope −1.85 ± 0.14-between 2 × 10 17 and 10 23 Mx. A power law suggests that the mechanisms creating surface magnetic features are scale-free. This implies that either all surface magnetic features are generated by the same mechanism, or that they are dominated by surface processes (such as fragmentation, coalescence, and cancellation) in a way which leads to a scale-free distribution.

On 1996 March 7, the Solar and Heliospheric Observatory spacecraft conducted a multi-instrument campaign to observe polar plumes in the south polar coronal hole. Recent time-domain analyses of EUV Imaging Telescope images from that campaign show filamentary substructure in the plumes, on a length scale of ∼5Љ, which changes on timescales of a few minutes, and coherent quasi-periodic perturbations in the brightness of Fe ix and Fe x line emission at 171 Å from the plumes. The perturbations amount to 10%-20% of the plumes' overall intensity and propagate outward at 75-150 km s Ϫ1 , taking the form of wave trains with periods of 10-15 minutes and envelopes of several cycles. We conclude that the perturbations are compressive waves (such as sound waves or slow-mode magnetosonic waves) propagating along the plumes. Assuming that the waves are sonic yields a mechanical energy flux of 150-400 W m Ϫ2 (1.5-ergs cm Ϫ2 s Ϫ1 ) in the plumes.

Abstract. A complete halo coronal mass ejection (CME) was observed by the SOHO Large-Angle and Spectrometric Coronagraph (LASCO) coronagraphs on May 12, 1997. It was associated with activity near Sun center, implying that it was aimed earthward. Three days later on May 15 an interplanetary shock and magnetic cloud/flux rope transient was detected at the Wind spacecraft 190 RE upstream of Earth. The long enduring southward magnetic fields associated with these structures triggered a geomagnetic storm. The CME was associated with a small coronal arcade that formed over a filament eruption with expanding double ribbons in Hot emission. The flare was accompanied by a circular EUV wave, and the arcade was flanked by adjacent dimming regions. We surmise that these latter regions marked the feet of a flux rope that expanded earthward into the solar wind and was observed as the magnetic cloud at Wind. To test this hypothesis we determined key parameters of the solar structures on May 12 and compared them with the modeled flux rope parameters at Wind on May 15. The measurements are consistent with the flux rope originating in a large coronal structure linked to the erupting filament, with the oppositepolarity feet of the rope terminating in the depleted regions. However, bidirectional electron streaming was not observed within the cloud itself, suggesting that there is not always a good correspondence between such flows and ejecta.