Soft-gamma-ray repeaters (SGRs) are galactic X-ray stars that emit numerous short-duration (about 0.1 s) bursts of hard X-rays during sporadic active periods. They are thought to be magnetars: strongly magnetized neutron stars with emissions powered by the dissipation of magnetic energy. Here we report the detection of a long (380 s) giant flare from SGR 1806-20, which was much more luminous than any previous transient event observed in our Galaxy. (In the first 0.2 s, the flare released as much energy as the Sun radiates in a quarter of a million years.) Its power can be explained by a catastrophic instability involving global crust failure and magnetic reconnection on a magnetar, with possible large-scale untwisting of magnetic field lines outside the star. From a great distance this event would appear to be a short-duration, hard-spectrum cosmic gamma-ray burst. At least a significant fraction of the mysterious short-duration gamma-ray bursts may therefore come from extragalactic magnetars.

No abstract

No abstract

We present X-ray imaging and spectral analysis of all microflares the Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) observed between March 2002 and March 2007, a total of 25,705 events. These microflares are small flares, from low GOES C Class to below A Class (background subtracted) and are associated with active regions. They were found by searching the 6-12 keV energy range during periods when the full sensitivity of RHESSI's detectors was available (see paper I). Each microflare is automatically analyzed at the peak time of the 6-12 keV emission: the thermal source size is found by forward-fitting the complex visibilities for 4-8 keV, and the spectral parameters (temperature, emission measure, power-law index) are found by forward fitting a thermal plus nonthermal model. The combination of these parameters allows us to present the first statistical analysis of the thermal and non-thermal energy at the peak times of microflares. On average a RHESSI microflare has a fitted thermal loop width 8 Mm (11 ′′ ), length 23 Mm (32 ′′ ) and volume 1×10 27 cm 3 , temperature 13 MK, emission measure 3 × 10 46 cm −3 and density of 6 × 10 9 cm −3 . There is no correlation between the loop size and the flare magnitude, either flux in the loop or GOES class, indicating that microflares are not necessarily spatially small. There is also no clear correlation between the thermal parameters except between the RHESSI and GOES emission measures, the GOES values are generally twice the RHESSI emission measures. The microflare thermal energy at the time of peak emission in 6-12 keV ranges over 10 26 to 10 30 erg and has a median value of 10 28 erg. The frequency distribution of the thermal energy deviates from a power-law at low and high energies arising from a deficiency of events due to instrumental and selection effects. It is difficult to compare this energy distribution to previous thermal energy distributions of transient events, as the work sought nanoflares through imaging in EUV or soft X-rays and covered just a few hours. There are large uncertainties in the majority of the non-thermal parameters, due to the steep spectra down to low energies. We typically find a power-law index of 7 above a break energy of 9 keV, which corresponds to a low-energy cut-off in the electron distribution as low as 12 keV. The resulting non-thermal power estimates, covering 10 25 to 10 28 erg s −1 with median value of 10 26 erg s −1 , therefore have large uncertainties as well. The few microflares with unexpectedly large non-thermal powers 10 28 erg s −1 have the smallest uncertainties, of about 10%. The total non-thermal energy however is still small compared to that of large flares as it occurs for shorter durations.

We summarize Reuven Ramaty High Energy Solar Spectroscopic Imager (RHESSI) hard X-ray (HXR) and gray imaging and spectroscopy observations of the intense (X4.8) g-ray line flare of 2002 July 23. In the initial rise, a new type of coronal HXR source dominates that has a steep double-power-law X-ray spectrum and no evidence of thermal emission above 10 keV, indicating substantial electron acceleration to tens of keV early in the flare. In the subsequent impulsive phase, three footpoint sources with much flatter double-power-law HXR spectra appear, together with a coronal superhot ( MK) thermal source. The north footpoint and the coronal T ∼ 40 source both move systematically to the north-northeast at speeds up to ∼50 km s Ϫ1 . This footpoint's HXR flux varies approximately with its speed, consistent with magnetic reconnection models, provided the rate of electron acceleration varies with the reconnection rate. The other footpoints show similar temporal variations but do not move systematically, contrary to simple reconnection models. The g-ray line and continuum emissions show that ions and electrons are accelerated to tens of MeV during the impulsive phase. The prompt de-excitation g-ray lines of Fe, Mg, Si, Ne, C, and O-resolved here for the first time-show mass-dependent redshifts of 0.1%-0.8%, implying a downward motion of accelerated protons and a-particles along magnetic field lines that are tilted toward the Earth by ∼40Њ. For the first time, the positron annihilation line is resolved, and the detailed high-resolution measurements are obtained for the neutron-capture line. The first ever solar g-ray line and continuum imaging shows that the source locations for the relativistic electron bremsstrahlung overlap the 50-100 keV HXR sources, implying that electrons of all energies are accelerated in the same region. The centroid of the ion-produced 2.223 MeV neutron-capture line emission, however, is located ∼ away, implying 20 ‫ע‬ 6 that the acceleration and/or propagation of the ions must differ from that of the electrons. Assuming that Coulomb collisions dominate the energetic electron and ion energy losses (thick target), we estimate that a minimum of ∼ ergs is released in accelerated 1∼20 keV electrons during the rise phase, with ∼10 31 ergs in ions above 31 2 # 10 2.5 MeV nucleon Ϫ1 and about the same in electrons above 30 keV released in the impulsive phase. Much more energy could be in accelerated particles if their spectra extend to lower energies.