• DOI:10.1086/323256
• Corpus ID: 10278561

The Giant Flare of 1998 August 27 from SGR 1900+14. II. Radiative Mechanism and Physical Constraints on the Source

@article{Thompson2001TheGF,
  title={The Giant Flare of 1998 August 27 from SGR 1900+14. II. Radiative Mechanism and Physical Constraints on the Source},
  author={Christopher Thompson and Robert C. Duncan},
  journal={The Astrophysical Journal},
  year={2001},
  volume={561},
  pages={980 - 1005}
}

• C. Thompson, R. Duncan
• Published 30 October 2001
• Physics
• The Astrophysical Journal

The extraordinary 1998 August 27 giant flare places strong constraints on the physical properties of its source, SGR 1900+14. We make detailed comparisons of the published data with the magnetar model, which identifies the soft gamma repeaters as neutron stars endowed with ~1015 G magnetic fields. The giant flare evolved through three stages, whose radiative mechanisms we address in turn. The extreme peak luminosity L > 106LEdd, hard spectrum, and rapid variability of the initial ~0.5 s spike emission all point to an expanding pair fireball with very low baryon contamination. We argue that this energy must have been deposited directly through shearing and reconnection of a magnetar-strength external magnetic field. Low-order torsional oscillations of the star fail to transmit energy rapidly enough to the exterior, if the surface field is much weaker. A triggering mechanism is proposed, whereby a helical distortion of the core magnetic field induces large-scale fracturing in the crust and a twisting deformation of the crust and exterior magnetic field. After the initial spike (whose ~0.4 s duration can be related to the Alfvén crossing time of the core), very hot (T ≲ 1 MeV) plasma rich in electron-positron pairs remains confined close to the star on closed magnetic field lines. The envelope of the August 27 flare can be accurately fitted, after ~40 s, by the contracting surface of such a "trapped fireball." The form of this fit gives evidence that the temperature of the trapped pair plasma decreases outward from its center. We quantify the effects of direct neutrino pair emission on the X-ray light curve, thereby deducing a lower bound μmin ~ 1 × 1032 G cm3 to the magnetic moment of the confining field, comparable to the strongest fields measured in radio pulsars. The radiative flux during the intermediate ~40 s of the burst appears to exceed the trapped fireball fit. The lack of strong rotational modulation and intermediate hardness of this smooth tail are consistent with the emission from an extended pair corona, in which O-mode photons are heated by Compton scattering. This feature could represent residual seismic activity within the star and accounts for ~10% of the total flare fluence. We consider in detail the critical luminosity, below which a stable balance can be maintained between heating and radiative cooling in a confined, magnetized pair plasma, but above which the confined plasma runs away to a trapped fireball in LTE. The emergence of large-amplitude pulsations at ~40 s probably represents a transition to a pair-depleted photosphere whose main source of opacity is electrons (and ions) ablated from the heated neutron star surface. The best-fit temperature of the blackbody component of the spectrum equilibrates at a value that agrees well with the regulating effect of photon splitting. The remarkable four-peaked substructure within each 5.16 s pulse, as well as the corresponding collimation of the X-ray flux, has a simple explanation based on the strong inequality between the scattering cross sections of the two photon polarization modes. The width of each X-ray "jet" is directly related to the amount of matter advected outward by the high cross section ordinary mode.