An Extended Burst Tail from SGR 1900+14 with a Thermal X-Ray Spectrum

Abstract

The Soft Gamma Repeater, SGR 1900+14, entered a new phase of activity in April 2001 initiated by the intermediate flare recorded on April 18. Ten days following this flare, we discovered an abrupt increase in the source flux between consecutive Rossi X-Ray Timing Explorer RXTE orbits. This X-ray flux excess decayed over the next several minutes and was subsequently linked to a high fluence burst from SGR 1900+14 recorded by other spacecraft (Ulysses and Wind/Konus) while the SGR was Earth-occulted for RXTE. We present here spectral and temporal analysis of both the burst of 28 April and the long X-ray tail following it. We find strong evidence of an exclusively thermal X-ray tail in this event and bring this evidence to bear on other bursts and flares from SGR 1900+14 that have shown extended X-ray excesses (e.g. 1998 August 29). We include in this comparison a discussion of the physical origins of SGR bursts and extended X-ray tails.

The dynamic behavior of soft gamma repeaters

“…The ratio of tail energy to burst energy is consistent with being constant among the group ( Fig. 7; Lenters et al, 2003). Timing anomalies and pulse profile changes are seen following the two most energetic bursts and the pulse fraction increases for the three weakest.…”

“…For the August 29 and April 28 burst tails, the location of the heating is also constrained (Lenters et al, 2003). In each of these events, we have precise pulse phase information prior to, during, and after the tail.…”

“…This event was detected on April 28 and like previous tails, showed a transient increase in pulse fraction and a cooling blackbody spectral component (Lenters et al, 2003). In fact, this is the only tail thus far to show an exclusively thermal X-ray tail.…”

“…However, the bremsstralung model yields a column density $5 times larger than the measured column from the persistent emission whereas the two component model fit yields a column consistent with the persistent emission value. The pulsed fraction increases above the quiescent level (11% RMS) up to $20% during this tail (Lenters et al, 2003), and the phase of the pulsations does not shift during the tail relative to the pre-burst pulse phase.…”

“…The second of these events was recorded on 1998 August 29 (Ibrahim et al, 2001;Lenters et al, 2003), just two days after the flare. This burst had a high gamma-ray fluence and an X-ray tail whose flux decayed more rapidly than a power-law in time (Lenters et al, 2003). We define the onset of the tail emission when the intensity quickly drops by $4 orders of magnitude within a few tens of milliseconds.…”