The 2004 Hyperflare from SGR 1806-20: Further Evidence for Global Torsional Vibrations

We report an analysis of the archival Rossi X-ray Timing Explorer (RXTE) data from the December 2004 hyperflare from SGR 1806-20. In addition to the approx. equal to 90 Hz QPO first discovered by Israel et al., we report the detection of higher frequency oscillations at approx. equal to 150, 625, and 1,835 Hz. In addition to these frequencies there are indications of oscillations at approx. equal to 720, and 2,384 Hz, but with lower significances. The 150 Hz QPO has a width (FWHM) of about 17 Hz, an average amplitude (rms) of 6.5%, and is detected in average power spectra centered on the rotational phase of the strongest peak in the pulse profile. This is approximately half a rotational cycle from the phase at which the 90 Hz QPO is strongly detected. The 625 Hz oscillation was first detected in an average power spectrum from nine successive cycles beginning approximately 180 s after the initial hard spike. It has a width (FWHM) of approx. equal to 2 Hz and an average amplitude (rms) during this interval of 9%. We find a strong detection of the 625 Hz oscillation in a pair of successive rotation cycles beginning about 230 s after the start of the flare. In these cycles we also detect the 1,835 Hz QPO with the 625 Hz oscillation. The rotational phase in which the 625 Hz &PO is detected is similar to that for the 90 Hz QPO, indeed, this feature is seen in the same average power spectrum. During the time the 625 Hz QPO is detected we also confirm the simultaneous presence of 30 and 92 Hz QPOs, first reported by Israel et al. The centroid frequency of the 625 Hz QPO detected with RXTE is within 1 Hz of the M 626 Hz oscillation recently found in RHESSI data from this hyperflare by Watts & Strohmayer, however, the two detections were made in different phase and energy intervals. Nevertheless, we argue that the two results likely represent detections of the same oscillation frequency intrinsic to the source, but we comment on some of the difficulties in making direct comparisons between the RXTE and RHESSI measurements