GCN Circular 2936

S. Boggs (UCB), K. Hurley (UCB), D.M. Smith (UCSC), R.P. Lin (UCB),
G. Hurford (UCB), W. Hajdas (PSI), C. Wigger (PSI)

RHESSI observed both the precursor and the giant flare from SGR 1806-20
in their entirety, staring at 21:28:03.44 UT and 21:30:26.65 UT 
2004-12-27 respectively. The SGR was 5 degrees from RHESSI's pointing 
axis which was directed toward the Sun. This placed the SGR outside the 
normal imaging FOV of the instrument. During the main peak of the flare 
the RHESSI spectroscopy detectors were saturated for ~0.5s after the 
rise, but observed the decay of the main peak and the 400-s long 
oscillatory component.

While the main RHESSI spectroscopy detectors (9 segmented germanium
detectors operating from 3 keV to 15 MeV) were saturated during the
peak, the RHESSI particle detector (used for detecting SAA passages)
was able to measure the incident flux with 0.125-s time resolution
in two energy channels determined by thresholds in the electronics:
>65 keV, and >650 keV. These data indicate significant emission above
650 keV for ~0.25 s, during the giant flare, and softening of the giant
peak during its evolution. In addition, even though the RHESSI 
spectroscopy detectors are saturated during the giant peak, we can use
the reset rates of the preamplifiers to constrain the rise and fall 
times of the giant peak to <1 ms and ~65 ms respectively.

The particle detector data and the spectroscopy detector reset rates
allow us to set conservative lower limits to the total fluence of the
primary giant peak: >0.1 erg/cm^2 and >0.3 erg/cm^2 respectively. This
fluence is >1-2 orders of magnitude higher than the 1998 flare of
SGR 1900+14, which had a fluence of 7e-3 erg/cm^2 (Hurley et al.,
Nature 397, 41, 1999). Assuming a distance of ~15 kpc for SGR 1806-20
(Corbel & Eikenberry, A&A 419, 191,2004) and isotropic emission, we
derive a lower limit on the total hard X-ray/gamma-ray energy released
in the giant peak to be >8e45 erg. We note that given BATSE's 
sensitivity of <1e-8 erg/cm^2, this type of giant flare would have been 
detectable by BATSE (as a short, hard GRB) out to >80 Mpc.

When they came out of saturation during the giant peak, the RHESSI
spectroscopy detectors were measuring a peak count rate of
~280,000 cnt/s. After the giant peak, RHESSI recorded a series of 51 
pulsations with a period of 7.56 s, similar to the INTEGRAL, KONUS, and 
Swift-BAT observations (Borkowski et al., GCN 2920
Not found
 
 
; Mazets et al., GCN 
2922
Not found
 
 
; Palmer et al., GCN #2925
Not found
 
 
). The pulse profile shows evidence for 
both spectral variations throughout the pulse, and evolution of the 
pulse shapes throughout the decay. The average 20-100 keV pulse profile 
shows 3-4 peaks in its structure. During this decay phase, the average 
20-100 keV flux is well modeled by the trapped fireball model of 
Thompson & Duncan (ApJ 561, 980, 2001), with an evaporation time 
t_evap=382 s, and index a=0.606, where flux ~ (1-t/t_evap)^(a/1-a). 
Preliminary lightcurves show indications of significant spectral 
softening during the 400-s oscillatory decay.

RHESSI observed the precursor during 21:28:03.44-21:28:04.49 UT, with a 
peak count rate in the spectroscopy detectors of ~30,000 cnt/s, and 
~25,000 counts total. The profile is square, as reported by Swift-BAT 
(Palmer et al., GCN #2925
Not found
 
 
), with emission extending up to 150 keV. We 
see a rise time for the precursor <50 ms, and a fall time <100 ms.

Results of this analysis will be posted as they come available at:
http://www.ssl.berkeley.edu/ipn3/041227