1. High-energy gammas from the giant flare of SGR 1806-20 of December 2004 in AMANDA Juande D. Zornoza (zornoza@icecube.wisc.edu) on behalf of the IceCube collaboration The AMANDA neutrino telescope consist of a 3D array of 677 optical modules distributed along 19 strings deployed at depths of 1500-2000 m in the South Pole. These optical modules can detect the Cherenkov light emitted by the muons produced in the CC interactions of high energy neutrinos in the ice. The ice above the detector reduces the background of the atmospheric muons by 6 orders of magnitude. Using the information of the position and the time of the hits, the track of the muon can be reconstructed. Other signatures are also possible, like cascades, which produce almost spherical light distributions. SatelliteTime at Earth (ms) GEOTAIL21:30:26.71 INTEGRAL21:30:26.88 RHESSI21:30:26.64 CLUSTER 421:30:26.15 Double Star21:30:26.49 Time window: 1.5 s 0.4 s Swift-BAT light curve ABSTRACT On December 27th 2004, a giant  -flare from the Soft Gamma-ray Repeater 1806-20 saturated many satellite gamma-ray detectors. This event was by more than two orders of magnitude the brightest cosmic transient ever observed. Analysis of previous SGR flares pointed out the possibility of a hard power law energy spectrum. If the gamma emission extends up to TeV energies, photo-produced muons could be observed in surface and underground arrays. AMANDA-II was used to search for down-going muons indicative of high-energy gammas. The data revealed no significant signal. The upper limit on the gamma flux at 90% CL is dN/dE < 0.05 (0.5) TeV -1 m -2 s -1 for  =-1.47 (-2). CC-  interactions: long (~km) tracks NC- and CC- e /  interactions: cascades (tracks short wrt the OM distance) 15 m AMANDA detector Soft Gamma-Ray emitters Observation: Soft-Gamma Ray sources (SGRs) are X-ray pulsars which emit periodically (5-10 s) soft X-ray (2-10 keV), with luminosities of the order of 10 33-35 erg/s. Shorter bursts (0.1 s) with luminosities reaching 10 41 erg/s are also observed sporadically. In rare occasions, these burst are giant flares of extraordinary luminosity (10 47 erg/s). Theoretical model: The “magnetar” model is a neutron star with very high magnetic field (B~10 15 G) which suffers global crust fractures when these magnetic fields are rearranged. The energy spectrum measured in previous SGRs suggested that a significant component of high energy gammas could be detected. The showers produced by these photons are muon poor, but the large number of secondary photons would produce some pions, which would decay into muons. Analysis technique The detector stability is important to avoid the effect of non-physical-related behavior on the event rate, something critical when investigating flare events. It was checked that the detector was stable in the moment of the flare. Also, other indicators, like the dead time, had normal values during the run of interest. References [2] F. Halzen, H. Landsman, T. Montaruli, TeV photons and Neutrinos from giant soft-gamma repeaters flares, astro-ph/0503348. [1] A. Achterberg et al, Limits on the high-energy gamma and neutrino fluxes from the SGR 1806-20 giant flare of December 27th, 2004 with the AMANDA-II detector, submitted to Phys. Rev. Lett. Results The “blindness” policy aims to avoid to introduce biases in the analysis. After the performing the optimization of the time and angular windows, the flare data were unblinded. No event was observed correlated with the position and time of the source. angular window  Since no event was found, upper limits were set. Assuming a flux with shape dN/dE=AE , the limit in the normalization constant at 90% confidence level is:  = – 1.47  A 90 = 0.05 TeV -1 m -2 s -1  = –2  A 90 = 0.5 TeV -1 m -2 s -1 Sensitivity (red dotted line) and limit in the normalization constant of the gamma flux as a function of the spectral index. The limit on the neutrino flux is also indicated. These results allow to exclude some of the theoretical hypothesis mentioned in [1]. In order to transform the number of observed events at the detector into a incoming flux at the Earth (or an upper limit on it), the effective area is used: The effective area for neutrinos is lower due to the small column density for down-going neutrinos. Angular window optimization The procedure to choose the optimum angular window is as follows: - the background on-source and off-time is calculated using real data. - the signal is simulated to estimate the angular resolution and the effective area of the detector. - the optimum cone search is found by minimizing the Model Discovery Factor: where  is the Poisson mean of the number of signal events which would allow to reject the only-background hypothesis (at a CL confidence level) in SP% of equivalent measurements. Time window selection The time window is chosen taking into account the flare width and the differences in the arrival times calculated from several satellites. The length of the burst is <0.6 s, and the differences in the satellite times (translated to time in the AMANDA detector) are ~0.7 s. Detector stability Monte Carlo studies Optimum cone: 5.8 deg Best MDF: 2.31 # required events: 4 background: 0.060 In this analysis the MC is used for the estimation of the angular resolution and the calculation of the effective areas. Several variables were investigated to compare MC simulation and data. The agreement between MC and data is good. The angular resolution was also checked to be almost independent of the spectral index. This is important because of its possible impact on the angular window optimization. ~80% signal retained >90% bg in 20 o rejected. Duration < 0.6 s The saw-profile of the MDF curves comes from the discreteness of the Poisson distribution. The Model Rejection Factor determines the upper limit in case of no event observed. Since it is quite flat, it means that the upper limit would not be far from the optimum value. We chose a conservative time window. Provided the flare is inside the window, the result does not depend critically on its size. Number of events on-source off- line. No instabilities are observed Time difference between events, which determines the dead time : 17%. The experimental approach consist in observing down-going muon events which could have been produced by TeV gamma rays. Those events have to be discriminated from the background of muons produced by cosmic rays in the atmosphere. This is the first time that a high-energy neutrino telescope has been used as a gamma observatory. The advantage of detectors like AMANDA is the large duty-cycle and the wide field of view [2]. Number of events predicted in AMANDA for several spectral indexes. The two values correspond to a maximum energy of 200 and 500 TeV. The units of F  are 10 -12 cm -2 s -1 TeV -1 [1]. Theoretical prediction of muon spectrum from a differential photon spectrum  0. Sketch of the AMANDA-II detector photons neutrinos Effective area: the size of a 100% efficient flat surface that observes the same number of photons/neutrinos than the detector. This means On the other hand, this analysis has also shown that given the level of background expected in flares lasting for a few seconds, the AMANDA detector can be used to observe TeV-gamma flares, which the advantage of long duty-cycle and large field of view. Other spectral indices: Further information: http://icecube.wisc.edu