Title First results of a study of TeV emission from GRBs in Milagrito Permalink

Milagrito, a detector sensitive to γ-rays at TeV energies, monitored the northern sky during the period February 1997 through May 1998. With a large field of view and high duty cycle, this instrument was used to perform a search for TeV counterparts to γ-ray bursts. Within the Milagrito field of view 54 γ-ray bursts at keV energies were observed by the Burst And Transient Satellite Experiment (BATSE) aboard the Compton Gamma-Ray Observatory. This paper describes the results of a preliminary analysis to search for TeV emission correlated with BATSE detected bursts. Milagrito detected an excess of events coincident both spatially and temporally with GRB 970417a, with chance probability 2.8 × 10 within the BATSE error radius. No other significant correlations were detected. Since 54 bursts were examined the chance probability of observing an excess with this significance in any of these bursts is 1.5×10. The statistical aspects and physical implications of this result are discussed. I OBSERVATIONS AND ANALYSIS Milagro, a new type of TeV γ-ray observatory sensitive at energies above 100 GeV, with a field of view of over one steradian and a high duty cycle, began operation in February 1999, near Los Alamos, NM. A predecessor of Milagro, Milagrito [5], operated from February 1997 to May 1998. During this time interval, 54 γ-ray bursts (GRBs) detected by BATSE [1] were within Milagrito’s field of view (less than 45◦ zenith angle). A search was conducted in the Milagrito data for an excess of events above the cosmic-ray background coincident with each of these γ-ray bursts. For each burst, a circular search region was defined by the BATSE 90% confidence interval, which incorporates both the statistical and systematic position errors [2]. The size of this 90% confidence interval ranged from 4◦ to 26◦ for the 54 GRBs in the sample. The search region was tiled with an array of overlapping 1.6◦ radius bins centered on a 0.2◦ × 0.2◦grid. This radius was derived from the measured angular resolution of Milagrito and was selected prior to the search. The number of events falling within each of the 1.6◦ bins was tallied for the duration of the burst reported by BATSE. This duration is defined as the time required for BATSE to accumulate 90% of the γ-rays(T90). T90 ranged from 0.1 seconds to 195 seconds for the 54 bursts examined. The angular distribution of background events on the sky was characterized using two hours of data surrounding each burst. This distribution was normalized to the number of events detected by Milagrito over the entire sky during the T90 interval (NT90). The resulting background data were also binned in 1.6 ◦ bins spaced 0.2◦ apart. The Poisson probability that the excess of events in each 1.6◦ bin was due to a background fluctuation was calculated and the bin with lowest probability was taken as the candidate position of a TeV γ-ray counterpart to the BATSE burst. The background and signal counts in this bin were used to calculate a fluence or fluence upper limit for each burst.