Outliers to the peak energy-isotropic energy relation in gamma-ray bursts

The peak energy-isotropic energy (EpEi) relation is among the most intriguing recent discoveries concerning gamma-ray bursts (GRBs). It can have numerous implications for our understanding of the emission mechanism of the bursts and for the application of GRBs to cosmological studies. However, this relation has been verified only for a small sample of bursts with measured redshifts. We propose here a test of whether a burst with an unknown redshift can potentially satisfy the EpEi relation. Applying this test to a large sample of BATSE bursts, we find that a significant fraction of those bursts cannot satisfy this relation. Our test is sensitive only to dim and hard bursts, and therefore this relation might still hold as an inequality (i.e. there are no intrinsically bright and soft bursts). We conclude that the observed relation seen in the sample of bursts with known redshift might be influenced by observational biases and the inability to locate and to localize well hard and weak bursts that have only a small number of photons. In particular, we point out that the threshold for detection, localization and redshift measurement is essentially higher than the threshold for detection alone. We predict that Swift will detect some hard and weak bursts that would be outliers to the EpEi relation. However, we cannot quantify this prediction. We stress the importance of understanding the detection-localization-redshift threshold for the coming Swift detections.