Long-term observations of gamma-ray bursts: The complex connection between gamma-ray bursts and supernovae

We present results from optical and infrared observations of 153 gamma-ray bursts (GRBs) using the ANDICAM dual-channel optical/IR imager on the SMARTS 1.3m telescope at CTIO. Our rapid observations of GRBs detected by HETE-2, Integral and Swift, produce 38 afterglow detections and numerous limits on optical/IR afterglow brightness. Many bursts are observed repeatedly during a month post-burst in an attempt to detect GRB-related supernovae (SNe).

Monitoring of GRB 031203 reveals the presence of SN 20031w. GRB 031203/SN 20031w is reminiscent of GRB 980425/SN 1998bw, the prototypical GRB-SN. We also observe SN 2006aj associated with GRB 060218. These events are broadly similar, in that they are underluminous in g-rays compared to typical long-duration GRBs and their optical light curves are dominated by SNe. These events appear to be intrinsically different from and much more common than high-luminosity GRBs.

Surprisingly, no SN was detected following the low-redshift GRB 060614, which suggests a new paradigm is required to explain the formation of at least a subset of long-duration GRBs. However, the redshift of this burst was inferred from the spatial coincidence of the afterglow with a z = 0.125 galaxy. GRB 060614 could represent a chance coincidence between a GRB and a foreground galaxy. Examining a uniform sample of 72 GRB fields indicates that there is generally a ~1% change that an optical afterglow will coincide with a galaxy at random. While spurious host galaxy detections will, therefore, be rare, the possibility must be considered when examining individual GRB/host galaxy examples.

Our attempt to detect SNe associated with an additional 99 GRBs fails to reveal any brightening sources. We assume that the bursts are produced by 98bw-like SNe and calculate their lower redshift limits based on the non-detection of these SNe. For bursts at known redshifts, we determine the upper limit on the brightness of the GRB-SNe. The lack of SNe after long-duration GRBs supports the theory that alternate GRB formation mechanisms may be required, while the lack of 98bw-like events following the majority of short-duration bursts supports the theory that short bursts are produced by the merger of binary neutron stars.