Signal and background in the underwater neutrino telescope ANTARES

At the bottom of the Mediterranean, the neutrino telescope ANTARES is being constructed. Its purpose is to detect cosmic neutrinos, which can yield information on distant and energetic processes that cannot be obtained from the more traditional study of light or charged particles. ANTARES searches for neutrinos by detecting the light they produce in interactions with the surrounding rock or water. Since neutrinos only interact weakly, very few of them will be observed in ANTARES, and distinguishing a neutrino signal from the background is a big challenge. Two major sources of optical background are the decay of radioactive potassium, which creates a steady background rate of about 30 kHz, and bioluminescence, which gives an additional steady background as well as intense flashes of light. The optical background at the location of ANTARES was studied with a prototype detector. During ninety days, the photon count rate was measured with an accuracy of 10 kHz. The behaviour of the count rate was highly variable. Its characteristics are described, at the timescale of 15 minutes, by the base rate (the steady rate independent of bursts) and the burst rate (the number of bursts per second). These rates show strong fluctuations: the base rate varied from 60 kHz to over 350 kHz, and the burst rate varied between 0 and 0.8 Hz. Base rate and burst rate were correlated, and both were correlated with the speed of the water current. A periodicity of 17.7 hours, the period of inertial waves at the latitude of ANTARES, was observed in the current speed as well as in the bioluminescence. These results suggest that the bioluminescent bursts are caused by macroscopic organisms emitting light in reaction to collisions with the detector. While the trigger algorithm can handle count rates of up to 600 kHz, bioluminescence forms an unpleasant background for the detection of cosmic neutrinos. For future underwater neutrino telescopes, it is recommended to search for a location with less bioluminescent activity. The main method of detecting cosmic neutrino sources is to reconstruct muon tracks caused by neutrino interactions. At low energies, this method is inefficient, because not enough light is produced. A new method has been devised to search for a low-energy neutrino flux from a given direction, without reconstructing individual muon tracks. This velocity filter is based on the asymmetry in the patterns of photon arrival times caused by a muon. The velocity filter was tested on simulated data, assuming a neutrino flux as is expected from Supernova Remnant RX J1713.7-3946 (~ 10-4 (E/GeV)-2.2 m-2 s-1 GeV-1). The neutrino signal cannot be detected, partly due to the large systematic uncertainty on the flux of atmospheric muons, but mainly because the expected neutrino flux is too low to be detectable above the statistical error on the background. A neutrino flux about three orders of magnitude larger than the flux expected from RX J1713.7-3946 can be detected at the 5σ level. In larger detectors and with less optical background, the velocity filter may be able to detect cosmic neutrino sources