Supernovae in deep Hubble Space Telescope galaxy cluster fields:: cluster rates and field counts

We have searched for high-redshift supernova (SN) candidates in multiple deep Hubble Space Telescope (HST ) archival images of nine galaxy cluster fields. We detect six apparent SNe, with 21.6 less than or equal to I-814 less than or equal to 28.4 mag. There is roughly 1 SN per deep (I-814 >26 mag), doubly-imaged WFPC2 cluster field. Two SNe are associated with cluster galaxies (at redshifts z = 0.18 and z = 0.83), three are probably in galaxies not in the clusters (at z = 0.49, z = 0.60, and z = 0.98), and one is at unknown z . After accounting for observational efficiencies and uncertainties (statistical and systematic) we derive the rate of type Ia SNe within the projected central 500 h(50)(-1) of rich clusters: R = 0.20(-0.19)(+0.84) h(50)(2) SNu in 0.18 less than or equal to z less than or equal to 0.37 clusters, and R = 0.41(-0.39)(+1.23) h(50)(2) SNu in clusters at 0.83 less than or equal to z less than or equal to 1.27 (95 per cent confidence interval; 1 SNu = 1 SN century(-1) per 10(10) L-Bcircle dot ). Combining the two redshift bins, the mean rate is R-(z) over bar =0.41 = 0.30(-0.28)(+0.58) h(50)(2) SNu. The upper bounds argue against SNe Ia being the dominant source of the large iron mass observed in the intracluster medium. We also compare our observed counts of field SNe (i.e. non-cluster SNe of all types) to recent model predictions. The observed field count is N less than or equal to 1 SN with I-814 less than or equal to 26 mag, and 1 less than or equal to N less than or equal to 3 SNe with I-814 less than or equal to 27 mag. These counts are half of some of the predictions. Since the counts at these magnitudes are likely dominated by type II SNe, our observations may suggest obscuration of distant SNe II, or a SN II luminosity distribution devoid of a large high-luminosity tail.