Supernova surveys can be used to study a variety of subjects, such as: (i) cosmology using type Ia supernovae, (ii) star formationrates using core-collapse SNe, (iii) supernova properties and their connection to host galaxy characteristics. The Stockholm VIMOSSupernova Survey (SVISS) is a multi-band imaging survey aiming to detect supernovae at redshift ∼0.5 and derive thermonuclearand core-collapse supernova rates at high redshift. In this paper we present the supernovae discovered in the survey along with lightcurves and a photometric classification into thermonuclear and core-collapse types. To detect the supernovae in the VLT/VIMOSmulti-epoch images we used difference imaging and a combination of automatic and manual source detection to minimise the numberof spurious detections. Photometry for the found variable sources was obtained and careful simulations done to estimate correct errors.The light curves were typed using a Bayesian probability method and Monte Carlo simulations were used to study misclassification.We detected 16 supernovae, eight of which had a core-collapse origin and eight that had a thermonuclear origin. The estimatedmisclassification errors are quite small, on the order of 5%, but vary with both redshift and type. The mean redshift of the supernovaeis 0.64. Additionally, we found a variable source with a very extended light curve that could possibly be a pair instability supernova.

We present supernova rate measurements at redshift 0.1–1.5 from the Stockholm VIMOS Supernova Survey (SVISS). The sample contains 16 supernovae in total. The discovered supernovae have been classified into core collapse or thermonuclear (Ia) types based on their light curves, colour evolution and host galaxy photometric redshift. The rates we find for the core collapse supernovae are 1.25 (+2.27 +0.85 −0.97 −0.78) - with statistical and systematic errors, respectively - at z = 0.39 and 6.90 (+5.24 +3.04 −3.25 −2.14) at z = 0.73. For the Ia supernovae the rates are 2.02 (+1.57 +0.53−0.96 −0.57) at z = 0.39 and 1.03 (+0.92 +0.31−0.54 −0.36) at z = 0.80. All of these rate estimates have been corrected for host galaxy extinction. Using Monte Carlo simulations we make a thorough study of the systematic effects from assumptions made when calculating the rates and find that the most important errors comes from misclassification, the assumed mix of faint and bright supernova types and uncertainties in redshift. We compare our rates to other observations, to the star formation history for core collapse rates and to different models of the delay time distribution for Ia rates. Overall, our measurements agree quite well with these other rates when using redshift-dependent corrections for extinction. We do not find any evidence of a missing fraction of core collapse supernovae.

Context. The Stockholm VIMOS Supernova Survey (SVISS) is an optical broad-band photometry survey, aimed at the search fordistant (redshifts z~0.7) supernovae, which are detected by means of monthly imaging in R and I bands with the VIMOS instrumenton the ESO VLT. The distances of the supernovae are estimated using photometric redshifts facilitated by additional U, B and V imaging also with VIMOS. In this paper we present the deep stacked imaging data in all these bands and a catalogue of galaxiesextending to very faint magnitudes.

Aims. We discuss the various steps followed to create a galaxy catalog, including the reduction and stacking of the data, astrometricand photometric calibration and source detections.We derived completeness, corrected number counts and color selected Lyman breakgalaxies. Raw number counts are also calculated. Finally, making use of a template fitting code, photometric redshifts are derived.

Methods. UBVRI deep imaging of a 4x56 sq. arcmin field within the ELAIS-S1 field is presented in this paper. The data reduction has been done with a dedicated MIDAS pipeline. The final images have been weighted by their seeing, which was found to optimise depth and spatial resolution. Source detection and phortometry was done using SExtractor. The photometric accuracy was tested by comparing the colors of the sources to stellar libraries and redshifted galaxy templates. A mixture of synthetic point-like andextended sources are simulated in order to estimate completeness. Derived number counts have been compared to the results ofsimilar surveys. Lyman break galaxy candidates were selected using color criteria: UBR, BRI, BVI and VRI, and their numbers computed and compared to other surveys.We have obtained photometric redshifts by spectral energy distribution template fitting. The photometric redshifts were calibrated using a sample of galaxies in the HDF-S with available spectroscopic redshifts that have also been observed with the same filter set as used here.

Results. Using a seeing weighted method to combine the science frames resulted in an increase of the number of faint objects detectedand minimizing the photometric errors. The astrometric accuracy is 0.400 arcsec with respect to 2MASS and 5 sigma limiting magnitudes for UBVRI are 28.2, 29.3, 28.3, 28.5 and 27.2 respectively (Vega). The computed galaxy number counts are in good agreement with previous works, with the U band counts of this survey being one of the deepest ever obtained. The raw number counts from broadband color selected Lyman break galaxies show a good agreement for the U and B dropouts with other surveys. Photometric redshifts have a scatter of dz=0.085 with a 9.1% fraction of outliers.

Conclusions. We derive deep number counts which are found to be consistent with previous works, reproducing the bright and faintends. Especially, the stacked deep U and I band data are found to present a comptetetive combination of depth and area, comparableto HDF observations. As well as V band, comparable to Subaru and the Lockman hole observations. Our U and B dropouts are within the expected redshift range, with mean values of 3 and 3.8 respectively.

Context. The Stockholm VIMOS Supernova Survey (SVISS) is an optical broad-band photometry survey, aimed at the search fordistant (z0.7) supernovae (SNe), which are detected by means of monthly imaging in R and I bands with the VIMOS instrument onthe ESO VLT. The distances to the SNe are estimated using photometric redshifts facilitated by additional U, B and V imaging alsowith VIMOS.

Aims. We investigate the properties of host galaxies for two SN types, thermonuclear and core collapse, by comparing both samples.We summarize the steps necessary to produce a SN catalogue (from detection, to photometry and typing) and explain the processfollowed to identify the host galaxies and determine their properties.

Methods. Using a spectral energy distribution (SED) fitting 2 minimization code we compute spectral energy distribution, age, starformation rate (SFR), absolute magnitudes for the host galaxies of the SN sample. Such galaxies are chosen based on their photometricredshift (GOODZ) and SED (modified version of Hyperz). From comparisons of the different photometric parameters’ distributionsbetween CC and TN SN host galaxies, we derived different trends.

Results. We find that CC SNe tend to be more centrally concentrated than TN SNe, with the former being present in late-type/starforminggalaxies and the latter in all galaxy types. CC SNe seem to favor brighter galaxies than TN SNe. KS tests on the stellarmass-extinction distribution show that TN and CC are drawn from the normal galaxy population, and on the SFR-extinction indicatesthat TN SN host galaxies are culled from it as well, but CC SN host galaxies are not derived from it.