The aim of the work presented in this paper is to test and optimise supernova detection methods based on the optimal image subtraction technique. The main focus is on applying the detection methods to wide field supernova imaging surveys and in parti
cular to the Stockholm VIMOS Supernova Survey (SVISS). We have constructed a supernova detection pipeline for imaging surveys. The core of the pipeline is image subtraction using the ISIS 2.2 package. Using real data from the SVISS we simulate supernovae in the images, both inside and outside galaxies. The detection pipeline is then run on the simulated frames and the effects of image quality and subtraction parameters on the detection efficiency and photometric accuracy are studied. The pipeline allows efficient detection of faint supernovae in the deep imaging data. It also allows controlling and correcting for possible systematic effects in the SN detection and photometry. We find such a systematic effect in the form of a small systematic flux offset remaining at the positions of galaxies in the subtracted frames. This offset will not only affect the photometric accuracy of the survey, but also the detection efficiencies. Our study has shown that ISIS 2.2 works well for the SVISS data. We have found that the detection efficiency and photometric accuracy of the survey are affected by the stamp selection for the image subtraction and by host galaxy brightness. With our tools the subtraction results can be further optimised, any systematic effects can be controlled and photometric errors estimated, which is very important for the SVISS, as well as for future SN searches based on large imaging surveys such as Pan-STARRS and LSST.
