NAOKI YASUDA, MAMORU DOI (UTOKYO), AND TOMOKI MOROKUMA (NAOJ) SN Survey with HSC

SN Ia as standard candle Very bright (M B ~-19.3)  Observable at cosmological distances (z~1.5) Light-curve shape (  m 15, stretch) / luminosity relation  Broader light-curve -> intrinsically brighter  Accurate to ~7% Accelerated expansion of the Universe

Luminosity Normalization Jha 2002 Astier et al. 2006

Reiss et al. (2007)

Complementarities Constraints from SN Ia is complementary to the constraints from LSS Independent attempt is important Astier et al. 2006

SN Ia progenitors Sullivan et al. (2006) SN Ia rate as a function of SFR of host galaxies Two components  SN rate proportional to SFR and stellar mass Light curve shapes depend on host galaxies Sullivan et al  Bright Faint  PromptDelayed

ESA-ESO Working Groups : Fundamental Cosmology (2006) List of SN Survey

Advantage of HSC Large aperture  Other SN surveys except for LSST use 4m telescopes  SN Ia samples are limited to z<0.9  Extend to z~1.2 Wide field  1FoV is comparable to survey area of SNLS High sensitivity in red bands (z-, Y-band)  Most energy of SN z=1 fall in i-, z-, and Y-band

Advantage of HSC Large aperture  Other SN surveys except for LSST use 4m telescopes  SN Ia samples are limited to z<0.9  Extend to z~1.2 Wide field  1FoV is comparable to survey area of SNLS High sensitivity in red bands (z-, Y-band)  Most energy of SN z=1 fall in i-, z-, and Y-band 1,000 z= from 4FoV and 4month duration observation

Performance of Subaru/Suprime-Cam Number of candidates  i < 25mag 1 month separation  SNe / deg 2 / month  1,000 SNe / 4FoV / 3months Photometry  Good enough for light-curve fitting for z~1  Comparable to HST photometry Oda et al. (2007)

Proposal 1,000 SN z = combined with previous surveys Expanding history of the Universe  Limit on the time variation of dark energy SN Ia rate and its environmental effect, evolution  Clue to the progenitor of SN Ia  Two evolutionary channel?

Observing Strategy “Multi-color rolling search”  Observe the same field repeatedly with multi colors  Maximum brightness  photometric typing / redshift  Not enough facilities for spectroscopy 5nights (every 5 days) x 4months x 2 in (r,)i,z, and Y-bands: ~1000 SN light curves Most SNe are observable over 2months

Comparison with on-going SN Surveys SDSS-II : ~60nights/yr x 3yrs (2.5m)0.1 < z < 0.3 SNLS : ~60nights/yr x 5yrs (3.6m)0.3 < z < 0.8 HSC : ~40nights/yr x 1yr (8.2m)0.6 < z < 1.2  1,000 SNe from 4FoV, 4months  Much cheaper than HST

Sample Observation Plan

Photometric typing / redshift Fitting to multi-epoch spectral templates Typing  ~90% of SN Ia candidates are confirmed spectroscopically from the data of a few epochs (SDSS-II) -> details in Ihara’s talk Redshift   z/(1+z) ~ 2-3% (SNLS) Guy et al. 2007

Photometric Redshift Simulation  Cosmology :  M = 0.3,   = 0.7, w = -1, w’ = 0.0  1hour exposures of i-, z-, and Y-band at (-8, -3, 0, +3, +8) days from new moon over 3months  Stretch parameter : / (Max magnitude : +/- 0.2)  Explosion time : from -15 days to +15 days  Color is fixed to 0.0 : same intrinsic color and no extinction  Redshift : 0.8, 0.9, 1.0, 1.1, and 1.2 Photo-z by light curve fitting program (SALT)  SALT is developed for SNLS analysis

Photo-z Results

Offset of mean value  Difference of spectral templates between light curve simulation (Hsiao template) and light curve fitting program (SALT)? Dispersion   z/(1+z) ~ 1-2% Catastrophic errors  Misidentification of colors  Degeneracy due to wavy feature of SNe spectrum?

Cosmology Errors on  M and w reduce by a factor of 2 Area encircled reduce by a factor of 2 Contour : 1 

Cosmology Systematic error due to photo-z error Contour : 1 

Cosmology Redshift should be determined well below 1% level  Difficult only with photometric information Need spectroscopic information  Combine with photo-z of host galaxies?  Different error properties are expected  Slitless (Grism) spectroscopy?  High sky noise  More observing time  Spectroscopy of host galaxies  Need large observing time  Only for elliptical hosts (no extinction)?

SN Ia rate, progenitor, … Do not need very accurate redshift Correlation with host galaxy  Brighter SNe are in later spirals SN rate  Two component model Proportional to  SFR  Stellar mass Two evolutional path Effect on chemical evolution Neill et al. 2007

Summary HSC can detect ~1000 SNe with reasonable observing time (~40 nights). Photometric Redshift can be determined to 1-2% level. For cosmology we need more accurate redshift. Nature of SNe Ia and their evolution can be explored with large sample.