1. Decelerating and Dustfree: Dark Energy Studies of Supernovae with the Hubble Space Telescope Kyle Dawson March 16, 2008 For the SuperNova Cosmology Project RDCS 1252.9 @ z=1.23 (ISAAC and ACS)

2. Supernovae and the Accelerating Universe Brief discussion of systematics – probably heard it all before from J. Guy Our HST program to combat the dust systematic Some initial results

3. The Problem with Dust Host galaxy dust extinction If host galaxy is similar to Milky Way then:  M B =R B E(B-V) with R B ~ 4 Color error ~ 0.1 => Magnitude error ~0.4 Extinction correction dominates measurement error! Before extinction correction After extinction correction

4. What About Intrinsic Color of SNe? Extinction:  M B =R B E(B-V) Conventional dust extinction predicts R B ~ 4. However, dispersion in SNe color contributes to observed B-V as well. Introduce color correction  to account for intrinsic color + host extinction  B =m max B – M +  (s-1) -  c Measure s, c, m max B, fit M, , and  SNe are better fit with  ~ 2.5 (depending on who you ask). Sullivan et al. 08 Residual without c-correction β=4.1

5. The Problem with Dust Host galaxy dust extinction  M B =R B E(B-V) with R B ~ 4 Degeneracy between intrinsic color of SNe and reddening from host lead to poorly constrained extinction correction Before extinction correction After extinction correction The error increases at higher redshift AND Corrections may be biased due to incomplete understanding of dust properties at high z! Or is it 2.5???

6. One Possible Solution: Use SNe Hosted by Dust-Free Elliptical Galaxies Galaxy type Dispersion (no Ext Corr) Elliptical: E/S0  mag Spiral: Sa/Sb/Sc  mag Irregular Scd/Irr  mag Solution: Use Ellipticals and don’t apply the extinction correction. But, at z~0.5, only ~1 in 5 SNe Ia is hosted by an elliptical galaxy. M. Sullivan et al. (SCP) 2003, MNRAS, 340, 1057.

7. Minimizing host galaxy dust extinction Elliptical Galaxies & Clusters Elliptical galaxies (E/S0) have little star-formation & dust. Clusters: concentration of nearly dust free Elliptical galaxies. B-V color of low-z SNe SN color: With few exceptions, SNe hosted by early type galaxies exhibit color consistent with no host extinction

8. No dust in elliptical (cluster) galaxies? Ellipticals in massive clusters are among the earliest (oldest) objects in the Universe. How do we know these galaxies are dust-free? * Small scatter in color for cluster galaxies – uniform population * Uniform color within a galaxy (dust is clumpy) * No emission at ~200 microns (dust emission)

9. Coherent evolution of cluster environment Scatter of color of ellipticals galaxies in clusters is very small (consistent with current resolution). Evolution from date of formation (z~3) appears consistent for clusters over wide range of redshifts. Strong constraints on host environment. 25 clusters with HST resolution will be added Blakeslee et al. ApJL (2003)

10. Decelerating and Dustfree: Targeting SNe in Very High Redshift Galaxy Clusters Decelerating and Dustfree: Targeting SNe in Very High Redshift Galaxy Clusters Major cycle 14 HST Program in collaboration with several galaxy cluster groups (219 orbits over 18 months). searched 25 recently discovered massive galaxy clusters at z ≥ 1 with ACS. Why clusters? Dominated by nearly dust-free Elliptical (E/S0) galaxies. 5 times higher density of Elliptical galaxies at high redshift. Scientific goals: Significantly improve SN constraints of dark energy - reduce the statistical and systematic uncertainties. Cluster studies: weak lensing, galaxy morphology, and color-magnitude relationship. SpitzerXMM+VLTVLTACS

11. Left Panel: Simulated 68% confidence region on w' vs. w 0 for the current literature SN sample but with underlying cosmology(w0 = -1; w’ = 0). The parameters are poorly constrained because color errors are magnified by R B ~ 4. Center Panel: Demonstration of bias introduced from assuming a prior. Solid filled curve-R B =4.1, dashed curve, R B =2.5. Right Panel: The goal of this project is shown as a confidence region for a simulated new sample of 10 z > 1 SNe Ia found in ellipticals, together with 5 in ellipticals from the past and ongoing GOODS searches, as well as 120 SNe Ia in ellipticals at the lower redshifts now being produced by the ground-based CFHT SN Legacy Survey, the CTIO Essence survey, and (at z < 0.1) the Nearby SN Factory What Sort of Improvement are we Looking For?

12. The Clusters Weak lensing mass estimates for 18 clusters Red-sequence modeled from extensive ground-based spectroscopy + 2 band ACS imaging Richness estimates for all 25 clusters – N gal from red-seq Consistent with model from SDSS (dotted line) M proportional to N gal 1.2 All SNe discovered in richest clusters  unshaded region

13. HST Rolling Search 20-26 day cadence ACS z’, i’ bands. Observing Program: 25 massive galaxy clusters at z ≥ 1 Search and follow: Continuous flow of ACS data Followup: NICMOS J-band scheduled ~10 days after discovery Spectroscopy: At least one night/month

14. Intensive Ground-based Spectroscopy Follow-up Goal: Establish redshift, host type, and often SN type Subaru: 10 half- nights/semester Keck: 2 nights/semester VLT: 10 hours/semester (queue mode) + Director ’ s Discretionary time E Host  SNIa with high confidence Preliminary spectra from this program

15. Results: SNe Discoveries in HST Program 15 z>1 SNe discovered, 9 hosted by E type galaxies Full multi-color light curves for 8/9 hosted by E type galaxies Full light curves for 4/7 z>0.95 SNe hosted by late-type galaxies Twice the SN/orbit as previous SN searches in random fields using HST Three times the E hosted-SN/orbit!

16. Factors to Consider for ACS Photometry Many issues contributing at the ~5% level were solved in 2007: Time variation of distortion corrections Time variation of bias levels Relative astrometry corrections PSF variation across FOV Variation/gradient in flat field Backscattering of NIR photons K-corrections of different templates Gravitational lensing of SNe behind clusters K-corrections using various lightcurve templates Almost there! Still blinded but expect to put E-hosted SNe on Hubble diagram very soon

17. The Elliptical host (E/S0) Hubble Diagram No extinction correction Preliminary Sullivan et al. & Knop et al. This program Seven SNe Ia in elliptical galaxies observed with complete lightcurves. No extinction correction applied (David Rubin) Example of E–only Hubble Diagram other analyses SN Ia rate in cluster environment Host Properties Cluster studies (collaborators) Weak lensing Red Sequence Scatter Etc.

18. Summary 219 orbits with HST to search for SNe in high z clusters Efficient observing strategy, finding 3x E hosted SNe/orbit over previous GOODS searches Each E-hosted SN worth 9 normal SNe Results soon, photometry complete, unblinding SNe results will include cosmology, properties of SNe by host type, rates Lots of parallel science: weak lensing mass estimates, morphology, evolution, spectral properties