Frontier Exploring the Depths of the Universe Jennifer Lotz, Matt Mountain, & the Frontier Fields Team Space Telescope Science Institute,Spitzer Science Center www.stsci.edu/hst/campaigns/frontier-fields Challenge: Can we peer deeper into the Universe than the Hubble Ultra Deep Field before the launch of the James Webb Space Telescope? HUDF ACS (optical) = 416 orbits WFC3 (IR) = 163 orbits =579 orbits of HST Hubble Deep Fields Initiative science working group: use Hubble + nature’s telescopes x 6 (strong lensing clusters) ⇒ Go intrinsically deeper than HUDF ⇒ Go wider than HUDF+parallels 6 Lensed Fields + 6 parallel “Blank Fields” = New Parameter Space James Bullock (Chair, UCI), Mark Dickinson (NOAO), Steve Finkelstein (UT), Adriano Fontana ( INAF, Rome), Ann Hornschemier Cardiff (GSFC), Jennifer Lotz (STScI), Priya Natarajan (Yale), Alexandra Pope (UMass), Brant Robertson (Arizona), Brian Siana (UC-Riverside), Jason Tumlinson (STScI), Michael Wood-Vasey (U Pitt) Brammer, VLT/Hawk-I K 6 strong-lensing clusters + 6 adjacent parallel fields 140 HST DD orbits per pointing Blank Field ACS/ WFC3-IR in parallel ~29th ABmag in 7 bands 2 clusters per year x 3 years → 840 total orbits 1000 hours Spitzer DD time for ~26.5 ABmag in IRAC 3.6, 4.5 μm Cluster http://www.stsci.edu/hst/campaigns/frontier-fields/ Frontier Fields will also observe 6 fields in parallel with the clusters, the second deepest observations of ‘blank’ fields ever obtained. Simultaneous images are taken with Hubble’s infrared camera WFC3/IR and the optical camera ACS; cameras will swap positions ~6 months later. Deep observations of the Frontier Fields will: • probe galaxies 10-20x intrinsically fainter than any seen before, particularly those before and during reionization • study the early formation histories of galaxies intrinsically faint enough to be the early progenitors of the Milky Way • study highly-magnified high-z galaxies in detail: structures, colors, sizes and provide targets for spectroscopic followup • provide a statistical picture of galaxy formation at early times Average Star Formation Histories From Z=0-8 13 Behroozi et al 2013 Time [Gyr] Time [Gyr] 13 10 7 421 13 10 7 421 10 11.0 10 11.0 M = 10 M M = 10 M h O• h O• 12.0 12.0 12 M = 10 M 12 M = 10 M 10 h O• 10 h O• 13.0 13.0 M = 10 M M = 10 M 11 h O• 11 h O• 10 14.0 10 14.0 M = 10 M ] M = 10 M ] • • h O • h O O • O 10 10 [M 10 10 * M ICL [M 9 9 10 10 8 8 10 10 7 7 10 0 0.5 12345 6 7 8 10 0 0.5 12345 6 7 8 z z FIG.9.—Left panel: Stellar mass growth histories of galaxies of different masses. Lines shown the amount of stellar mass remaining incentralgalaxiesatthe presentUDF day z>~8 that was galaxies in place (intoo any massive progenitor evolve galaxy) at into a given today’s redshift forMilky our bestWay fit model. Right panel: Amount of stellar mass in the intracluster light 15 14 (ICL) remaining at the presentneed day in placeto go at a10x given deeper redshift. Note that plots for 10 M! halos are not shown as they are nearly identical to those for 10 M! halos. Shaded regions in both panels shown the one-sigma posterior distribution. in Fig. 7 of rising to a peak efficiency and later falling is con- sistent with all available data. 1 5.3. Stellar Mass and Intracluster Light Growth Histories 0.8 Our model constrains the buildup of stars in the intraclus- ter light (see definition in §2.4) purely from observational galaxy data and measurements of the halo-halo merger rate 0.6 in simulations (see also Watson et al. 2012 for an alternate method). In our best-fitting model, only 5% of stellar mass 14 0.4 in mergers for 10 M! halos is allowed to be deposited onto 13 z = 0.1 the central galaxy since z =1,andonly10%for10 M! ha- z = 0.5 los. For Milky Way-sized and smaller halos, this number rises 0.2 z = 1 z = 2 rapidly to 70-80%. Yet, due to the sharply decreasing stellar z = 4 mass to halo mass ratio for lower-mass halos, most of the in- coming stellar mass will be in (rare) major mergers. Central 0 10 11 12 13 14 15 10 10 10 10 10 10 galaxies are therefore relatively uncontaminated by stars from M [M • ] Fraction of Stellar Growth from In Situ Star Formation h O smaller recently-merged satellites. At higher redshifts, how- ever, larger fractions of the stellar mass in merging galaxies FIG.10.—Thefractionofstellarmassgrowthingalaxiesduetoin situ star formation (as opposed to growth by galaxy-galaxy mergers) asafunctionof are allowed to be deposited onto the central galaxy. halo mass and redshift. In Fig. 9, we show the amount of galaxies’ present-day stel- lar mass and intracluster light (ICL)/halo stars that was in are significantly different than at lower reshifts, with more place at a given redshift. The left-hand panel (stellar mass) stellar mass per unit halo mass. However, concerns about the shows that almost all stars in the central galaxy in present-day reliability of the stellar mass functions at those redshifts(see cluster-scale halos were in place at z =2.However,sincethat §3.1) urge caution in interpreting the physical meaning of this time, a large number of their satellites have been disrupted result. into the ICL. Thus, for cluster scale halos, the stellar mass in Ausefulperspectiveontheseresultscanbeobtainedby the ICL exceeds the stellar mass in the central galaxy by a fac- considering the historical stellar mass to halo mass ratio of tor of 4-5, consistent with observations (e.g., Gonzalez et al. halos, as shown in Fig. 8. Despite the large systematic un- 2005). We note that our model predicts what may seem to be 12 certainties, it is clear that halos go through markedly dif- alargeICLfractionforMilkyWay-sizedgalaxies(10 M!). ferent phases of star formation. This evolution is most ap- However,the MilkyWay is a specialcase. It hasnothad a ma- parent for massive halos, as observations have been able to jor merger for 10-11 Gyr (Hammer et al. 2007); however, probe the properties of the progenitor galaxies all the way to as noted above,∼ only major mergers can contribute substan- z =8.Specifically,high-redshiftprogenitorsoftoday’sbright- tially to the ICL. A major merger 11 Gyr ago would have, 14 est cluster galaxies (Mh 10 M! were relatively efficient in however, contributed less than 3% of the present-day stellar converting baryons to stars—comparable∼ to the most efficient mass of the Milky Way into the ICL; allowing for passive galaxies today. However, between redshifts 2 − 3, their effi- stellar evolution, this would result in less than 2% of the lumi- ciencies peaked, and thereafter they began to form stars less nosity of the Milky Way coming from the intrahalo light, in rapidly than their host halos were accreting dark matter. At excellent agreement with observations (Purcell et al. 2007). the present day, such galaxies have an integrated star forma- Finally, in Fig. 10, we show the inferred fraction of stel- tion efficiency that is two orders of magnitude less than at lar mass growth in galaxies coming from in situ star forma- their peak. The picture is less clear for progenitors of lower- tion (as opposed to galaxy-galaxy mergers) as a function of mass galaxies because current observations cannot probe their halo mass and redshift. At all redshifts greater than 1, the progenitors as far back. Nonetheless, their apparent behavior vast majority of stellar mass growth is from star formation, The Astrophysical Journal,762:32(21pp),2013January1 Coe et al. 8 Figure 2. Three images of MACS0647-JD as observed in various filters with HST. The leftmost panels show the summed 11 hr (17-orbit) exposures obtained in eight filters spanning 0.4–0.9 µm with the Advanced Camera for Surveys. The five middle columns show observations with the Wide Field Camera 3 IR channel in F105W, F110W, F125W, F140W, and F160W, all shown with the same linear scale in electrons per second. The F125W images were obtained at a single roll angle, and a small region near JD2 was affectedCoe by persistence et due toal. a moderately 2013 bright star in- our parallelCLASH observations immediately / prior (see also Figure 4). The right panels zoom in by a factor of 2 to show F110W+F140W+F160W color images scaled linearly between 0 and 0.1 µJy. (A color version of this figure is availableMACS0647.7+7015 in the online journal.) Table 2 detection threshold. We set this threshold equal to the rms mea- Seventeen Clusters Searched in This Work sured locally near each object. Isophotal fluxes (and magnitudes) High Magnification?a Clusterb Redshift are measured within these isophotal apertures. SExtractor de- Abell 383 (0248.1 0331) 0.187 rives flux uncertainties by adding in quadrature the background − Abell 611 (0800.9+3603) 0.288 rms derived from our inverse variance maps and the Poisson Abell 2261 (1722.5+3207) 0.244 uncertainty from the object flux. MACSJ0329.7 0211when 0.450 didSince the our images lights are drizzled to a 0"".065 pixel scale, which − YMACSJ0647.8+7015 0.591 is 2–3 times smaller than the WFC3 point-spread function YMACSJ0717.5+3745 0.548 (PSF), the resulting images contain significant correlated noise. MACSJ0744.9+3927 0.686 comeThe weight on? maps produced by drizzle represent the expected MACSJ1115.9+0129 0.355 variance in the absence of correlated noise.