Lines of sight with multiple, projected, cluster-scale halos have high total masses and complex lens plane interactions that can boost the area of magnification, or etendue, making detection of faint background sources more likely than elsewhere. To
identify these new compound cosmic telescopes, we have found lines-of-sight with the highest integrated mass densities, as traced by the projected concentrations of Luminous Red Galaxies (LRGs). We use 1151 MMT Hectospec spectra to derive preliminary magnification maps for two such lines of sight with total mass exceeding ~ 3 x 10$^{15}$ Msun -- J0850+3604 (0850) and J1306+4632 (1306). We identify 2-3 group- and cluster-scale halos in each beam over 0.1 < z < 0.7, all of which are well-traced by LRGs. In Subaru Suprime-Cam imaging of beam 0850, we discover serendipitously a candidate multiply-imaged V-dropout source at z = 5.03, whose location is consistent with the critical curves for a source plane of $z_s$ = 5.03 predicted by our mass model. Incorporating the position of the candidate multiply-imaged galaxy as a constraint on the critical curve location in 0850 narrows the 68% confidence band on lens plane area with mu > 10 for a source plane of $z_s$ = 10 to [1.8, 4.2] square arcminutes, comparable to that of MACS 0717+3745 and El Gordo, two of the most powerful known single cluster lenses. The 68% confidence intervals on the lens plane area with mu > 10 for 1306 are [2.3, 6.7] square arcminutes. The significant lensing power of our beams makes them powerful probes of reionization and galaxy formation in the early Universe.
We present uncontaminated rest-frame u - R colors of 78 X-ray-selected AGN hosts at 0.5 < z < 1.5 in the Chandra Deep Fields measured with HST/ACS/NICMOS and VLT/ISAAC imaging. We also present spatially-resolved NUV - R color gradients for a subsampl
e of AGN hosts imaged by HST/WFC3. Integrated, uncorrected photometry is not reliable for comparing the mean properties of soft and hard AGN host galaxies at z ~ 1 due to color contamination from point-source AGN emission. We use a cloning simulation to develop a calibration between concentration and this color contamination and use this to correct host galaxy colors. The mean u - R color of the unobscured/soft hosts beyond ~6 kpc is statistically equivalent to that of the obscured/hard hosts (the soft sources are 0.09 +/- 0.16 magnitudes bluer). Furthermore, the rest-frame V - J colors of the obscured and unobscured hosts beyond ~6 kpc are statistically equivalent, suggesting that the two populations have similar distributions of dust extinction. For the WFC3/IR sample, the mean NUV - R color gradients of unobscured and obscured sources differ by less than ~0.5 magnitudes for r > 1.1 kpc. These three observations imply that AGN obscuration is uncorrelated with the star formation rate beyond ~1 kpc. These observations favor a unification scenario for intermediate-luminosity AGNs in which obscuration is determined geometrically. Scenarios in which the majority of intermediate-luminosity AGN at z ~ 1 are undergoing rapid, galaxy-wide quenching due to AGN-driven feedback processes are disfavored.
