CATS: Optical to Near-Infrared Colors of the Bulge and Disk of Two z=0.7 Galaxies Using HST and Keck Laser Adaptive Optics Imaging
E. Steinbring, J. Melbourne, A. J. Metevier, D. C. Koo, M. R. Chun, L. Simard, J. E. Larkin, C. E. Max
aa r X i v : . [ a s t r o - ph ] J u l CATS: Optical to Near-Infrared Colors of the Bulge and Disk of Two z = 0 . GalaxiesUsing HST and Keck Laser Adaptive Optics Imaging E. Steinbring , J. Melbourne , A. J. Metevier , , D. C. Koo , M. R. Chun , L. Simard , J.E. Larkin , & C. E. Max ABSTRACT
We have employed laser guide star (LGS) adaptive optics (AO) on the KeckII telescope to obtain near-infrared (NIR) images in the Extended Groth Strip(EGS) deep galaxy survey field. This is a continuation of our Center for AdaptiveOptics Treasury Survey (CATS) program of targeting 0 . < z < Hubble Space Telescope (HST) are already in hand. OurAO field has already been imaged by the Advanced Camera for Surveys (ACS)and the Near Infared Camera and Multiobject Spectrograph (NICMOS). Our AOimages at 2.2 µ m ( K ′ ) are comparable in depth to those from HST, have Strehlratios up to 0.4, and FWHM resolutions superior to that from NICMOS. Bysampling the field with the LGS at different positions, we obtain better qualityAO images than with an immovable natural guide star. As examples of the powerof adding LGS AO to HST data we study the optical to NIR colors and colorgradients of the bulge and disk of two galaxies in the field with z = 0 . Subject headings: instrumentation: adaptive optics — galaxies: field galaxies
1. Introduction
Although the bulges of disk-dominated galaxies can appear to be the remnants of long-ago mergers, their kinematics may point to close ties with ongoing formation processes in All authors except L.S. are affiliated with the Center for Adaptive Optics. Herzberg Institute of Astrophysics, National Research Council Canada, Victoria, BC V9E 2E7, Canada UCO/Lick Observatory, Department of Astronomy and Astrophysics, University of California, SantaCruz, CA 95064 NSF Astronomy & Astrophysics Postdoctoral Fellow Institute for Astronomy, University of Hawaii, 640 North A´ohoku Place, Hilo, HI 96720 Division of Astronomy & Astrophysics, University of California, Los Angeles, CA 90095 z = 0 . . ′′ H = 70 km s − Mpc, Ω m = 0 .
3, Ω Λ = 0 .
7) which we adoptthroughout. The
Hubble Space Telescope (HST) has been used to decompose many z > . z > . V filter desirably samples rest-frame U , whichis below the 4000 ˚A break. Our initial efforts to do high-resolution NIR photometry of0 . < z < z = 0 .
2. Data and Reductions
The AO field was chosen from the EGS where HST data exist from several programs.Besides WFPC2 data in F606W and F814W (GO-5090: PI-Groth, GO-8698: PI-Mould),ACS Wide Field Camera (WFC) data in the equivalent filters and NICMOS NIC3 in F160Whave been taken (GO-10134: PI-Davis). We make use of the ACS data in our analysis, asthese are deeper and have finer pixel sampling than the previous WFPC2 data. Exposureswere 2260 s in ACS F606W, 2100 s in ACS F814W, and 640 s in NICMOS F160W, enoughto provide excellent
S/N ( >
10) in the peak pixel (0 . ′′
05 for ACS and 0 . ′′
20 for NICMOS)for galaxies brighter than I ∼
22 mag. All HST images were processed with the standardpipelines. The reduced ACS data were kindly provided to us by J. Lotz, and the reducedNICMOS data by S. Kassin. Fluxes were converted to Vega magnitudes using the zeropointsin the Data Handbooks for ACS (Pavlovsky et al. 2006) and NICMOS (Mobasher et al.2004). We will henceforth refer to the HST filters F606W, F814W, and F160W as V , I , and H , respectively.The EGS ACS data were taken in a rectangular mosaic designed to avoid bright stars.Thus, no V <
12 mag star is available to serve as a natural guide star (NGS) for the KeckAO system. Several stars in EGS are, however, bright enough (
V <
18) for tip-tilt guidingin the LGS mode, and one of the stars happens to lie within one of the parallel NICMOSpointings. Using this V = 14 star to guide tip-tilt correction, we obtained Keck II LGS-mode AO observations with the NIRC2 camera with the K ′ filter on 2 March 2005. SeeWizinowich et al. (2006) and Le Mignant et al. (2006) for more information on the Keck IILGS AO system and its operation. Individual exposures were 120 s, obtained in a nonre-peating circular dither pattern of 4 ′′ radius, for a total integration of 5640 s. For these smalldithers no re-acquisition is necessary with each laser move, and so overheads are no morethan for NGS mode. The center of the dither pattern was aligned with the center of theNICMOS field, which required an 18 ′′ offset between the laser spot and tip-tilt star. Expo-sures were obtained with NIRC2 with 0 . ′′
04 pixel wide-field mode (40 ′′ × ′′ ) which slightlyundersamples the Keck diffraction limit in the NIR (FWHM ≈ . ′′
05 at 1.6 µ m). This is agood match to NICMOS NIC3 field (50 ′′ × ′′ ) although the 0 . ′′
20 NIC3 pixels significantlyundersample the HST PSF (FWHM ≈ . ′′
16 at 1.6 µ m). Data reductions closely followed 4 –those described in our earlier NGS work (Steinbring et al. 2004). A photometric zeropointof Z K ′ = 24 . ± .
07 (from the Keck LGS AO NIRC2 webpages) was used to convert tostandard Johnson K ′ .Apart from the tip-tilt guide star, we detect seven objects with Keck AO in the regionof overlapping HST coverage. One is a star, bright but unsaturated, which is suitable asa PSF estimator for HST. Three are galaxies of known redshift, listed in Table 1. Two ofthese were measured as part of the DEEP2 redshift survey (Davis et al. 2007). A third hasa photometric redshift obtained from the Canada-France-Hawaii Telescope Legacy Survey(Ilbert et al. 2006). There is another galaxy in the ACS images of unknown, although pre-sumably low, redshift. It is only marginally detected at K ′ . No redshifts are published forthe remaining two objects either, and they are too faint in all bands for our analysis. The HST PSF is well characterized. The field distortions of ACS and NICMOS aremodest and can be modeled in detail with the Tiny Tim software (Krist 1995). The PSFStrehl ratio - the peak intensity relative to that of the ideal diffraction pattern - is high,approaching S = 0 .
86 for NICMOS NIC3 H at optimal focus (Barker et al. 2007). So tofirst order the FWHM will describe the PSF, and differences in Strehl ratio can indicate PSFvariation over the field.The HST PSF FWHM was measured on the bright but unsaturated star in the fieldusing a Gaussian fit with the IRAF task IMEXAM. The ACS WFC has a PSF FWHM of0 . ′′
11 in V and 0 . ′′
10 in I . The NICMOS NIC3 H PSF has FWHM ≈ . ′′
35. Variation inPSF FWHM for the ACS data was estimated by comparing the PSF star in the field withanother suitable star 74 ′′ to the southwest, outside the overlapping NICMOS coverage. Thiswas found to differ in FWHM by less than 0 . ′′
01 (∆FWHM = 4% in V , 6% in I ). We cannotmeasure PSF variation in the NICMOS field as there is only one unsaturated star available,but variation in PSF FWHM over the NIC3 field is also known to be slight, less than 0 . ′′ ∼ S = 0 .
02 in V and ∆ S = 0 .
09 in I over a range of 74 ′′ . The largest separationof PSF and target is 37 ′′ , so variation in Strehl ratio over our field will be less. The caseshould be similar for NICMOS H , as its PSF FWHM variation is comparable to ACS. 5 – Variations in LGS AO performance, both in time and angular separation from the guidesource, are similar to those experienced in NGS mode. See van Dam et al. (2006a) andvan Dam et al. (2006b) for a discussion of Keck II LGS AO performance characterization.Changes in seeing with time result in fluctuations in AO correction over the entire fieldof view. Increasing angular separation θ between the target and laser spot results in adrop in Strehl ratio proportional to exp θ / . It also increases radial elongation of the PSFtowards the laser spot. Although we did not record the seeing at V , previous experiencewith Keck AO indicates that for the correction achieved it was ∼ . ′′
7, and fairly stable duringthe night. Two steps were taken to track changes in AO performance and characterize thedata. First, we interleaved the scientific observations with short, unsaturated exposures ofthe tip-tilt star. This establishes the PSF at the tip-tilt guide location immediately prior toeach scientific exposure. Second, we followed the scientific observations by duplicating thedither pattern used for science observations in a dense star field. The globular cluster M5was observed, which was at comparable airmass to EGS and is familiar from our previousAO calibration work (Steinbring et al. 2002). We used a V = 12 star for tip-tilt guiding.That this was brighter than the star used for EGS should not impact the results, as V =14 is already sufficient for excellent tip-tilt correction. We applied the same offset as thescience observations and duplicated the circular dither pattern. We then shut off the laser,and continued the observations in M5 using this tip-tilt star as the NGS-mode guide forcomparison with our LGS observations.These calibration data were combined to give the frame-by-frame performance of theAO system. The Strehl ratio of the unsaturated images of the tip-tilt star and another faintstar visible in the scientific exposures were used to interpolate the “on-axis” Strehl ratio -the value at the center of each frame, coincident with the laser spot - via an exp θ / fit.This was repeated in each of the M5 fields. In the latter case, the fit was to all Strehl ratiosof isolated unsaturated stars. Measured Strehl ratios are plotted in Figure 1, along with theelongation of the PSF at the tip-tilt guide star position. Image quality improved during theobservations and peaked at a Strehl ratio of S = 0 .
38 just before we moved the telescopeto M5. These last science images also correspond to the least elongated images, and thestrong anti-correlation with Strehl ratio is mirrored in the calibration data. Strehl ratio andelongation data are plotted again in the lower panel of Figure 1 as a function of separationbetween the tip-tilt guide star and the laser spot. Note the expected trend towards increasingPSF elongation with increasing offset. However, one can see that Strehl-ratio performanceis fairly stable as a function of dither position, and so it is probably reasonable to neglectthe effect of anisoplanatism due to separation of the tip-tilt star and the laser spot. 6 –The final images of both the target and the M5 field were combined after pruning poordata. Image quality and per-pixel
S/N were found to be optimal by setting a cutoff foron-axis Strehl ratio of S = 0 .
15. This eliminated 10 of the EGS frames or about a fifth ofthe data; and similarly for M5. The Strehl ratio at the center of the combined M5 frames is S = 0 .
23, which is consistent with the interpolated mean on-axis Strehl ratio for the scienceframes ( S = 0 .
25, see Figure 1). It is worth emphasizing that this is based on a model fitto the on-axis Strehl ratio in each frame, not a measured value there. Although the truePSF field variation may differ from the model, this method allows us to uniformly compareall of the M5 and EGS frames based on the best correction expected in each. The resultsare shown in Figure 2. The variation in image quality over the combined images can bedetermined from the M5 field. Although LGS observations still suffer from anisoplanatism,the image quality across the field should be improved over our previous NGS observations.This is due to being able to move the laser spot around the field, and thus spread out thebest correction over the combined frames . To help characterize the improved performance of LGS over NGS we employed an an-alytic model of AO performance discussed in Steinbring et al. (2005). A synthetic starfieldwas generated by producing a spatially varying PSF for the location of every star in eachM5 frame. This is a simple PSF model comprising a diffraction-limited Gaussian core and aGaussian halo into which core light is scattered as the Strehl ratio diminishes. Anisoplanaticdegradation of the Strehl ratio is determined by the fit to the calibration data. These fakeframes were then combined and their Strehl ratio determined for each isolated star in thesame manner as the original data. The results are shown along the left hand column in Fig-ure 3. These are contour plots of equal Strehl ratio. Notice how the region of best correctionin the LGS observation is predicted to be larger than for NGS and closer to the center of thefield. This is due to dithering the laser spot in a broad circle around the center of the field,which is indicated by the dashed curve. The observed results are shown in the right-handpanels. The advantage of this LGS geometry is demonstrated by inferring the Strehl ratio atthe positions of the three targets in the field. The greatest disparity in Strehl ratio betweenany two target positions is roughly ∆ S ≈ . − .
14 = 0 .
03 for NGS. For LGS, correctionis both better and more uniform, with a maximum disparity of ∆ S ≈ . − .
19 = 0 . As a final reduction step, the combined NIRC2 science and M5 calibration frames wereresampled to match the ACS pixellation. No smoothing was done prior to resampling.Figure 4 shows a radial plot of the PSF star in the science frames. This has a FWHM of0 . ′′
16, measured with a Gaussian using the IRAF task IMEXAM. Although this is a radialaverage, the elongation of the PSF is slight, less than 0.2. The PSF FWHM is smaller atthe positions of the three targets: for GSS 044 6719 it is 0 . ′′
11; GSS 044 6712, 0 . ′′
14; and141811.0+523149, 0 . ′′
15. The ACS and NICMOS profiles are also plotted for comparison.The three targets are shown in Figure 5 along with the HST data. Each panel is 6 ′′ × ′′ .North is up, and east is left. The PSF star is shown along the bottom row in all bands withthe same scaling. There is a good match between this PSF in K ′ and that obtained fromM5, both having FWHM = 0 . ′′
16. Galaxy 141811.0+523149 ( z = 1 . V and I , but it is only marginally resolved in our NIRC2 K ′ image andno disk is evident. We therefore choose to exclude it from further analysis and instead focuson the two z = 0 . The ACS V and I images reveal this z = 0 .
67 galaxy to be nearly face on, with thedisk easily seen in both images. Also apparent are well defined spiral arms with bright star-formation regions. The central bulge-like core is very compact; only a few pixels across inthese images, and appears more pronounced in the I image. With 0 . ′′
20 pixels and a PSFFWHM of 0 . ′′
35, the NICMOS H image does not resolve the central core. In contrast, our K ′ galaxy image has a PSF more comparable to that of the ACS and shows a marginally resolvedcore with clear signs of the disk. Figure 6 shows color images produced by combining theACS V and I with either NICMOS H or NIRC2 K ′ . In each case the ACS images have beendegraded with a Gaussian filter to match the resolution of either the NICMOS or NIRC2PSF. Notice that the improved resolution of NIRC2 over NICMOS NIC3 helps reveal theseparate bulge and disk components of the galaxy. The ACS V and I images of this z = 0 .
70 galaxy are dramatic, clearly showing a smooth,inclined disk and an elongated core, more bulge-like in I . There is a broad dust lane towardsthe center of the disk in the I image. The NICMOS H image also shows the galaxy disk and 8 –a core which is elongated in the same direction as the disk. The clump of white pixels justnorth of center in this image is an artifact due to a poorly removed cosmic ray. The diskis also faintly visible in our NIRC2 K ′ image. The K ′ core is asymmetric, elongated in thesame sense as in ACS I .
3. Photometry
Synthetic aperture photometry was performed on the two z = 0 . ′′ aperture was used for GSS 044 6719 and a 5 ′′ aperture for the larger GSS 044 6712. Thezeropoints are those discussed in Section 2. The results are presented in Table 2. Our total K ′ magnitude for GSS 044 6719 of 18 . ± .
02 is comparable to seeing-limited photometryfrom Palomar giving K = 18 . ± .
04 (Noeske 2006). Note the similar colors of the twogalaxies, especially in H − K ′ . In order to probe the colors of the cores, we repeated thephotometry using a smaller circular aperture centered on the brightest pixel in each galaxyin K ′ . The size of 0 . ′′ z = 0 . V and I . We find that the cores of both galaxies are redder than their total light,especially GSS 044 6712 in H − K ′ .As an improvement over our small aperture we employed GIM2D (Marleau & Simard1998, Simard et al. 2002) to obtain separated bulge and disk photometry. The GIM2Dcode models a galaxy’s total flux, B/T , bulge and disk sizes and orientations, and galaxypixel center from an input image, or simultaneously from multiple images. Sersic bulge plusexponential disk models are convolved with an input PSF and directly compared to dataduring an optimization with the Metropolis algorithm (Metropolis et al. 1953). Once thealgorithm converges, 99% confidence intervals are determined via Monte-Carlo sampling ofparameter space ( N sample = 300).A method similar to that described in Steinbring et al. (2004) was applied, adopting ade Vaucouleurs bulge (Sersic index n = 4) plus exponential disk ( n = 1) model and usingthe PSFs discussed in Section 2. First, we determined the center of the galaxy in the reddestband (NIRC2 K ′ ), as that fit was least likely to be affected by dust or by star-forminghotspots. Then, because GIM2D does not allow simultaneous fits in more than two bands,we fit the galaxy structural parameters from the I -band ACS image, which has both highresolution and high signal-to-noise. Structural parameters fit were bulge and disk positionangles (PAs) and sizes (disk R scale and bulge R eff ), disk inclination i , and bulge ellipticity e .In all bands we kept the center fixed to that found in K ′ , except for our H -band fit, in which 9 –we allowed the center to wander 0 . ′′
05 in both the x andy y directions. Because the intrinsicpixel size of this image is much larger than for the images in other bands, the H -band imagewas difficult to register against the other bands at the sub-pixel level. With the galaxy centerand structural parameters fixed, the following parameters were then allowed to float in allbands: total flux, bulge-to-total flux ratio ( B/T ), and sky background.The results are shown for GSS 044 6719 in Figure 7. In the left column of panels inthis figure, the ACS V and I , NICMOS H , and NIRC2 K ′ data are shown. Best-fit modelsfor each passband are displayed in the center column, and the residual images are to theright. Note how the positive residuals from modeling the ACS V and I images highlightclumpy star formation regions that are not accounted for in our (smooth) models. Negativeresiduals in V and I may indicate significant dust in the disk. For GSS 044 6712, a dustlane is prominent, and is visible even in K ′ . In fact, the asymmetric core hindered a robustcentering of the model, preventing meaningful constraints on what could be a small bulge.Worse results were obtained by centroiding instead on the I image, which is better resolvedbut less symmetric than in K ′ (see figure 8). Masking out the dust lane in the I image didnot significantly improve the fit.The resulting structural parameters for GSS 044 6719 are shown in Table 3, quoted with99% confidence limits. The photometry of the disk and bulge components for GSS 044 6719was derived from the total fluxes and the B/T . Bulge magnitudes correspond to limitsthat have been propagated from the total magnitude measurements and their uncertainties.Total, bulge, and disk colors are presented in Table 4; errors again represent 99% confidencelimits. The redness of the bulge relative to the disk is apparent in both I − K ′ and H − K ′ ,but the former is likely a more reliable measurement due to the better match between NIRC2and ACS PSFs.Two tests were carried out on the photometry of GSS 044 6719 as a check on the colorsobtained from GIM2D fitting. The first test was to repeat the GIM2D analysis but insteadas a simultaneous fit in ACS V and I only and using PSFs derived from TinyTim. Theresults of the fit and the correponding photometry are shown in Tables 3 and 4. That theACS-only and the combined ACS, NICMOS, and NIRC2 datset agree on an upper limit of R eff = 0 . ′′ V , I , and K ′ in concentric annuli.First, the ACS I image was degraded slightly with a Gaussian filter to match the resolutionof the NIRC2 K ′ image. Then, the IRAF task ELLIPSE was used to measure photometryin annular apertures with outer radii ranging from 0 . ′′ . ′′ . ′′ I
10 –band image of the galaxy. The resulting aperture photometry is given in Table 5. Note thatonly for apertures of outer radius smaller than 0 . ′′ V − I and H − K ′ colors redderthan 1.35 and 2.55 magnitudes respectively, the total galaxy colors derived from GIM2Dfitting. This confirms that most of the red light is confined to what GIM2D finds as thebulge effective radius.
4. Discussion
The integrated ACS I magnitude of GSS 044 6719 is 21.1, which gives an absoluterest-frame B magnitude of -21.0 using the K -corrections of Simard et al. (2002). For GSS044 6712 ( I = 20 .
3) this is M B = − .
4. These M B , along with a disk scale length R scale ≈ . . < z < . σ = 24 ± s − , consistent with a disk-dominatedgalaxy. From our GIM2D fitting, we can see that while the bulge is weak in the optical(restframe UV), it may comprise 20% of the K ′ (restframe ∼ J ) light. Interestingly thisgalaxy is detected at 24 µ m (restframe 14 µ m) with Spitzer (Davis et al. 2007), indicatingsignificant dust may be present.In an attempt to understand the very red optical to NIR colors of these galaxies, andespecially the bulge of GSS 044 6719, we used the Galaxev code of Bruzual & Charlot (2003)to compare population synthesis models with the data. Galaxev constructs spectra forarbitrary star formation histories by adding together the single-stellar-population spectrafrom various ages, weighted appropriately. We produced V − I versus I − K ′ colors from thesespectra by scaling them according to the Calzetti (1997) dust attenuation curve, shifting themto z = 0 .
7, and then convolving this with the appropriate filter bandpasses. We constructedtwo models, both with solar metalicity ( Z = 0 . τ = 3Gyr) star formation rate, and the other passively evolves - that is, there is no active starformation - after a brief (50 Myr) burst. These were also run assuming a large amount ofdust ( A V = 10), for a total of four models. The results are shown in Figure 9, for ages of10 Myr, 100 Myr, 1 Gyr, 2 Gyr, 4 Gyr, and 7 Gyr (the age of the universe at z = 0 . τ model with no dust; a value of τ > . ′′ τ = 3) the same models can also hold for the othergalaxy, GSS 044 6712. However, the bulge of GSS 044 6719 can only be fit by either a τ model or a single-burst+passive model after invoking significant dust. One way to reconcilethis bulge color with the total light, which seems essentially unreddened, would be a clumpydistribution of dust in the galaxy. We could be seeing an old bulge through a dust clump in athick disk. The bulge is very small, and so any obscuration would affect its color much moredramatically than the disk. The same might be true for an AGN which, highly obscured,could contribute to these red colors, as well as produce significant mid-IR emission.The combination of the high resolution and long wavelength of our NIRC2 AO imaging,along with existing HST imaging, give us an opportunity to measure and interpret the opticalto NIR color gradients of the bulge and disk of two galaxies at z = 0 .
7. Aperture photometryshows that the centers of the galaxies are significantly redder than the outer regions. Whilewe have only explored a few stellar population models in our interpretation of this, we findthat the color gradient can be explained by two different star-formation histories for bulgeand disk, with similar ages. However, aperture photometry is much more difficult to interpretin terms of galaxy subcomponents, as the light in any given aperture is a combination oflight from the galaxy bulge and disk. Furthermore, our GIM2D analysis shows that thebulge of GSS 044 6719 is extremely small, so only the central aperture we used is likely tocontain a significant amount of bulge light.A much larger sample of galaxies is needed, but it seems we have established a pathtowards obtaining it. Employing the Keck laser in our NIRC2 AO observations has resolvedthe PSF issues that troubled our NGS work and made it particularly difficult to measureand analyze galaxy subcomponent colors. We have demonstrated a method of dithering thelaser guide star that stabilizes the PSF across the field of view during science observations.Furthermore, our method of interleaving science and PSF calibration frames, and of imaginga crowded stellar field for further calibration, allows us to much better characterize the PSFin our LGS science images than in the NGS images we analyzed in our previous work. Thisis important because LGS AO is quickly becoming a standard feature of large telescopes,opening up the sky to deep NIR imaging surveys in archival HST faint galaxy fields. Thisis the path that the CfAO Treasury Survey is pursuing.We gratefully acknowledge Matthew Barczys and Shelley Wright for their contributionsto CATS. Our thanks go to Jennifer Lotz and Susan Kassin for providing the HST imagesand Kai Noeske for providing us the Palomar K magnitude of GSS 044 6719. We appreciatethe hard-working Keck staff, especially Observing Assistant Steven Magee and Keck AOteam members Randy Campbell, Antonin Bouchez, David Le Mignant (currently at CfAO),and Peter Wizinowich, who made the huge complexity of laser operations seem (to us) like 12 –routine observations. We acknowledge the great cultural significance of Mauna Kea to nativeHawaiians, and express gratitude for permission to observe from its summit. Data presentedherein were obtained at the W. M. Keck Observatory, which is operated as a scientificpartnership among the California Institute of Technology, the University of California, andthe National Aeronautics and Space Administration. The Observatory was made possibleby the generous financial support of the W. M. Keck Foundation. This work was supportedby the National Science Foundation Science and Technology Center for Adaptive Optics,managed by the University of California at Santa Cruz under cooperative agreement No.AST-9876783. AJM appreciates support from the National Science Foundation from grantAST-0302153 through the NSF Astronomy and Astrophysics Postdoctoral Fellows program. REFERENCES
Andredakis, Y.C., Peletier, R.F., & Balcells, M. 1995, MNRAS, 275, 874Baggett, S., et al. 2002, in HST WFPC2 Data Handbook, v. 4.0, ed. B. Mobasher, Baltimore,STScIBarker, E., Dahlen, T., et al. 2007, in NICMOS Instrument Handbook, v. 10.0, Baltimore,STScIBender, R., Burstein, D. & Faber, S.M. 1992, ApJ, 399, 462Boehm, A. & Ziegler, B.L. 2007, ASP Conf. Ser., 379, 278Bruzual, G., Charlot, S. 2003, MNRAS, 344, 1000Calzetti, D. 1997, AJ, 113, 162Davis, M., et al. 2007,ApJ, 660, L1de Jong, R.S. 1996, A&A., 313, 45Farrarese, L. & Merritt, D. 2000, ApJ, 539, L9Gebhardt, K., Bender, R., Bower, G., Dressler, A., Faber, S.M., Filippenko, A.V., Green,R., Grillmair, C., Ho, L.C., Kormendy, J., Lauer, T.R. 2000, ApJ, 539, L13Gebhardt, K., Faber, S.M., Koo, D.C., Im, M., Simard, L., Illingworth, G.D., Phillips, A.C.,Sarajedini, V.L., Vogt, N.P., Weiner, B., & Willmer, C.N.A. 2003, ApJ, 597, 239 13 –Ilbert, O., Arnouts, S., McCracken, H.J., Bolzonella, M., Bertin, E., Le Fevre, O., Mellier,Y., Zamorani, G., Pello, R., Iovino, A., Tresse, L., Bottini, D., Garilli, B., Le Brun,V., Maccagni, D., Picat, J.P., Scaramella, R., Scodeggio, M., Vettolani, G., Zanichelli,A., Adami, C., Bardelli, S., Cappi, A., Charlot, S., Ciliegi, P., Contini, T., Cucciati,O., Foucaud, S., Franzetti, P., Gavignaud, I., Guzzo, L., Marano, B., Marinoni, C.,Mazure, A., Meneux, B., Merighi, R., Paltani, S., Pollo, A., Pozzetti, L., Radovich,M., Zucca, E., Bondi, M., Bongiorno, A., Busarello, G., De La Torre, S., Gregorini,L., Lamareille, F., Mathez, G., Merluzzi, P., Ripepi, V., Rizzo, D., Vergani, D.,astro-ph/0603217Koo, D.C., Simard, L., Willmer, C.N.A., Gebhardt, K., Bouwens, R.J., Kauffmann, G.,Crosby, T., Faber, S.M., Harker, J., Sarajedini, V.L., Vogt, N.P., Weiner, B.J.,Phillips, A.J., Im, M., & Wu, K.L. 2005, ApJS, 157, 175Koo, D.C., Ammons, S.M., Barczys, M., Larkin, J.E., Max, C.E., Melbourne, J., Metevier,A.J., Steinbring, E., Wright, S.A. & the CATS Team 2006, Galaxy Evolution Acrossthe Hubble Time, IAU Symp. no. 235, held 14-17 August, 2006 in Prague, CzechRepublic, S235, 233Kormendy, J. & Kennicutt, R.C. 2004, ARA&A, 42, 603Krist, J. in ASP Conf. Ser. 77, Astronomical Data Analysis Software and Systems IV, ed.R. A. Shaw, H. E. Payne, & J. J. E. Hayes (San Francisco: ASP), 349Le Mignant, D., van Dam, M.A., Bouchez, A.H., Chin, J.C.Y., Chock, E., Campbell, R.D.,Conrad, A., Doyle, S., Goodrich, R.W., Johansson, E.M., Kwok, S.H., Lafon, R.E.,Lyke, J.E., Melcher, C., Mouser, R.P., Summers, D.M., Stomski, P.J. Jr., Wilburn,C., Wizinowich, P.L. 2006, Advances in Adaptive Optics II. Eds. Ellerbroek, BrentL., & Bonaccini Calia, D. Proc. SPIE, Volume 6272, 1Maggorian, J., Tremaine., S., Richstone, D., Bender, R., Bower, G., Dressler, A., Faber,S.M., Gebhardt, K., Green, R., Grillmair, C., Kormendy, J., & Lauer, T. 1998, AJ,115, 2285Marleau, F.R., & Simard, L. 1998, ApJ, 507, 585Melbourne, J., Wright, S.A., Barczys, M., Bouchez, A.H., Chin, J., van Dam, M.A., Hart-man, S., Johansson, E., Koo, D.C., Lafon, R., Larkin, J., Le Mignant, D., Lotz, J.,Max, C.E., Pennington, D.M., Stomski, P.J., Summers, D., & Wizinowich, P.L., ApJ,625, L27 14 –Metropolis, N., Rosenbluth, A., Rosenbluth, M., Teller, A., & Teller, E. 1953, J. Chem.Phys., 21, 1087Mobasher, B., et al. 2006, “NICMOS Data Handbook”, Version 6.0, (Baltimore: STScI)Noeske, K. G. 2006, private communicationPavlovsky, C., et al. 2006, “ACS Data Handbook”, Version 5.0, (Baltimore: STScI)Pierce, M.J. & Tully, R.B. 1992, ApJ, 387, 47Peletier, R.F. & Balcells, M. 1996, AJ, 111, 2238Salpeter 1955, ApJ, 121, 161Schade, D., Lilly, S.J., Crampton, D., Hammer, F., Le Fevre, O. & Tresse, L. 1995, ApJ,451, L1Simard, L., Willmer, C.N.A., Vogt, N.P., Sarajedini, V.L., Phillips, A.C., Weiner, B.J., Koo,D.C., Im, M., Illingworth, G.D. & Faber, S.M. 2002, ApJS, 142, 1Steinbring, E., Faber, S.M., Macintosh, B.A., Gavel, D., Gates, E.L., 2005, PASP, 117, 847Steinbring, E., Metevier, A.J., Norton, S.A., Raschke, L.M., Koo, D.C., Faber, S.M.,Willmer, C.N.A., Larkin, J.E., Glassman, T.M., 2004, ApJS, 155, 15Steinbring, E., Faber, S.M., Hinkley, S., Macintosh, B.A., Gavel, D., Gates, E.L., Christou,J.C., Le Louarn, M., Raschke, L.M., Severson, S.A., Rigaut, F., Crampton, D., Lloyd,J.P., Graham, J.R. 2002, PASP, 114, 1267van Dam, M.A., Bouchez, A.H., Le Mignant, D., Johansson, E.M., Wizinowich, P.L., Camp-bell, R.D., Chin, J.C.Y., Hartman, S.K., Lafon, R.E., Stomski, P.J. Jr., & Summers,D.M. 2006, PASP, 118, 310van Dam, M.A., Sasiela, R.J., Bouchez, A.H., Le Mignant, D., Campbell, R.D., Chin, J.C.Y.,Hartman, S.K., Johansson, E.M., Lafon, R.E., Stomski, P.J. Jr., Summers, D.M., &Wizinowich, P.L. 2006, Proc. SPIE, in Advances in Adaptive Optics II. Eds. B.L.Ellerbroek & D. Bonaccini Calia, 6272, 95Vogt, N.P., Koo, D.C., Phillips, A.C., Wu, K., Faber, S.M., Willmer, C.N.A., Simard, L.,Weiner, B.J., Illingworth, G.D., Gebhardt, K., Gronwall, C., Guzm´an, R., Im, M.,Sarajedini, V., Groth, E.J., Rhodes, J., Brunner, R., Connolly, A., Szalay, A., Kron,R., & Blandford, R. 2005, ApJS, 159, 41 15 –Weiner, B.J., Phillips, A.C., Faber, S.M., Willmer, C.N.A., Vogt, N.P., Simard, L., Gebhardt,K., Im, M., Koo, D.C., Sarajedini, V.L., Wu, K.L., Forbes, D.A., Gronwall, C., Groth,E.J., Illingworth, G.D., Kron, R.G., Rhodes, J., Szalay, A.S., & Takamiya, M. ApJ,620, 595Willmer, C.N.A., Faber, S.M., Koo, D.C., Weiner, B.J., Newman, J.A., Coil, A.L., Con-nolly, A.J., Conroy, C., Cooper, M.C., Davis, M., Finkbeiner, D.P., Gerke, B.F.,Guhathakurta, P., Harker, J., Kaiser, N., Kassin, S., Konidaris, N.P., Lin, L., Lup-pino, G., Madgwick, D.S., Noeske, K.G., Phillips, A.C., Yan, R. 2006, ApJ, 647,853Wizinowich, P.L., Le Mignant, D., Bouchez, A.H., Campbell, R.D., Chin, J.C.Y., Contos,A.R., van Dam, M.A., Hartman, S.K., Johansson, E.M., Lafon, R.E., Lewis, H., Stom-ski, P.J., Summers, D.M., Brown, C.G., Danforth, P.M., Max, C.E., & Pennington,D.M. 2006, PASP, 118, 297Wyse, R.F.G., Gilmore, G. & Franx, M. 1997, ARA&A, 35, 637Ziegler, B.L., Bohm A., Fricke, K.J., Jager, K., Nicklas, H., Bender, R., Drory, N. Gabash,A., Saglia, P., Seitz, S., Heidt, J., Mehlert, D., Mollenhoff, C., Noll, S. & Sutorius, E.2002, ApJ, 564, L69
This preprint was prepared with the AAS L A TEX macros v5.2.
16 –Table 1. Targets
Coordinates (J2000.0)Name IAU R.A. Dec. z GSS 044 6719 141809.0+523200 14 18 09.05 52 32 00.3 0.67368GSS 044 6712 141808.8+523207 14 18 08.87 52 32 07.2 0 . +0 . − .
02 2 · · · Units of right ascension are hours, minutes, and seconds, and units ofdeclination are degrees, arcminutes, and arcseconds. Photometric redshift. Uncertainties are 99% confidence limits.
Table 2. Galaxy Aperture Photometry Target Aperture
V I H K ′ V − I I − K ′ H − K ′ GSS 044 6719 3 . ′′ . ± .
03 21 . ± .
02 19 . ± .
03 18 . ± .
01 1 . ± .
05 2 . ± .
02 0 . ± . . ′′ . ± .
03 23 . ± .
02 21 . ± .
03 20 . ± .
01 1 . ± .
05 2 . ± .
02 1 . ± . . ′′ . ± .
03 20 . ± .
02 18 . ± .
03 17 . ± .
01 1 . ± .
05 3 . ± .
02 0 . ± . . ′′ . ± .
03 22 . ± .
01 20 . ± .
03 19 . ± .
01 2 . ± .
04 3 . ± .
02 1 . ± . Uncertainties are 1- σ limits assuming only Poisson noise. Table 3. Model Galaxy Parameters for GSS 044 6719 B/T
Dataset Disk R scale2 i disk PA disk3 Bulge R eff2 e bulge PA bulge3 V I H K ′ (arcsec) (deg) (arcsec)ACS V and I , NICMOS H , & NIRC2 K ′ . +0 . − . . +4 . − . − . +9 . − . . +0 . − . . +0 . − . . +25 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . ACS V and I . +0 . − . . +3 . − . . +18 . − . . +0 . − . . +0 . − . . +12 . − . . +0 . − . . +0 . − . · · · · · · Structural parameters are described in Section 3. Uncertainties are 99% confidence limits. For z = 0 .
7, 1 kpc corresponds to 0 . ′′
14 on the sky assuming H = 70 km s − Mpc, Ω m = 0 .
3, and Ω Λ = 0 . Counter-clockwise from north. Independent fit in each band, with a star used as PSF. Simultaneous fit in V and I using a TinyTim model PSF. Table 4. Photometry of GSS 044 6719 from GIM2D Component
V I H K ′ V − I I − K ′ H − K ′ ACS, NICMOS, & NIRC2 data fit with I -band parametersTotal 22 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . Disk 22 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . Bulge 27 . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +1 . − . . +1 . − . . +0 . − . ACS data onlyTotal 22 . +0 . − . . +0 . − . · · · · · · . +0 . − . · · · · · · Disk 22 . +0 . − . . +0 . − . · · · · · · . +0 . − . · · · · · · Bulge 27 . +2 . − . . +2 . − . · · · · · · . +0 . − . · · · · · · Uncertainties are 99% confidence limits.
20 –Table 5. Annular Aperture Photometry of GSS 044 6719 on ACS and NIRC2 Images Aperture V I K ′ V − I I − K ′ . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . . ′′ . ′′ . +0 . − . . +0 . − . . +0 . − . . +0 . − . . +0 . − . Uncertainties are 1- σ limits assuming only Poission noise. The inner and outer radii of the annular aperture are given.
21 –Fig. 1.— Top panel shows extrapolated on-axis Strehl ratio (squares) and elongation(crosses) for each frame we observed. Science frames are indicated by filled symbols andthick crosses; calibration frames by open symbols and thin crosses. Note that approximatelytwo minutes separates each frame for the science data, with less than one minute betweeneach calibration frame. The bottom panel plots the LGS results as a function of separationbetween the tip-tilt guide star and the laser spot. 22 –Fig. 2.— Keck LGS AO NIRC2 K ′ image of the EGS field (left) and M5 PSF calibrator field(right). North is up, and east is to the left in these 49 ′′ × ′′ fields. Tip-tilt guiding wasprovided with the bright star at upper-left in the EGS field, and the corresponding centralone of three bright stars in the M5 field (indicated by ‘GS’). The latter was also used forfull NGS-mode correction (image not shown). We reproduced the laser dither pattern of thescientific observations in the calibration field. This provides surrogate PSF stars throughoutthe field. The star used as a PSF estimator in the science field is indicated by ‘PSF’. 23 – XXXXXX
Fig. 3.— Contours of equal Strehl ratio over the NIRC2 field for NGS mode (top row) andLGS mode (bottom row). The location of the NGS is marked by ‘GS’ and a dashed ringlabelled ‘L’ indicates the positions of the laser spot. Axis labels are units of arcsecondsfrom the center of the field. The left hand panels are model results. Laser observationsare expected to have less spatial variation in the PSF. The best correction should also beachieved in a more useful location - offset from the guide star, close to the center of the field.Although the models do not account for field distortion in the camera optics, they give aqualitatively correct picture. We observed a significant improvement with LGS performanceover NGS, and less variation in the PSF between target locations, each indicated by an ’X’. 24 –Fig. 4.— Keck LGS AO K ′ PSF after pruning data with Strehl ratio less than 0.15. TheHST ACS and NICMOS PSFs are shown for comparison. 25 –Fig. 5.— Images of the three targets and PSF star in each filter. North is up, east is tothe left. The field of fiew of each panel is 6 ′′ × ′′ . The surrogate AO PSFs from theM5 calibration frame are shown along the right-hand column. Note the good agreementbetween the K ′ PSF from the science field and that obtained from M5. The physical scalecorresponding to the PSF FWHM is given for each image. For z = 0 . V , I , H , and K ′ are 0.32, 0.48, 0.97, and 1.29 µ m respectively in the restframe of thegalaxy (0.24, 0.35, 0.72, and 0.95 µ m for z = 1 . V , I , andNICMOS H data (top) and ACS V , I , and NIRC2 K ′ data (bottom). ACS images have beendegraded to match the NICMOS H PSF. Note how the improved resolution of NIRC2 helpsreveal the bulge. The field of view and orientation of each panel is the same as Figure 5. 27 –Fig. 7.— Images of the data (left column), models (center column), and residuals (rightcolumn) for GSS 044 6719. The field of view and orientation of each panel is the same asFigure 5. The positive residuals in the ACS images are consistent with clumpy star formationregions in the disk of the galaxy. Similarily, negative residuals may indicate dust lanes. Thesimple two-component GIM2D model fits best in K ′ , unconfused by spiral arms and star-formation regions. Note that the NICMOS image has been rebinned to the ACS pixel scale,which makes it appear smoothed. 28 –Fig. 8.— Same as Figure 7 except for GSS 044 6712. The asymmetric core prevented areliable centering of the model. 29 – I - K ’ Aperture phot., total, A and BAperture phot., core, A and BGIM2D phot., total, A onlyGIM2D phot., disk, A onlyGIM2D phot., bulge, A onlyAnnular aperture phot., A onlySingle-burst model (no dust)Single-burst model (dusty)Tau model (no dust)Tau model (dusty)
10 Myr 100 Myr 1 Gyr2 Gyr4 Gyr7 Gyr
A A B B
No dust added to modelsDusty models
A A A A Fig. 9.— Total (black square) and core (black circle) colors from aperture photometry.Galaxy GSS 044 6719 is indicated by ‘A’ here; GSS 044 6712 by ‘B’. Colors from GIM2Dfitting for galaxy A are also shown: total (grey square), disk (grey triangle), and bulge (greycircle). The results from annular aperture photometry of galaxy A are plotted, connectedby solid lines. Error bars are shown only for the smallest aperture used, but the othersare similar, and are omitted for clarity. Overplotted are the single-burst+passive-evolutionmodel (thin dashed curve) and an exponential τ = 3 Gyr model (thin solid curve) withoutdust, and again with A V = 10 magnitudes of extinction (thick upper curves). Stars indicateages since the onset of star-formation of 10 Myr, 100 Myr, 1 Gyr, 2 Gyr, 4 Gyr, and 7 Gyr;the dot-dashed lines connect the oldest ages in the two star-formation models, delineatingthe region encompassed by different star-formation histories. The colors of both galaxies canplausibly be explained by the models, with some reddening. The red I − K ′′