! Hubble Space Telescope Cycle 6 (1996) Phase II Proposal Template ! $Id: 6610,v 12.1 1996/08/20 22:08:17 pepsa Exp $ ! Hubble Space Telescope Cycle 6 (1996) Phase II Proposal Template ! $Id: 6610,v 12.1 1996/08/20 22:08:17 pepsa Exp $ ! ! Refer to the HST Phase II Proposal Instructions to fill this out ! ! Anything after a "!" is ignored, and may be deleted ! ! All keywords with multiple entries are comma delimited except the ! Visit_Requirements and Special_Requirements keywords which can be ! delimited with carriage returns or semi-colons, but not commas ! ! For help call your Program Coordinator: van Orsow ! Phone: 410-338-4568 , E-mail: vanorsow@stsci.edu ! ! This partially completed template was generated from a Phase I proposal. ! Name of Phase I Proposal: archive-0569.windhorst.prop ! Date generated: Fri Dec 22 18:56:01 EST 1995 ! Proposal_Information ! Section 4 Title: WFPC2 Ly-alpha imaging of galaxy clusters at z=2.4: galaxy formation from compact sub-galactic clumps? Proposal_Category: GO Scientific_Category: COSMOLOGY Cycle: 6 Investigators PI_name: Rogier Windhorst PI_Institution: Arizona State University CoI_Name: Stephen Odewahn CoI_Institution: Arizona State University, Tempe Contact: ! Y or N (designate at most one contact) CoI_Name: Sam Pascarelle CoI_Institution: Arizona State University, Tempe Contact: ! Y or N (designate at most one contact) CoI_Name: William Keel CoI_Institution: Univ. of Alabama, Tuscaloosa Contact: ! Y or N (designate at most one contact) CoI_Name: Nicholas Scoville CoI_Institution: Caltech, Pasadena Contact: ! Y or N (designate at most one contact) CoI_Name: Simon Driver CoI_Institution: Univ. of New South Wales, Sydney Contact: ! Y or N (designate at most one contact) Abstract: ! Free format text (please update) We request 27+28 orbits in Cycle 6 and 45 in Cycle 7 to obtain narrow-band redshifted LyAlpha, plus B & I-band imaging of rich groups and clusters of young compact galaxies at z~eq 2.4. The goal is to obtain definite clues to galaxy formation, and to see how this proceeded from (the merging of) compact ``sub-galactic clumps'' since z~2--3. In recent deep WFPC2 images, we discovered a spectacular number of well-detected, compact ( 1 kpc) Ly-alpha emitters in F410M at z~2.4, with 3 confirming spectroscopic z's. These numerous sub-galactic objects make up a signifcant part of the faint blue galaxy counts for B~eq 23-28 mag. Our hypothesis is that they occur in the range z~1--3.5, being the building blocks of the luminous galaxies that we see today. If so, this would have far-reaching consequences to theories of galaxy formation (e.g.\ CDM) and of large scale structure. Hence, we need confirmation of the universality of these objects by observing independent WFPC2 fields, one with several known z=2.4 QSO's, and two blank sky fields. The images will be used to: (1) locate ~70 Ly-alpha emitting sub-galactic clumps at z~eq 2.37+/-0.05 within 3', determine their Ly-alpha fluxes and SFR; (2) determine their LF and constrain the luminous mass spectrum at z=2.4; (3) determine their size distribution and their relation to QSO absorption clouds; and (4) combined with CO-detections, determine the gas-to-stellar mass conversion rate and the efficiency of galaxy formation at z~2.4. Questions ! Free format text (please update) Observing_Description: \sn bullet (1) Field selection:\ The other z=2.4 cluster candidates were selected from the proximity of at least three quasars from the Hewett & Burbidge catalog with known redshifts in the range 2.35 z 2.40 and within 3--40\ ' on the sky, which corresponds to ~eq 1.8--24\hmin Mpc, so that we have a high probability to hit a ``supercluster'' or Broadhurst Etal (1988) ``spike'' in the redshift distribution at z~eq 2.37+/- 0.04. Such spikes have been observed out to z~1 in the redshift distribution of field galaxies by B88, C94, Gl95, L95, and recently for z\cge 1 by Cowie Etal (1995) and Koo Etal (1995) with Keck. The faint galaxy two-point correlation function (NW95a) suggests that the evolution of galaxy clustering has been relatively mild out to z\cge 1, so that the ``spikes'' observed in the redshift distribution out to z~1 may well exist out to z~1--3 (even if at lower correlation amplitude). In B88's redshift distribution, such spikes have a typical separation Delta z~0.043, possibly evolving (in an unknown manner) with redshift. Our WFPC2 narrow-band imaging of the z=2.40 cluster showed that such superstructures likely existed in the redshift distribution at z~2.4. The typical velocity structure and width of ``superclusters'' or spikes in the redshift distribution requires the filter's velocity width to be no less than Delta z~ 0.05, or we loose a good fraction of the z=2.4 candidates. Searches in WFPC2 filters narrower than F410M are therefore not efficient. The best fields with at least three spectroscopically confirmed quasars within ~3--40' at z~eq2.37+/- 0.05 were found in the 2340+01 field (z=2.32, 2.34, 2.35; V= 18.7--20.8 mag) and the 1308+28 field (z=2.37, 2.390, 2.420, 2.42, (2.46); V=19.8--21.5 mag). All can be covered within the 147Angstrom\ FWHM of F410M. The 1308+28 field serves as a backup, if the 2340+01 field cannot be scheduled due to the shorter duration of Cycle 6 before the SSM in 1997. \sn bullet (2) Time justification and feasibility:\ Since most significant Ly-alpha candidates in Fig. 3 have a (Ly-alphaband-\Bband) excess ~1.0 mag below the featureless power-law, we need to have errors in color less than 0.3 mag to get significant detections for each individual object. We therefore need to detect and do better than 5Sigma photometry on compact (\rhl 0"pt 2) z=2.4 Ly- alpha emitters down to m_4100~ 26.5 mag, and the associated field galaxies down to B 27.5, and I 26.0 mag (or equivalently (V 27m ag). We measured the sky-brightness to be 24.2 \magarc in Ly-alphaband, 24.1 in \Bband, 23.0 in \Vband (23.4 in V_F555W), & 22.45 in \Iband, so we can rescale from experience with our Cycle 4--5 images: we need \cge 15 orbits (11.25-hr) in Ly-alphaband, 8 orbits (6.0-hr) in \Bband, and 4 orbits (3-hr) in \Iband to get the required S/N and errors comparable to Fig. 3. Shallower exposures will not allow us to find the bulk of the Ly-alpha emitters in Fig. 3, given the apparent steepness of their LF in Fig. 4b. We propose to use the F450W for continuum subtraction, because it overlaps with the F410M completely, and because it has the highest QE* T and so gives the highest S/N and smallest photometric errors per unit time (One could use F380W or F439W for continuum subtraction, but they have lower QE* T). The \Iband exposures are necessary to compute and apply the color terms in Fig. 3 (\Vband would suffice too, but \Iband has a larger color baseline, and all WFPC2 filters are below the 4000Angstrom\ break where the young galaxy SED is relatively straight, W91). This \Iband photometry cannot be obtained from the ground, as the z=2.4 candidates are so small that aperture or isophotal photometry becomes impossible, and total magnitudes would yield too large errors for the faint compact objects. We also request two 14-orbit generic parallel exposure to survey two blank sky comparison fields: 6-hr (8 orbits) in narrow-band redshifted Ly-alpha (F410M) plus 3.0-hr (4 orbits) in \Bband and 1.5-hr (2 orbits) in \Iband for both. These parallel exposures will roughly yield 0.4 mag lower sensitivity than the 27 orbit primary exposure, and so allow to detect in principle ~24 z=2.4 candidates, which we need for a definitive test, as explained in section 2. To improve the statistics for fainter Ly-alpha emitters, we plan to extend the LF at z=2.4 (Fig. 4b) in the 17\hh+50degrees field by another 1.4 mag, thereby increasing the statistics by a factor of \cge 2, or a total of ~40 Ly-alpha candidates. We can do this in 45 orbits of 2700 sec in F410M in Cycle 7 on our 53W002 field, which will increase the total of F410M exposures on this field to 60 orbits, and the sensitivity thus by 0.3 dex~eq0.75 mag. Another gain of ~0.6 mag can be obtained by getting better on -orbit bias and dark-frames for the dithered stack of read- noise limited F410M frames (as outlined in the Special Requirements section). In total we expect ~70 Ly-alpha candidates. The total cost of this proposal is about one orbit per discovered Ly-alpha protogalactic object. The 15 orbit primary Ly-alphaband images will be substepped in a 3*5 L-shaped pattern, to allow for minimal reconstruction of a fully sampled PSF. The 45 orbit primary Ly-alpha images will be substepped in a ``nonet'' pattern of 9*5 orbits to get optimal PSF sampling. \sn bullet (3) Insufficiency of ground -based searches:\ Even in the best ground-based seeing and sky-brightness, a ~150 hour Keck narrow-band exposure would be required to match our WFPC2 point-source sensitivity in F410M. Going to narrower filters than F410M from the ground doesn't help, because the typical velocity structure and spacing of superclusters or ``spikes'' in B88's N(z) would cause one to loose a significant number of objects outside the interval z~eq 2.37+/- 0.04. Going to narrower filters with HST makes little sense for the same reasons. Hence, the search for compact Ly-alpha-emitting proto-galactic clumps cannot be done from the ground. \sn bullet (4) Processing:\ We need to subtract the continuum B-band from the Ly-alpha- image, to obtain a Ly-alpha-image at 0"pt 1 resolution. Hence, we require higher signal-to-noise in F450W, to avoid corrupting the S/N of the F410M-image. The Cycle 6 & 7 will be reduced like our Cycle 4--5 images, as in D95a, DWG95, FW95, G94a, H95, WK95a, W94b, similar to the Cycle 1--3 WF/PC reduction techniques of G94a, W92, W94a, W94c, W95a. \mn bullet (5) TEAM OPERATION IN DATA REDUCTION AND ANALYSIS:\ We will systematically analyze the Cycle 6 & 7 images according to the following plan, similar to our plans in Cycle 4--5 (see attached ASU Bibliography for results): \n (a) Make a data base and complete WFPC2 galaxy catalogs from the Ly-alpha, B, and I band images using RGASP and FOCAS software, as we did in D95a, DWG95a, O95, P95b, which will allow us to address our scientific goals. \n (b) Study catalog completeness and reliability from simulations and comparison of independent data halves on the same field (as in F93, F94, W93), run catalogs on slightly convolved WFPC2 images to assure completeness for lower SB-galaxies (as in F94, N95a, N95c). \n (c) Run FOCAS and RGASP, adapted for HST to get: accurate galaxy positions in several filters (as in N95c, NW95b), light-profiles & scale-lengths (as in W94a, Sc95), color gradients (as W94c), etc. Classify all objects with the optimized neural network package MORPHO (as in O95) that is trained on brighter galaxy samples, which were classified by eye and have some spectroscopic confirmation, as in O95. \n (e) As a team, analyze the z=2.4 data base. Details are given in the Supporting Observations section, in \S\S 1.c, 1.d & 2 of the proposal, and in W91, P95a, P95b. The team will be coordinated through regular Email, Telecons, and team meetings. Real_Time_Justification: \sn bullet (1) Scheduling:\ Depending on HST scheduling constraints, the 1308+28 field serves as a backup in case our proposed 2340+01 field cannot be scheduled in Cycle 6 due to the Second Servicing Mission and SSMOV in spring 1997. We will image the brightest z=2.4 Ly-alpha emitters in a clone of the F410M filter at Steward/Kitt Peak in late 1995/early 1996. Somewhat depending on the outcome of these supporting observations, we may have to defer the choice between the two proposed fields until the start of Cycle 6. \sn bullet (2) Pointing:\ We will place the one of the brighter z=2.4 quasars inside the PC1, and if possible, one of the others in a WFC CCD, not requiring any special spacecraft orientation. \sn bullet (3) Calibration:\ Our most important constraint is the accuracy of the standard bias and dark-calibration. Our Cycle 5 F410M images were 2700 sec per orbit (the maximum one can do before the exposure sees the Earth's limb), and had a whopping sky of 5.5 DN (these were the most read-noise limited deep CCD frames that this PI has ever seen.). For the WFPC2 read-noise of 5.3in and dark-current of 0.003in/sec at --88degrees C, the relative contributions to the image- variance should have been: sky=39.3insq, read-noise=421insq, and dark=121.5insq, but the resulting sky-variance was ~ 1.75* (~0.6 mag.) worse than this. This could not have possibly come from the F410M sky (\cge 24.2 \magarc), but is due to the limited S/N of the bias and dark-frame stacks available for the calibration. It is therefore of utter importance for all our F410M exposures to reduce the image variance induced in the Pipeline by the current lack of good enough superdarks and superbiases. We will work with STScI to plan this accordingly. We request that you specifically approve the following: We must reduce the main component --- read-noise in the superbias and darks --- by a factor of ~3. Since ~ 8 were used so far, we therefore request 64*0 sec bias frames during bright-time, and 16 sets of 1-orbit dark frames, the latter as at least two sets of 8*2700 sec darks taken during bright-time between successive SAA transitions and as close as possible in time to the F410M data to reduce the effects of time-dependent hot pixels. This will reduce our image noise by ~0.65 mag, and together with the 4* larger Cycle 7 integration time, enhance our total sensitivity by 1.4 mag, allowing us to populate Figs. 4 with the expected ~70 objects. \sn bullet (1) Ground-based spectroscopy:\ We have carried out an extensive program of ground-based follow-up on our WFPC2 field around 53W002, and will carry out a similar program for the proposed 2340+01 field (and the backup field at 1308+28). For this, we will use the MMT Red Spectrograph with multi-aperture plates and the KPNO 4m or WYIN 3.5 m spectrographs for low-resolution spectroscopy. The goal is to measure redshifts for any compact objects, including QSO's, that show up in our HST F410M and ground-based narrow-band Ly -alpha images, as well as to constrain their nature in terms of starburst and/or weak AGN. \sn bullet (2) Ground-based narrow-band and broad-band imaging:\ We purchased a nearly identical filter to F410M from Custom Scientific Inc. (who also made the WFPC2 filters) to image a much wider field in redshifted narrow-band Ly-alpha (<=ff 4120Angstrom, DeltaLambda~eq 150Angstrom FWHM) with the Steward 90 inch, Kitt Peak 4 m, and WYIN 3.5 m telescopes. The WFPC2 Ly-alpha image will provide the best resolution and sensitivity at 0.07--0.2" resolution --- inaccessible to our best ground- based Ly-alpha images --- and will be combined with the deep ground-based Ly-alpha-images (~1"pt 0 FWHM) to provide good S/N at all spatial frequencies 0.07---1.0" (using the Lucy- Hook algorithm), and be used to trace larger Ly-alpha clouds (expected to be rare, based on our ground-based Ly-alpha images so far). We also used the Steward 90 inch with a 1200* 800^2 Lesser CCD (98\% QE at 4100Angstrom\ ) for deep UBVI imaging and the Steward NICMOS 256*256 arrays for JHK imaging (W91, W94c, M94b), and used the UBVRIgriJHK photometry to constrain the effects of dust in z=2.40 candidates and provide photometric redshift-estimates (C95) for other objects with V\cge 24 mag, where redshifts cannot be easily obtained on 4m class telescopes. \sn bullet (3) Keck follow- up:\ Co-I Dr. Scoville (Caltech) has agreed to attempt Keck LRIS follow-up on galaxies ~1.5 mag fainter (B 26, V 25.5 mag) than we can reasonably do with the MMT to confirm z=2.4 candidates, isolate AGN components, study line ratios, and determine SFR's, etc. For this, we will use a grating matched to the velocity dispersion expected for these small underluminous objects. Drs. Koo & Illingworth (UCSC) have tried to spend some Keck LRIS time in 1995 on our z=2.4 candidates in the 53W002 field, but were clouded out. We will prepare a joint publication on these objects. \sn bullet (4) mm follow-up:\ Dr. Scoville has also agreed to image the field with the OVRO mm-Array in redshifted CO lines at 34 and 102 GHz to get gas-mass estimates for the most gas-rich z=2.40 cluster candidates. Deep integrations (20-40 hr) should enable detection of the gas in all galaxies with mass \cge 10^11\Mo (in which a significant fraction of the mass is still gaseous) and provide angular resolution 1-2". Precise cluster redshifts will be available for the CO observations from the MMT and Keck spectroscopy. Calibration_Justification: ! Move appropriate text from Real_Time_Justification Additional_Comments: Fixed_Targets ! Section 5.1 Target_Number: Target_Name: Alternate_Names: Description: Position: ! Most common specification format is ! RA=0H 0M 0.00S +/- 0S, ! DEC=0D 0' 0.0" +/- 0", ! PLATE-ID=0000 Equinox: RV_or_Z: RA_PM: ! Units are seconds of time per year Dec_PM: ! Units are seconds of arc per year Epoch: Annual_Parallax: Flux: ! Include at least V and B-V Comments: Solar_System_Targets ! Section 5.2 Target_Number: Target_Name: Description: Level_1: ! Satellite of Sun Level_2: ! Satellite of Level_1 Level_3: ! Satellite of Level_2 Ephem_Uncert: ! Needed for REQ EPHEM CORR sp req Acq_Uncert: ! Needed for SAVE and USE OFFSET sp reqs Window: Flux: ! Include at least V and B-V Comments: Generic_Targets ! Section 5.3 Target_Number:1 Target_Name:PAR Description:CLUSTER OF GALAXIES Criteria:ABS_II > 10D Flux: Comments: Scan_Data ! Appendix B Scan_Number: FGS_Scan: Cont_or_Dwell: Dwell_Points: Dwell_Secs: Scan_Width: Scan_Length: Sides_Angle: Number_Lines: Scan_Rate: First_Line_PA: Scan_Frame: Length_Offset: Width_Offset: ! This is a template for a single visit containing a single exposure ! Repeat exposure and visit blocks as needed Visits ! Section 6 Visit_Number:1 Visit_Requirements: ! Section 7.1 ! Uncomment or copy visit level special requirements needed ! Most of these requirements (including ORIENT) will limit scheduling ! PCS MODE [Fine | Gyro] ! GUIDing TOLerance ! DROP TO GYRO IF NECESSARY [NO REACQuisition] ! ORIENTation TO ! ORIENTation TO FROM ! ORIENTation TO FROM NOMINAL ! SAME ORIENTation AS ! CVZ PARallel ! SCHEDulability ! AFTER [BY [TO ]] ! AFTER ! BEFORE ! BETWEEN AND ! GROUP WITHIN