%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % phase14-GO.tex (use only for General Observer and Snapshot proposals; % use phase14-AR.tex for Archival Research and % Theory proposals use phase14-DD.tex for GO/DD proposals). % % HUBBLE SPACE TELESCOPE % PHASE I OBSERVING PROPOSAL TEMPLATE % FOR CYCLE 14 (2004) % % Version 1.0, October 01, 2004. % % Guidelines and assistance % ========================= % Two documents are of particular importance for the preparation and % submission of HST proposals. The `Call for Proposals' discusses % policies and procedures, and explains how to submit a Phase I % proposal. The `HST Primer' provides a basic introduction to the % technical aspects of HST and its instruments, and explains how to % calculate the appropriate number of orbits for your Phase I observing % time requests. Both documents are available electronically from the % STScI World-Wide Web site at the Cycle 14 Announcement Web Page: % % http://www.stsci.edu/hst/proposing/docs/cycle14announce % % Please contact the STScI Help Desk if you need assistance with any % aspect of proposing for and using HST. Either send e-mail to % help@stsci.edu, or call 1-800-544-8125; from outside the United % States, call [1] 410-338-1082. % % What/How to submit % ============== % Cycle 14 proposals must be submitted electronically. Starting in % Cycle 12, a java-based software tool, APT (the Astronomer's % Proposal Tool see: http://apt.stsci.edu is the new interface for % all Phase I and Phase II proposal submissions for HST. % % THE PHASE I LATEX TEMPLATES THAT HAVE BEEN IN USE % FOR MANY YEARS WILL NOT BE ACCEPTED IN CYCLE 14. % % A Phase I proposal consists of two parts: % % * a completed APT proposal form; and % * an attached PDF file (see Chapter 9). % % Both get submitted to STScI directly from within APT. % % Please study Chapter 7, Chapter 8 and Chapter 9 of the Cycle 14 % Call for Proposals at: % http://www.stsci.edu/hst/proposing/docs/cycle14announce % carefully. Please do so well before the submission deadline, to % give the STScI Help Desk ample time to answer any questions that you % may have about the new procedures. % % LaTeX info % ========== % Most of the input you will be asked to provide must be given within % the curly braces "{}" following the Template-keywords (a Template- % keyword is preceded by a backslash: "\"). The LaTeX comment % character is the percent sign "%" -- anything you type that has a % percent sign to the left of it on the same line will be ignored. If % you need to have a percent sign appear in your text, use "\%" % instead. Experienced LaTeX users may be tempted to alter styles and % fonts to squeeze more text into the allowed space; % --- PLEASE DO NOT DO THIS! --- % % Since this is a LaTeX document, you can use the special LaTeX % characters you may need. In addition, a number of astronomical % symbols have been defined for you in the style file. A complete % list of those symbols can be found at the URL: % % http://www.stsci.edu/hst/proposing/docs/symbols.pdf % % Encapsulated PostScript % ======================= % Many LaTeX users will be familiar with ways to incorporate % encapsulated PostScript graphics into a document. If you are using % this form to prepare output for your POSTSCRIPT/PDF Submission, then % feel free to use any of these methods to add figures to your scientific % justification or to any of the questions that follow the scientific % justification. Information on how to encapsulate figures is % available at the URL: % % http://www.stsci.edu/hst/proposing/docs/postscript-guidelines/ % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % The template begins here. Do not alter the two lines that follow. \documentstyle[phase14,12pt,times]{article} \begin{document} \def\gs{\mathrel{\raise0.27ex\hbox{$>$}\kern-0.70em % Greater/squiggles \lower0.71ex\hbox{{$\scriptstyle \sim$}}}} \def\ls{\mathrel{\raise0.27ex\hbox{$<$}\kern-0.70em % Less than/squiggles \lower0.71ex\hbox{{$\scriptstyle \sim$}}}} \def\etal{{\it et al.}} \def\VEV#1{\left\langle #1\right\rangle} % 1. SCIENTIFIC JUSTIFICATION % (see Section 9.1 of the Call for Proposals) % % Present the SCIENTIFIC JUSTIFICATION for the proposed program. % This section should present a balanced discussion of background % information, the program's goals, its significance to astronomy in % general, and its importance for the specific subfield of astronomy % that it addresses. The members of the review panels will span a broad % range of scientific expertise, so you should write this section for % a general audience of scientists. % % There are page limits on the size of your PDF attachment. % % For Large GO proposals, Treasury GO proposals the page limits are: % - No more than 13 pages total. % - Any text in the `Scientific Justification' section, % may not extend beyond page 6. % - the increased limits allow 2 pages for a new section, Two Gyro % Mode Observations (see Section 9.3). This section must be used % to describe the science impact of two-gyro operations and, if % necessary, to outline an alternative observing strategy. % % For all other Regular GO and Snap proposals the limits are: % - No more than 9 pages total. % - Any text in the `Scientific Justification' section may not extend % beyond page 3. % - the increased limits allow 1 page for a new section, Two Gyro Mode % Observations (see Section 9.3). This section must describe the science % impact of two-gyro operations and, if necessary, to outline an % alternative observing strategy. % % In relation to these page limits, note the following: % % - Any pages beyond the page limits will be discarded by STScI % and will not be available to reviewers. % - The figures and tables must appear after the text of the Science % Justification. There are no limits on the numbers of figures, tables % and references in the PDF attachment. However, the total page limit % must be obeyed. % - The section on previous observations does not count towards the total % page limit. % - Your PDF attachment must be prepared with a font size of 12pt. % Do not change the font size or layout in any of the templates that % are provided by STScI. % % See Section 7.1.3 of the Call for Proposals for further information % on Page Limits. % \justification % Do not delete this command. % Enter your scientific justification here. \section{Using Type Ia Supernovae in Early-type Galaxies to Measure Cosmology} The signature project of the most ambitious cosmology projects being designed and/or built this decade (LSST/LST/Panstarrs, Dark Energy Camera, JDEM/SNAP/Destiny, and the South Pole Telescope and other SZ experiments) is the detailed, accurate measurement of the universe's expansion history, from deceleration through acceleration, to look for clues of the properties and identity of dark energy. Of the small handful of known measurement techniques (SN Ia, cluster counts, S-Z, weak lensing, and baryon oscillations), only Type Ia supernovae (SNe Ia) have actually been developed to the point of routine use. Only HST observations can provide the required signal-to-noise for those supernovae at $z\gs 1$, where the transition from deceleration to acceleration can be studied. Initial studies of the decelerating universe from both the Higher-Z Team and the Supernova Cosmology Project (Riess \etal\ 2004; Fadeyev, Aldering \etal\ 2004) clearly point to the limiting factor for both statistical and systematic uncertainties: extinction correction of the host galaxy. We propose a new approach to the measurements in this difficult decelerating redshift range. By studying ``clean'' supernovae discovered specifically in galaxy-cluster ellipticals, we can remove this primary statistical and systematic uncertainty --- and do so with a dramatically more efficient use of HST time (see Obs. Strategy section). The resulting data set also provides a deep z-band survey of $z \gs 1$ rich clusters, allowing us to take important steps in understanding other cosmological measurement techniques. In particular, cluster counting is an extremely sensitive measure of expansion history if their masses can be estimated with any reasonable precision. We propose to measure cluster masses via weak lensing from the same HST images, and also calibrate Sunyaev-Zel'dovich distance measurements to those same clusters. Finally, this data set will be used to study galaxy cluster assembly and star formation rates. We have intentionally included clusters that are X-ray-selected, optically-selected, and IR-selected for this purpose. \subsection{How problematic is the extinction correction uncertainty at $z\gs 1$?} The correction for the extinction of SNe from dust in the host galaxies is currently the single dominant source of both statistical and systematic error for SNe distances and the derived cosmological paramters. This is already true at $z \sim 0.5$, even with HST photometry, as shown in Knop \etal\ (2003), and dramatically worse at $z >1$ as shown in Figure 1. The typical color uncertainties for HST-studied $z>1$ SNe is 0.08 -- 0.1 in $B-V$, leading to uncertainties in extinction correction, after accounting for intrinsic color uncertainty, of $>$0.4 mag! This dispersion grows worse, $\sigma \approx 0.5$, after accounting for the uncertainty in the dust reddening coefficient, $R_V \equiv {A_B / E(B-V)}$, which Draine (2003) notes can vary from the fiducial value 3.1 by $\pm0.5$. (Note that the actual dispersion about the Hubble-line fit for $z>1$ SNe Ia corrected for extinction matches this 0.5 mag value.) Figure 2a shows the resulting poor constraints on the dark-energy equation-of-state parameter and its time variation, $w$ vs. $w^\prime$. These constraints do not distinguish between almost any current dark energy model. These large dispersions in extinction correction have been dealt with, e.g. in Riess \etal\ (2004), by applying a strong Baysean prior to the distribution, assuming knowledge of the dust and SN distribution in the $z>1$ host galaxies; the shaded contour of Fig.~2b results. However, such Baysean priors are necessarily one-sided (no negative reddening) and hence are known to introduce systematic biases when the error bars are larger at high-redshift than low-redshift (Perlmutter \etal\ 1999). This bias can be seen in Fig.~2b as the difference between the middle contour,and the dashed contour. This approach to the extinction analysis is also subject to other obvious sources of systematic biases, for example if the mean value of $R_V$ drifts from low to high redshift, as shown by the dotted contours of Fig.~2b. \subsection{How is this problem solved using SNe Ia in ellipticals?} Extinction systematics impose particular difficulty at high redshift because of the increased difficulty of precision color measurements of the SNe (especially when one must observe in the IR!). Until $z>1$ SN surveys achieve comprehensive extinction measurements, the only robust route to use of these high redshift SNe for cosmology is observation of elliptical galaxies where the dust correction is minimized. To check the bias of the cosmology derivation due to the possibility of a coherent systematic, it would be quite useful to build up an ``all elliptical'' Hubble diagram for SN at all redshifts. % These observations would provide a firm foundation for that goal. In Sullivan \etal\ (2003), we showed that the dispersion about the Hubble diagram for elliptical-hosted SNe is \emph{half} that of later-type galaxy hosts, primarily due to the absence of dust. Thus, SNe Ia in ellipticals are statistically each worth \emph{four times} that of SNe in spirals when making cosmological measurements -- and without the aforementioned systematics associated with extinction correction. This extra weight is even more dramatic at $z>1$ where color measurements are weaker. Without an extinction prior, the $z>1$ data in Riess et al. (2004) would give a dispersion of [???], compared with [???] for elliptical hosts, implying an elliptical advantage of [???] to 1. \subsection{How is it known that the targeted $z \gs 1$ cluster ellipticals are dust-free?} Nearby elliptical galaxies are well known to be almost entirely free of dust, with very few exceptions (like the disk in the center of Cen A). The clearest line of evidence for the lack of dust in elliptical galaxies comes from the tightness of the color-magnitude relation. The dispersion in the colors of early-type galaxies has long been known to be very small in clusters (Bower \etal\ 1992; Ellis \etal\ 1997; Stanford \etal -- WHAT IS THIS REFERENCE???). In fact, this relation has recently been shown by Hogg \etal\ (2004) to be universal for early-type galaxies in clusters and in lower-density environments. Recent results from ACS imaging suggests that these arguments for dust-free ellipticals hold true at redshifts $z\sim 1$. ACS imaging of RDCS1252-29 at $z=1.24$ by Blakeslee \etal\ (2003) found an intrinsic dispersion of $0.024 \pm 0.008$ mag for 30 ellipticals in the F775W - F850LP color, which approximates rest-frame $U-B$. This dispersion is comparable to that found by Bower \etal\ in Coma. Given the high likelihood of at least some age (and color) variation in the stellar populations of the member galaxies, the maximum amount of dust that could be in these ellipticals must be very small -- unless a conspiracy gives all these ellipticals the same amount of dust along the line of sight. %The Bower et al. number for the dispersion in U-V for Coma early-types was 0.04. In addition, the HST images themselves will provide another sanity check as to the dust-free nature of the supernova hosts. Elliptical galaxies are well-fit by 2-dimensional de Vaucouleur profiles. Subtracting such model fits from ellipticals reveal when dust lanes are present, as in Cen A. \subsection{Why is this cluster search much more efficient at finding (and studying) SNe Ia at $z > \sim 1$ than the previous HST searches in the GOODS fields?} This search centered on rich clusters is expected to yield twice as many supernovae in total as compared to a blank-field searches (such as the previous GOODS searches), and \emph{five times} as many supernovae in elliptical hosts. The number of Type Ia supernovae scales with luminosity, with a rate of approximately one per year per $10^{12} L{\cdot}$ in rest-frame B-band luminosity. [CORRECT NUMBERS???] Our cluster fields double the (rest-frame) luminosity in an ACS pointing, which implies the doubling of the expected number of supernovae. More importantly, this increased luminosity is primarily from early-type galaxies. Since only 1 in 5 Type Ia supernovae were discovered in early-type galaxies in the GOODS searches, our rate of SNe in early-type hosts will be increased by a factor of five. These rates are consistent with the rate seen in previous searches of clusters below $z=1$, including one we (SCP) performed with ground-based telescopes (Perlmutter \etal\ 1995; Pain 1997 [which ref???]), and the search of archival HST data by Gal-Yam \etal\ (2002, who rediscovered one of our cluster SNe). In addition to discovering more SNe Ia per HST orbit, there is another set of major efficiency gains in the {\it follow-up} of these SNe due to the knowledge that the host is elliptical, our knowledge of the redshift of the cluster, and the higher rate per field. First, the follow-up observations need significantly less signal-to-noise and fewer bands, since the extinction correction no longer dominates the requirements, and the bands can be pre-chosen to match each cluster's known redshift. Second, the higher discovery rate allows pre-scheduling observations for each cluster with a cadence guaranteed to well sample the lightcurves for every supernova discovered. For all but the highest redshift clusters, this eliminates the need for expensive TOO follow-up observations! \subsection{What is the optimal redshift range, and what limits on $w_0$ and $w^\prime$ are expected?} The two key goals of the SN Ia work at these redshifts is (1) to map the expansion history back in time to the epoch in which matter dominated over dark energy and the universe was decelerating, and (2) to use this epoch to begin to constrain the time derivative of the equation-of-state of dark energy, $w^\prime$, together with its value today, $w_0$. These measurements are most sensitive beyond $z \gs 1$, but not much more is gained above $z \sim 1.7$ since then the matter density completely dominates. In this $z=0.9-1.7$ redshift range, gravitational lensing enters as a systematic. Mass fluctuations along the light path amplify and deamplify the magnitude distribution of standard candles, distorting the Hubble diagram. Holz \& Linder (2005) show this is not a dominant source of error currently, but the sample observed here will serve as a valuable testbed for Monte Carlo techniques treating the influence of gravitational lensing on the derived cosmological model (due to the non-Gaussianity of the lensing effect, one cannot simply average SNe flux over large redshift bins). We have therefore chosen the richest clusters in this redshift range for this search, and with the resulting $\sim$10 cluster SNe Ia the statistical and systematic uncertainties due to extinction can be reduced to the level that the dashed contours of Fig.\ 1 can be ``recovered.'' \section{Developing other Dark Energy Techniques using these Cluster Data} The number of clusters of galaxies as a function of mass and redshift provides a powerful alternative probe of the cosmology thanks to its sensitivity to the comoving volume {\it and} the growth of large scale structure. The rare, massive clusters at high redshifts provide most of the discriminating power of the experiment, requiring large surveys of the sky to search for these systems. The regime $z=0.9--1.7$ is a key epoch when cluster potentials first begin decaying, as matter domination wanes and the effects of accelerating dark energy make their first appearance. A number of large cluster surveys are underway (e.g., Red-sequence Cluster Survey) or will start in the near future (e.g., South Pole Telescope Sunyaev-Zel'dovich Survey, Dark Energy Survey). These surveys, which detect clusters through a variety of techniques, all require estimates for the cluster masses in order to succesfully constrain cosmological parameters. None of the observables (richness, luminosity, temperature, etc.) measure mass directly, although masses can be inferred under certain assumptions (e.g., hydrostatic equibrium, spherical geometry). Nevertheless, many cluster properties correlate with mass. Unfortunately it is not (yet) possible to predict these relations to the required level of accuracy, because of the complex nature of cluster formation. This is more of an issue at higher redshifts, where clusters are dynamically young. Instead the mass-observable relations need to be calibrated empirically. Moreover, it is important to measure the {\it evolution} in these relations. Weak gravitational lensing of background galaxies provides a {\it direct} measurement of the projected cluster mass, without any assumptions about the geometry or dynamical state of the cluster. The proposed observations enable us to determine masses of clusters in a redshift range which is virtually unexplored. In conjunction with the X-ray, SZ and optical properties, the proposed observations enable us to extend the mass calibrations out to high redshifts. In addition, such a comparison will provide invaluable insights on cluster formation, at an era when clusters were young. \subsection{Weak lensing estimates of cluster masses} The next generation cluster surveys will yield samples of order $10^4$ clusters out to high redshifts. It will be impossible to obtain weak lensing masses for all of them. Fortunately, recent work by Majumdar \& Mohr (2003; 2004) on "self-calibration" has shown that follow-up studies of only a moderate sample of clusters are sufficient to allow these surveys to constrain $w$ better than $5-10\%$. The main requirement of such a sample is that the observations can constrain the evolution in cluster properties. Out to $z = 0.7$, large ground based projects (e.g., Canadian Cluster Calibration Project, CFHT Legacy Survey, Subaru observations) will provide a detailed study of the mass-observable relation and its scatter using $\sim 10^3$ clusters. Members of our team are leading some of these efforts. At higher redshifts, space-based observations are needed, driven by the need to measure the shapes of resolved {\it background} galaxies. The proposed observations target a redshift range which is virtually unexplored. Without the proposed observations, our knowledge of the cluster properties at $z>1$ will be too limited to use these systems for cosmology. Consequently, observations in this regime are desparately needed if cluster abundance studies are to succeed. As shown in Figure~???, the proposed observations enable us to determine masses for these high redshift clusters with a relative uncertainty of $\sim 25\%$ or better (depending on mass and redshift). The figure also shows that the zero-point of the mass-observable relation can be determined with sufficient accuracy out to $z\sim 1.3$! The proposed sample is larger by a factor of 4 from previous work, and targets clusters at higher redshifts. In addition, the clusters have been selected based on either their X-ray emission or the overdensity of red galaxies. The accuracy in the mass determination of individual clusters is sufficient to address the important question whether different detection methods select different populations of galaxy clusters. We will complement the sample proposed here with data already in the HST archive to extend the range in mass and redshift. In combination with our ground based efforts we will be able to study the evolution in the properties of galaxy clusters from the present day out to $z\sim 1.4$, thus paving the way for cluster abundance studies in an era of precision cosmology. % %The number of clusters of galaxies as a function of mass and redshift %provides a powerful alternative probe of the cosmology thanks to its %sensitivity to the comoving volume {\it and} the growth of large scale %structure. The rare, massive clusters at high redshifts provide most %of the discriminating power of the experiment, requiring large surveys %of the sky to search for these systems. A number of large cluster %surveys are underway (e.g., Red-sequence Cluster Survey) or will start %in the near future (e.g., South Pole Telescope Sunyaev-Zeldovich %Survey, Dark Energy Survey). % %These surveys, which detect clusters through a variety of techniques (from weak %lensing, X-ray emission, and SZ-decrements) all require estimates for the %cluster masses in order to succesfully constrain cosmological parameters. %Although many of the proposed observables correlate with mass, it is not (yet) %possible to predict these relations to the required level of accuracy, because %of the complex nature of cluster formation. This is more of an issue at higher %redshifts, where clusters are dynamically young. Instead, the mass-observable %relation needs to be calibrated empirically. % %%%%%%%%%%%%%%%%%%%%% The following is from Adam Stanford % %As the largest gravitationally bound structures in the universe, the properties %and histories of galaxy clusters are highly sensitive to the physics of cosmic %structure formation and to the values of the fundamental cosmological %parameters (Eke \etal\ 1996; Bahcall \etal\ 1997). While clusters in such %well-defined samples as the RDCS have been used to constrain $\Omega_m$ and %$\sigma_8$ (Borgani \etal\ 2001), the uncertainty in the relation between %cluster mass and measurables such as $L_x$ or $T_x$ limits the precision %obtainable on cosmological parameters to $\sim$50\%. To improve on this, a %better understanding of the mass--$T_x$ relation and its evolution is %necessary. To that end, we need to study in detail a well-defined sample of %clusters as a function of redshift with independent measures of the cluster %mass based on weak lensing (Tyson, Wenk, \& Valdes 1990; Lombardi \etal\ 2005) %and $T_x$ (Evrard \etal\ 1996) and on the total stellar mass (Lin, Mohr, \& %Stanford 2004). %%%%%%%%%%%%%%%%%%%% The following is from Eric Linder % %In order to develop the cluster mass function (cluster count) method %as a cosmological tool, it must be tried and proved in a variety of %environments and redshifts. The regime $z=0.9-1.7$ is a key epoch %when cluster potentials first begin decaying as matter domination wanes %and the effects of accelerating dark energy make their first appearance. %Understanding of this epoch will be crucial for understanding structure %formation, cluster properties, application of the integrated %Sachs-Wolfe effect, and eventually lensing of the cosmic microwave %background and the associated B modes. This proposal not only directly %leads to knowledge of clusters but also through important additional %information: preheating or entropy injection (through SN explosions) %is a major systematic in the use of the cluster mass function. \subsection{Weak lensing estimates of cluster masses} %\par Volume-redshift relation for clusters. %\par What is our target error for the masses? %\par Explain why this is a good sample. It spans redshifts 0.9 to 1.4, % and spans selection techniques (not just x-ray selected clusters). %\par What is the tie-down to clusters in the local universe? %\par HST is the only viable instrument for doing this at z=1! %\par Question: Can we use WFPC2 in parallel mode for measuring the % shears 5 arcmin away from the cluster center? %\par In Section 9.1: Discuss observing strategy, depth, FOV, bands. % % %Out to redshifts of $z =0.7$, large ground based projects (e.g., Canadian Cluster %Calibration Project, CFHT Legacy Survey, Subaru observations) will %provide a detailed study of the mass-observable relation and its %scatter. At higher redshifts, space based observations are needed, %driven by the need to measure the shapes of resolved {\it background} galaxies. %Constraints on cosmology are robust provided we can determine the %mass-observable relation within $\sim 5-10\%$ (Haiman, Mohr \& Holder %2001). As described below, the proposed observations enable us to %determine masses for these high redshift clusters with a relative %uncertainty per cluster of 25\% or better (depending on mass and redshift). In %combination with archival data and ground based efforts, we can %reach the required goals. [NEED MORE HERE!!!???] \subsection{Cluster masses from X-rays} %%%%%%%%%%%%%% The following is from Erica Ellingson X-ray observations of the intra-cluster medium also provide a robust method of estimating the virialized mass from clusters, which can be tied to cosmological models either via the X-ray luminosity function (e.g., Gioia 1991, Vikhlinin \etal\ 1998, Rosati \etal\ 2001, Nichols et al., 1999 and others) or the X-ray gas temperature (Henry \& Arnaud 1991, Eke \etal\ 1998, Donahue \& Voit 1999, Henry 2000). %These results form a well-established basis for %constraining cosmology via the growth of large mass structures. However, these methods are observationally very expensive, with current surveys to locate clusters at $\sim1$ limited to small areas or only the most luminous clusters (e.g., Pierre \etal\ 2004, Ebeling \etal\ 2001). Follow-up observations to determine cluster temperatures, providing a more robust mass estimate than luminosity alone, are more costly still. We thus rely on X-ray observations of a subsample of clusters as important calibrators for this method. Currently, ?? of our proposed X-ray selected clusters and ??? of our optically selected clusters have available X-ray measurements via XMM or Chandra that are sufficient to determine cluster temperatures and masses to $15\%$ accuracy. We will vigorously pursue additional observations to create a calibration sample of at least ??? clusters spanning our complete redshift and mass range. %Chandra observations of cluster morphology also provide %important information about the core mass distributions and %dynamical state of the clusters. These observations are important %in the interpretation of strong lensing results, and investigating %any discrepanies between weak lensing and virial mass %estimators. The dynamical state of the cluster may also provide an important %parameter in understanding the cluster galaxy stellar populations %(e.g. Biviano \etal\ 2003) and the history of star formation as %cluster-scale stuctures are assembled. \subsection{Sunyaev-Zeldovich estimates of cluster masses and $D_A$} %%%%%%%%%%%%%% The following is from Kyle Dawson Using Sunyaev-Zel'dovich (SZ) observations along with X-ray observations of a single galaxy cluster, one can decouple the dependence of the signals on electron density and determine the angular diameter distance ($D_A$) to the cluster. To date, the method has been used to determine distances to 26 different clusters. A fit to the sample of SZE and X-ray determined distances yields a Hubble constant $H_0 = 60 \pm 3$ km/s/Mpc, where the error bars represent statistical uncertainty assuming a $\Omega_M=0.3$, $\Omega_{\Lambda}=0.7$ cosmology. Systematic uncertainties are expected to be of order $30\%$, clearly dominating the statistical uncertainty (Reese \etal, 2002). The SZ Array (SZA) is one instrument in the next generation of dedicated telescopes for observing galaxy clusters and determining SZ cosmological-distance measurements. However, this method has never been calibrated against other distance measurements. The SN Ia measurements described in this proposal will provide a sample of distance-calibrated high redshift galaxy clusters which will be be targeted in future SZA observations. %The sample is being studied extensively in X-ray observations, which will be %used in conjunction with the SZ measurements to provide an independent distance %measurement to the cluster. Comparing the SZ/X-ray distance measurements to the SNe distance measurements, we can determine the extent of systematic uncertainties in the two methods. \section{Galaxy Cluster Science at $z \gs 1$} %\par Luminosity function and morphology-density relation %\par Fundamental plane (requires high-S/N spectroscopy too) %\par Luminosity evolution (M/L ratios) for cluster galaxies %\par Relation of everything to cluster mass! (Note that our present ignorance of some cluster properties does not impact upon the main science driver with SNe.) %%%%%%%%%%%%%%%%%%%% The following is from Adam Stanford The ACS GTO team has made significant progress in understanding moderate-$z$ galaxy clusters, which are efficient sites for the study of galaxy evolution. Postman \etal\ (2005) present results based on morphological analyses of ACS imaging of 7 clusters at $0.8 < z < 1.3$. They confirm previous work by Smith \etal\ (2004 -- REF???) that a morphology--density relation exists at $z \sim 1$, but that the slope of this relation flattens from the present epoch out to high redshift. Moreover, Postman \etal\ find that the fraction of ellipticals does not change up to $z \sim 1.25$, and that the fraction of S0s remains roughly constant at the 20\% level seen in $0.4 < z < 0.5$ clusters. Hence, the formation of all massive cluster ellipticals, and a significant fraction of the lenticulars as well, must be occuring at $z > 1.25$. This proposal will take the next step beyond the limits of the GTO program in terms of small sample size and redshift to probe the regime in which massive cluster galaxies are being formed. The GTO program has three clusters as $z>1$, to which our sample adds 20. The highest redshift of the clusters is pushed from $z=1.26$ to $z=1.5$. In conjunction with data we have obtained on our sample clusters from Chandra, XMM, Keck, the VLT, Magellan, and Spitzer, the ACS data to be obtained in this proposal will provide the ability to: {\bf 1)} determine the epochs of the assembly of ellipticals and the origin of S0s; {\bf 2)} measure the mass function of distant clusters and calibrate the relation between mass and X-ray luminosity and temperature at unprecedented redshifts; and {\bf 3)} compare the properties of clusters selected in the optical, NIR, and X-ray at redshifts $z > 1$ where clusters are first forming. %\subsection{Galaxy Evolution and Clusters} % %Progress in understanding galaxy evolution in clusters is being driven %by the need to reproduce simultaneously the lower number fractions of %early-types towards higher-$z$ (Dressler et al.\ 1997; van Dokkum et %al.\ 2001; Postman et al.\ 2005), strong homogeneity and slow %evolution in the stellar populations of ellipticals and S0s in %moderate redshift clusters ( Aragon-Salamanca et al.\ 1993; Ellis et %al.\ 1997; Stanford, Eisenhardt, \& Dickinson 1998; van Dokkum et %al. 1998a; Kelson et al.\ 2000; van Dokkum \& Stanford 2003; Blakeslee %et al.\ 2003; Holden et al.\ 2004; Postman et al.\ 2005), the %morphology-density relation (Postman et al.\ 2005), and galaxy %downsizing (Kodama et al.). % % %A self-consistent explanation of these results can be achieved by %invoking an observational bias: the progenitors of the youngest, %low-redshift early-types drop out of samples constructed in high %redshift clusters. van Dokkum \& Franx (2001) have shown the way that %morphological evolution at $z \lesssim 2$ coupled with star formation %at $z \sim 2-3$ ($\Omega_m = 0.3$, $\Lambda = 0.7$) can explain the %results seen in Figure 2. More complex semi-analytic models of %galaxy formation and evolution set in a CDM universe naturally predict %the morphological evolution that is a central tenet in the paradigm of %cluster galaxy evolution discussed here: ellipticals are formed by %mergers at higher redshift (e.g., van Dokkum et al.\ 1999), and the %Butcher-Oemler effect is the result of spirals being converted into %S0s at moderate redshifts (Dressler et al.\ 1997). The models of %Kauffman \& Charlot (1998) are able to match the small scatter in the %color-mag relation of the early-types, and even the $M/L$ evolution %found by e.g.\ van Dokkum et al.\ (1998) can be predicted (Diaferio et %al.\ 2000). % %However the accuracy of this scenario for galaxy evolution in clusters %is open to debate. In the case of early-types, the ideas are based %heavily on observations of only a few high-z clusters (e.g., MS1054 %and RDCS0848+4453). For disk galaxies, the details of when field %spirals fall in to a cluster and have their star formation stopped are %essentially unknown. Several types of critical tests must be %performed on multiple, well-defined cluster samples that span a range %of cluster mass at high redshift (mainly $z > 1$), where qualitative changes in the %cluster galaxy population seem to be occuring: 1) determine the %relative fractions of ellipticals and S0s (separately), and of disk %galaxies to investigate dependencies %on cluster mass, position within the cluster, and redshift; 2) measure %the early-type $M/L$ ratios (now possible up to $z \sim 1.3$ [van %Dokkum \& Stanford 2003; Holden et al.\ 2005] in conjunction with Keck+VLT %spectroscopy) using the intercept of the fundamental plane (FP) in %order to determine if there is variation in the galaxy $M/L$ ratios at the %same redshift; and 3) precisely measure the slope and intrinsic scatter in the %color-mag relation in a larger number of high-$z$ clusters to %determine if there are any variations with respect to cluster mass or %dynamical state . \subsection{Supernovae rates and star formation} \par The rate of supernovae in galaxy clusters will address oustanding questions about the intracluster medium (ICM). Specifically, the high metal abundances and the high energetics of the ICM are as-yet unexplained. The metals seen in luminous elliptical galaxies is explained by Type Ia supernovae rates and mass loss from evolving stars (which were originally enriched by Type II supernovae). However, the mystery arises when one looks at the iron abundances in the ICM. The iron produced from Type II SNe of a Salpeter IMF do not produce enough iron, nor do Type I SNe. It has been suggested by Brighenti \& Mathews 1999 that a higher rate for Type Ia supernovae can explain the ICM metal abundances as well as the ``entropy floor'' seen in the X-ray gas (e.g., Lloyd-Davies, Ponman \& Cannon 2000). The only observational measures of SN rates in $z \sim 1$ clusters is from the work of Gal-Yam \etal\ (2002). Searching the HST archive for repeated WFPC2 pointings of galaxy clusters, they identified 2 or 3 supernovae in 18 pointings (with overlaps of 1.3 to 4.7 square arcminutes). The derived SNe rate of $0.41^{+1.23}_{-0.39} h^2_{50}$ SNu in clusters at $z\sim 1$ argues against SNe Ia being the dominant source of iron in the ICM. However, these statistics from a mere 2 or 3 detections could clearly use improvement. \subsection{Strong lenses at $z \gs 1$, and the possibility of finding strongly lensed supernovae} %{\bf Studies of gravitational lensing arcs and cluster structure:} %%%%%%%%%%%%%%%%% The following is from Eric Linder and Joe Hennawi Massive clusters are rich test grounds for understanding of structure formation and cosmological geometry through gravitational lensing of background galaxies. The probability of formation of giant arcs has been found to be greatly enhanced, beyond expected, around clusters (see, e.g., Gladders \etal\ 2003; Ho \& White 2004), showing the need for increased understanding of their mass distribution. As one eye-opening example, Broadhurst \etal\ (2004) find over 130 images of 35 multiply lensed galaxies behind Abell 1689 using ACS imaging. This includes radial arcs and near-central images. Our target list includes one already known high redshift lensing cluster, RCS2319+0038 at z=0.91 (Gladders \etal\ 2003), exhibiting three dramatic giant arcs. With deeper z-band imaging, and the advantage of ACS for observing long, thin, low surface brightness giant arcs, this cluster will likely become a the poster child for high-redshift cluster lensing. Higher redshift cluster lensing will be even more useful in constraining cluster mass profiles and using arcs for geometric measures of cosmological distances. Availability of X-ray and other wavelength imaging of these clusters would provide much needed information for understanding the role of clusters in cosmology. Finally, cluster lenses act as natural gravitational telescopes which facilitate the detection and study of faint, high redshift galaxies (e.g., Metcalfe \etal\ 2003; Pello \etal\ 2004). In particular, if our z-band images are combined with deep i-band data, the i-z color can be used to search gravitationally lensed high redshift galaxies out to $z\sim 7$. Such lensing leads to the possibility of finding strongly lensed supernovae and extremely distant supernovae, otherwise to faint to be detected (Gunnarsson \& Goobar 2003). By combining the observed (lensed) supernova brightness with the weak and strong lensing cluster information, important bounds may be set on both cosmological parameters and the mass profile of the lensing cluster. \section{Conclusions} The observations proposed here will comprise a springboard for a wide array of astrophysical investigations: high redshift supernovae, cluster profiles, gravitational lensing, and multiwavelength studies of large scale structure and cosmology. These are the opening steps in bringing to maturity cosmological methods of the next generation, and the data will serve as a bedrock scientific legacy for extragalactic astronomy. \noindent {\bf References} Blakeslee, J.P. \etal\ 2003, ApJL, 596, 143 \\ Brighenti, F. \& Mathews, W.G. 1999, ApJ, 515, 542 \\ Broadhurst, T.\ \etal\ 2004, ApJ [astro-ph/0409132] \\ Cardelli, J.A., Clayton, G.C. \& Mathis, J.S. 1988, ApJ, 329, 33 \\ Dalal, N., Holder, G., \& Hennawi, J.\ 2004, ApJ, 609, 50 \\ Draine, B.T. 2003, ARA{\&}A 41, 241 [astro-ph/0304489] \\ Ebeling \etal\ 2001 Fontana, A. et al. 2000, AJ, 120, 2206 \\ Gal-Yam, A., Maoz, D. \& Sharon, K. 2002, MNRAS, 332, 37 \\ Gunnarsson \& Goobar 2003, A{\&}A, 405, 859 \\ Ho, S. \& White, M. 2004, submitted to ApJ [astro-ph/0408245] \\ Gladders, M.D. \etal\ 2003, ApJ 593, 48 [astro-ph/0303041] \\ Gladders, M.~ D., Hoekstra, H., Yee, H.~K.~C., Hall, P.~B., \& Barrientos, L.~F.\ 2003, ApJ, 593, 48 [astro-ph/0303041] \\ Haiman, Mohr \& Holder 2001 \\ Hennawi, J. \etal\ 2005, in prep. \\ %Holz, D.E. \& Linder, E.V. 2005, submitted to ApJ [astro-ph/0412173] \\ Hogg, D.W. \etal\ 2004, ApJ, 601, 29 \\ Kolatt \& Bartelmann 1998, MNRAS, 296, 763 \\ Linder, E.V. \& Miquel, R. 2004, Phys.\ Rev.\ D 70, 123516 [astro-ph/0409411] \\ Lloyd-Davies, E.J., Ponman, T.J. \& Cannon, D.B. 2000, MNRAS, 315, 689 \\ Metcalfe, L. et al. 2003, A{\&}A, 407, 791 \\ Majumdar, S. \& Mohr, J.J. 2003, ApJ 586, 603 \\ Majumdar, S. \& Mohr, J.J. 2004, ApJ 613, 41 \\ Pell{\' o}, R., Schaerer, D., Richard, J., Le Borgne, J.-F., \& Kneib, J.-P.\ 2004, A{\&}A, 416, L35 \\ Perlmutter, S. \etal\ 1995, ApJ, 440, 41 \\ Pierre \etal\ 2004 \\ Postman, M. \etal\ 2005, ApJ, in press [astro-ph/0501224] \\ Reese, E.D., Carlstrom, J.E., Joy, M., Mohr, J.J., Grego, L., and Holzapfel, W.L. 2002, ApJ, 581, 53\\ Tripp, T.\ 1998, A{\&}A, 331, 815 \\ %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 2. DESCRIPTION OF THE OBSERVATIONS % (see Section 9.2 of the Call for Proposals) % % Please provide a short description of the proposed observations. % Explain the amount of exposure time and number of orbits requested % (e.g., number of objects, examples of exposure-time calculations and % orbit estimates for some typical observations). You should summarize % your target acquisition strategies and durations where relevant. For CVZ % targets, state the number of CVZ opportunities available in the cycle. % % Discuss and justify any non-standard calibration requirements. You should % estimate the number of orbits required for these special calibrations, and % include them in the Observation Summary section of APT. % % Depending on the type of proposal, the following items should also be % included: % % * Long-term projects should provide summary information for the entire % project, along with a cycle-by-cycle breakdown of the requested % spacecraft orbits. % * Treasury Programs should discuss the data products that will be made % available to the community, the method of dissemination, and a realistic % time line. It is a requirement that data products be delivered to STScI in % suitable digital formats for further dissemination via the HST Data Archive % or related channels. Any required technical support from STScI and % associated costs should be described in detail. % % Proposers submitting Large and Treasury Programs should discuss % how they have designed their program with regard to schedulability. % % - All proposers should include the number and durations of the % schedulability windows for each observation. % - Proposers of programs with timing constraints and timing relationships % between observations should describe those constraints, % including allowable flexibility. % - Proposers of programs containing large blocks of orbits at constrained % orientation angles, such as mosaics and surveys, should % describe those constraints and allowable flexibility. % - Proposers requesting large blocks of orbits with a single target % should estimate how many orbits/day will be required for their % programs, during the scheduling windows that meet their timing % and orientation restrictions. % - If a target visibility other than that indicated on the Large Program % web page is used, state the value used and explain why you used % that value. % * Calibration proposals should present a detailed justification of how they % will achieve the goals of the program, and if applicable, a description of % the conditions under which these goals will be achieved. % * Calibration proposals should discuss what documentation, and data % products and/or software will be made available to STScI to support % future observing programs. % * Proposers submitting Large and Treasury programs are asked to % include additional technical detail to provide information on the % scheduling aspects of their program. % \describeobservations % Do not delete this command. % Enter your observing description here. \noindent {\bf The supernova observing strategy} The high rate of SNe Ia produced by the clusters {\it in addition} to the SNe Ia in the fore/background field galaxies makes possible an elegantly simple combined search-and-follow-up scheduling strategy. Generally, we observe each cluster with the ACS F850LP filter every $\sim$24 days for $\sim$8 visits, with the exact cadence depending on the cluster redshift and the exact number of visits depending on HST observing constraints for the cluster. This means that every supernova that appears in these clusters during the entire period, modulo small end effects, will have a lightcurve measured --- without the need for expensive Target of Opportunity (ToO) observations. We obtain ground-based spectroscopy of the supernovae's host galaxies after the fact, for precise redshifts and confirmation of elliptical galaxy type, which also confirms the SN Ia identification. For only the three highest redshift SNe, we supplement this basic lightcurve with an additional ($>$2-week-advance-notice) ToO follow-up observation sequence with NIC2 F110W, to provide restframe $B$ band lightcurve data --- since the NICMOS fields are too small to be used in the pre-scheduled search. (The very lowest redshift clusters are observed as part of the pre-scheduled cadence in ACS 814W to match the restframe B band.) Figure YYY shows this strategy. Note that this is a dramatically more efficient approach than the previous ACS/GOODS searches, performed by both our team and the Riess et al team, for several reasons. First, there is a large net savings in the number of search-plus-follow orbits necessary to study $\sim$10 SNe Ia; the previous method required $\sim$35\% more orbits. Second, the number of required ToO's is reduced from 10 to 3, in itself equivalent to many orbits of observing time, and that these are ToO's with more than two weeks advance notice and hence much less disruptive to schedule. Third, the resulting data has much smaller statistical uncertainties, since the large contribution from extinction correction is removed, making each of these SNe worth $\sim$10 SNe not found in ellipticals. Fourth, not only will the $\sim$10 cluster SNe be supplemented by $\sim$10 SNe from the field that would have been studied in the GOODS searches, but there is still an additional population of useful SNe that have their lightcurves obtained, which previously could not be followed given the limited number of ToO's. (Including the cluster and the field, each cluster target field will thus be expected to yield on average one SN Ia in the z>1 range and more at z<1 -- all with z-band lightcurves.) This observing strategy is also considerably simpler to schedule than the previous GOODS searches. The observing simulations show that there is flexibility in the exact choice of repetition dates; the cluster observations can be at any orient (unlike the contiguous GOODS tiles); there are clusters on the target list observable at different times of the year, and they do not all need to be observed in the same cadence like the GOODS tiles, so the observing load is spread. The observing requirements for this program are based on our experience with ACS and NICMOS measurements of supernovae in precisely this redshift range (with even a few in ellipticals), so we have direct tests of our exposure time, filter, and cadence requirements. For the bulk of the target clusters, between $z = 1.1$ and 1.4, we require a single ACS orbit with the 850LP filter every 24 observer days (i.e., $\sim$10.5 SN-restframe days) over the lightcurve. After propagating errors for the fit of the lightcurve timescale (used to calibrate the peak magnitude), this yields a SN Hubble-diagram point with uncertainty below the $\sim$0.15 mag intrinsic SN Ia uncertainty (in fact, 90 \% of the SNe will have distance modulus uncertainties below 0.10 mag). At redshift smaller than 1.1 one can afford to increase the cadence to 28 observer days and still meet the criterion that 95 \% of the SNe observed will have an uncertainty in the distance modulus of less than 0.15 mag. Correspondingly, for redshifts $z>1.4$ we use choose a shorter cadence of 20 days to compensate for the increase of the photometric errors. The fluxes/magnitudes, cadences and error bars shown in Figure YYY demonstrate the range that are expected for this program. [CHECK ALL THE NUMBERS IN THIS PARAGRAPH -- they are currently estimates.] For an elliptical host galaxy, with its symmetric smooth morphology, the images before and/or after the SN lightcurve generally provide a fit of the host galaxy light that gets subtracted from each image of the SN-plus-galaxy. However, for $\sim$2 of the targeted $\sim$10 SNe Ia we expect the SN to be so close to the core of a bright host that we will need an additional final image the following year, with a 3-orbit depth so as not to degrade the signal-to-noise after subtraction. More than offsetting this requirement for 6 additional orbits is the probability of cancelling the final cadence observations for clusters that have not produced a SN in time to be followed for sufficient lightcurve points. This will happen, on average, for 15 of the 25 clusters, saving 15 orbits from the total required. (The cancellations would be more than three weeks in advance, so will not introduce scheduling complications.) The total number of orbits thus required for this program is: 12 clusters x 7 planned 1-orbit ACS visits + 13 clusters x 8 planned 1-orbit ACS visits + 3 SNe x 5 NICMOS orbits (primarily TOO) + 2 SNe x galaxy-only 2-orbit ACS visit the following year -15 final cancelled orbits = 194 orbits. [To fill in: How do we use the info from the higher-redshift clusters --IRAC and x-ray selected -- to select which supernovae to follow-up with NICMOS?] [Issues to decide: Do we want full lightcurve follow-up of z>1.3 SNe, or just at max? Do we want an additional point in a NICMOS color at max for a number of z ~ 1.1 SNe to check the intrinsic color (which would help the measurements of spiral-hosted SNe)? What about a point in the CMAGIC period of the lightcurve?] The fundamental dataset of F850LP-band lightcurves for every SN shown (and more!) in Figure YYY will be obtained without need of any ToO's. However, for specific cluster redshift ranges the science will be greatly enhanced by strategic use of additional low-impact ToO orbits. Primarily these consist of a simple single-orbit or two-orbit observation in an additional NICMOS F100W filter {\em that can be scheduled over woo weeks in advance}. For 6 of the SNe ($z \approx$ 1 -- 1.25 clusters) these observations provide a color measurement at maximum light to calibrate the intrinsic SN Ia colors at high-redshift. For two of the SNe ($z \approx$ 1.25 -- 1.4 clusters) two maximum-light orbits will provide the restframe B-band point to tie these SNe to the same restframe-B Hubble diagram as the others. (Restframe U in 850LP will provide the lightcurve date and timescale ``stretch" measurement, as for previous very-high-redshift SNe.) Note that these ToO's are much less expensive in orbits and disruption to the HST calendar than the fast-run-around $\sim$20-orbit-timeseries ToO's needed for every one of the GOODS supernova discoveries by the two SN research teams. For only one [[[***two???***]]] of our ToO's, for a SN at$z \gs 1.5$ will this time series of 15 orbits be needed, to provide the restframe B lightcurve, when the F850LP lightcurve is too far into the restframe UV of the SN to provide significant signal-to-noise. [[[NOTE: This would be 6 x 1 orbit + 2 x 2 orbits + 1? x 15 orbits = 25 total ToO orbits.]]] \noindent {\bf Observing requirements for the cluster science.} The primary requirement for achieving our objectives on the evolution and formation of galaxies in clusters is to obtain reliable visual classifications of galaxies down to 2 magnitudes below $L^\ast$ at $z \sim 1.2$ using the standard system of elliptical, lenticular, and spiral/irregular over an area reaching out to $r_{200}$. The secondary requirement is to measure galaxy sizes, in particular $r_e$, with sufficient accuracy to enable estimates of $M/L$ ratios for early-type galaxies when coupled with velocity dispersions being obtained with ground-based facilities. These requirements can be met with [CHECK THIS NUMBER -- WHAT IF WE OBSERVE ONLY 7 ORBITS SOMETIMES?] 8 orbit exposures in the F850LP band of the central $\sim$200 arcsec area of our target clusters, as shown by Postman et al.\ (2005) and by Holden et al.\ (2005). \noindent {\bf Observing requirements for the weak lensing studies.} Most current weak lensing studies of galaxy clusters with HST use a mosaic of observations in order to extend the measurement of the lensing signal out to larger radii, thus improving the accuracy of the mass measurement. Typically the motivation for this is the need to break the mass-sheet degeneracy; in practice this is not feasible, and instead we intend to fit parameterized mass models (e.g., the NFW profile predicted by cold dark matter simulations) to the data. This approach still benefits from covering a large area, but our detailed calculations show that the gain is minimal for clusters beyond $z=1$ (where the FOV of ACS is well matched to the cluster). We find that we obtain more accurate masses by obtaining deeper observations instead, thus increasing the number density background galaxies and the accuracy in the shape measurements -- taking advantage of the unique capabilities of HST. Hence the observing strategy for the SNe search is well matched to the requirements of the weak lensing mass determination. For clusters below $z=1$ and clusters that show evidence of a complicated mass distribution we might consider mosaiced observations (with a large overlap in the centre), but most of the clusters will be observed to the full $\sim 8$ orbit depth. Figure XXX shows the expected relative uncertainty in the mass of a massive cluster with a velocity dispersion of $\sigma=1000$ km/s. *** want to include something on the relative accuracy we get for the zeropoint of the mass-observable relation, which is the ultimate goal of this***. These results are based on a photometric redshift catalog of the UDF, under the assumption that the cluster mass profile is well described by a singular isothermal sphere model. As we proposing observations of high redshift clusters, which are known to have a complex mass distribution in their cores (e.g., Hoekstra et al. 2000; Jee et al 2005), we also assume that we can only use the shape measurements at $r>60"$. We compared the outcome of this model calculation to actual published results (Jee et al. 2005; Lombardi et al. 2005) and find very good agreement. *** some of this can be moved to a figure caption if we include a figure*** Although the ACS observations provide a tremendous improvement over ground based observations (absolutely crucial for studies of high redshift clusters), they are not free from systematics. Large programs such as COSMOS and GEMS (Heymans et al 2004) show the PSF undergoing cyclic variation with a period of 15-20 days. These variations are caused by thermal fluctuations in the telescope and result in temporal variations in the PSF, which need to be corrected for. However, our team includes members with extensive expertise in this area. Using the Program TinyTim (Krist \& Hook 2004) we can create models of the PSF for the entire range of allowed focus values. Then, the $\approx 20$ stars in each exposure suitable for PSF measurement can be used to determine the focus value for that exposure. The appropriate TinyTim model can then be used for PSF correction. **** might rephrase or cut the following *** We should also note that because of the nature of this program, the clusters will be observed with different roll angles, thus partly averaging out some of the PSF effects. In addition, studies of galaxy clusters are less sensitive to residual systematics compared to cosmic shear measurements. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 3. Two-Gyro Mode Observations % (see Section 9.3 of the Call for Proposals) % % This new section has been added to take account of the possible % transition from standard three-gyro operations with HST to the % Two-Gyro Science Mode currently under development by the % HST project. This is likely to result in reduced scheduling flexibility % (fewer orbits available for a given target), lower observational % efficiency (longer overheads per target) and, possibly, degraded % scientific performance. Those issues are discussed % thoroughly in The HST Two-Gyro Science Handbook. % % Based on the information in the HST Two-Gyro Handbook, % proposers should address the following issues: % % ´ Are the scientific goals of this proposal attainable with observations % undertaken in two-gyro mode? % ´ How will the observations be modified to take account of two-gyro % operations? Modifications might include increased integration times % and/or fewer instrumental configurations for GO and SNAP Programs; % GO Programs might envisage a reduced number of targets % and/or increased orbit allocations. % ´ In the case of Large, Treasury and Long-Term GO Programs, what % are the scientific implications if only a subset of the data are obtained % in normal three-gyro mode? % \describetwogyro % Do not delete this command. % Enter your Strategy for Two-Gyro Mode Observations here. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 4. SPECIAL REQUIREMENTS % (see Section 9.4 of the Call for Proposals) % % List and justify any special scheduling requirements, including % requests for: % % - Target of Opportunity (TOO) observations: % estimate the TOO's probability of occurrence during Cycleæ14, and % state how soon HST must begin observing after the occurrence % - Continuous Viewing Zone (CVZ) observations % - Shadow time (SHD) or Low-sky (LOW) observations % - Time-critical observations % - Early acquisition observations % - Coordinated Parallel (CPAR) or Pure Parallel (PPAR) observations % - Target acquisitions that use the `Reuse target offset' function % - Real-time interactions % - Scheduling of STIS/MAMA and STIS/CCD observations % (other than target acquisitions) in the same visit % - Scheduling of coronographic observations in the same % orbit with a roll of the spacecraft between observations % - Requests for expedited data access % - Other special scheduling requirements (e.g., requests for % non-SAA impacted orbits % % Also, if applicable, discuss the benefits of or need for a % non-default proprietary period request. % \specialreq % Do not delete this command. % Justify your special requirements here, if any. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 5. COORDINATED OBSERVATIONS % (see Section 9.5 of the Call for Proposals) % % If you have plans for conducting coordinated observations with % other facilities that affect the HST scheduling, please describe % them here (examples are coordinated or simultaneous observations % with other spacecraft or ground-based observatories). Describe how % those observations will affect the scheduling. % % If you have plans for supporting observations that do not affect % HST scheduling, then don't describe them here. If they improve % your science case, then describe them in the `Scientific Justification' % section of the proposal (see Section 9.1). % % A limited feasibility check on coordinated observations between % HST and a few other observations (Chandra, XTE, and FUSE) may % be performed via the Visual Observation Layout Tool (VOLT), developed % at the Goddard Space Flight Center. See the VOLT web site at % http://pioneer.gsfc.nasa.gov/public/volt/home.html % for more information. Note that VOLT cannot be used for detailed % (Phase II) scheduling of HST observations. % % Joint HST-Chandra Observations % % Proposers requesting joint HST-Chandra observations must provide a % full and comprehensive technical justification for the Chandra portion % of their program. This justification must include: % % * the choice of instrument (and grating, if used), % * the requested exposure time, justification for the exposure time, target % count rate(s) and assumptions made in its determination, % * information on whether the observations are time-critical; indicate % whether the observations must be coordinated in a way that affects the % scheduling (of either the Chandra or HST observations), % * the exposure mode and chip selection (ACIS) or instrument configuration % (HRC), % * information about nearby bright sources that may lie in the field of view, % * a demonstration that telemetry limits will not be violated, % * a description of how pile-up effects will be minimized (ACIS only). % % Technical documentation about Chandra is available from the % Chandra X-ray Center (CXC) Web Page at: http://asc.harvard.edu/ % which also provides access to the Chandra Help Desk. The primary document % is the Proposer's Observatory Guide, available from the Chandra User % Documents Web Page and the Chandra Proposer Web Page. Full specification % of approved observations will be requested during the Chandra Cycle 7 % period when detailed feasibility checks will be made. % % Proposers requesting joint HST-Chandra observations must specify % whether they were awarded Chandra time in a previous Chandra or HST % cycle for similar or related observations. % % Joint HST-Spitzer Observations % % Proposers requesting joint HST-Spitzer observations must provide a full % and comprehensive technical justification for the Spitzer portion of their % program. This justification must include: % % ´ the choice of instrument and Astronomical Observation Template(s), % ´ the requested observing time, justification for the requested time, target % fluxes, required sensitivity and assumptions made in its derivation, % ´ information on whether the observations are time-critical; indicate % whether the observations must be coordinated in a way that affects % scheduling of either HST or Spitzer observations. % % Technical documentation about the Spitzer Space Telescope is available % from the Spitzer Science Center (SSC) Web Page, which also provides % access to the Spitzer Help desk (help@spitzer.caltech.edu). The primary % document is the Spitzer ObserverÍs Manual, available, together with other % relevant documents, from the Proposal Kit Web Page. The SSC strongly % recommends that observers proposing Spitzer observations estimate the % required observing time using Spot, the Spitzer proposal planning software, % also available from the online proposal kit. % % Proposers for Collaborative HST-Spitzer Programs must also complete the % Spitzer proposal submission process by the HST Cycle 14 deadline. Both % submissions should include the same PDF attachment. Detailed % instructions for the Spitzer submission are in the Spitzer Cycle 2 Call for % Proposals. In the case of Regular HST-Spitzer Programs, full specification % of the Spitzer observations will be requested by Spitzer after the proposals % are selected, at which time detailed feasibility checks will be made. % % Proposers requesting joint HST-Spitzer observations must specify whether % they were awarded Spitzer time in a previous cycle for similar or related % observations. % % Joint HST-NOAO Observations % % Proposers requesting joint HST-NOAO observations must provide a full % and comprehensive technical justification for the NOAO portion of their % program. This justification must include: % % * the telescope(s) and instrument(s) on which time is requested, % * the requested observing time per telescope/instrument, a specification % of the number of nights for each semester during which time will be % required, a breakdown into dark, grey and bright time, and an % explanation of how the required exposure time was estimated, % * information on whether the observations are time-critical; indicate % whether the observations must be coordinated in a way that affects % the scheduling (of either the NOAO or the HST observations), % * a description of any special scheduling or implementation requirements, % % Successful proposers will be asked to supply additional details about % the observations, i.e., the same details required for NOAO proposals for % the particular telescope/instrument. This "Phase II - NOAO" information % must be submitted by the April 30, 2005 NOAO deadline for the Fall 2005 % semester. Submission instructions will be forthcoming following notification % of the results of the HST review. % % Technical documentation about the NOAO facilities is available from the % NOAO Web Page at: http://www.noao.edu/. Questions may be directed to % the NOAO Proposal Help Desk by email to noaoprop-help@noao.edu. % NOAO will perform feasibility checks on any approved proposals. % % Proposers requesting joint HST-NOAO observations must specify whether % they were recently (in the last two years) awarded NOAO time for similar or % related observations. % % A full and comprehensive scientific justification for the requested NOAO % observing time and facilities must be given in the `Scientific Justification' % section of the proposal. % \coordinatedobs % Do not delete this command. % Enter your coordinated observing plans here, if any. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 6. JUSTIFY DUPLICATIONS % (see Section 9.6 of the Call for Proposals) % % Justify, on a target-by-target basis, any potential duplication with % previously accepted GO or GTO observing programs. Use the % `Duplication' checkbox in the OS to identify the duplicating observations. % \duplications % Do not delete this command. % Enter your duplication justifications here, if any. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % 7. PREVIOUS RELATED HST PROGRAMS % (see Section 9.9 of the Call for Proposals) % % Please list the program number and status of the data for all % accepted GO/SNAP/AR programs of the PI which are relevant to % the current proposal. Include a list of major publications resulting % from the previous HST observations and provide a sentence or two % describing the significance of each paper. Related publications based % on theory and/or data from other telescopes should be included. % Unpublished data from early cycles should be explained. A significant % publication record will be regarded by the Review Panels and TAC as % a strong plus. GTO programs and publications may be included at the % PI's discretion. % % The information in this section is not counted against % the total page limits for the proposal. % \previousprograms % Do not delete this command. % Enter your previous HST programs here, if any. \emph{To be updated} By combining observations from a series of GO programs over a number of HST cycles we have obtained a cumulative sample of high redshift SNe which has yielded new determinations of cosmological parameters $(\Omega_M, \Omega_\Lambda, w)$. Equally important, these HST observations have been the basis for studies of possible systematics of the SN technique, such as host-galaxy extinction or evolution. Two such multi-cycle HST studies were published in the past year and both provided confirmation and improved precision on the earlier ground-based acclerating universe results. Knop \etal, 2003 (based on G0-7336, GO-7590, GO-8346) presented an analysis of an independent set of 11 high redshift SNe. The high-quality lightcurves available from photometry on WFPC2 make it possible for this sample alone to provide measurements of the cosmological parameters comparable in statistical weight to the previous results. In addition to high-precision lightcurve measurements, this data offered greatly improved color measurements of the high-redshift supernovae, and hence improved host-galaxy extinction estimates. These extinction measurements show no anomalous negative $E(B-V)$ at high redshift. The precision of the measurements is such that it was possible to perform (for the first) time a host-galaxy extinction correction directly for individual supernovae without any assumptions or priors on the parent $E(B-V)$ distribution. Sullivan \etal\ 2003 (based on GO-8313, GO-9131) presented the Hubble diagram of distant type Ia supernovae (SNe Ia) segregated according to the type of host galaxy. This allowed us to confirm our previous evidence for a cosmological constant by explicitly comparing SNe residing in galaxies likely to contain negligible dust with the larger sample. % These data demonstrate that host galaxy extinction is unlikely to systematically dim % distant SNe Ia in a manner % that would produce a spurious cosmological constant. These data provide a key test of evolutionary systematics. Other such multi-cycle analyses, described below, are in progess. In particular, this year we are completing final observations of host galaxies after the SNe faded for SNe discovered in GO-9075 and GO-8585. \noindent {\bf GO-9727}: This cycle 12 program will begin observations in April 2004 using ACS to do a new search for very high redshift $(1.2 > z > 1.6)$ SNe Ia in the GOODS-N field. In coordination with Riess (GO-9728), images from 15 ACS pointings will be taken approximately every 45 days and searched for candidates. Followup photometry will be obtained with ACS and NICMOS for approximately three very high redshift SNe. \noindent {\bf GO-9075}: In this program, we pushed our SNe Ia studies to the highest redshifts that are feasible for a ground-based discovery and spectroscopic identification campaign. HST follow-up observations for this program started after servicing mission 3B in March 2002 and have been completed for the most part --- final reference images are still to be taken. Coordinated with three large search campaigns using the Subaru 8.2~m and also with simultaneous smaller searches using the CTIO 4~m and CFHT 3.6~m, we obtained ACS/WFC and NICMOS/NIC2 photometry for multi-epoch lightcurves of eight Type Ia SNe at high redshift $(0.9 < z < 1.3)$. For two of the highest redshift SNe, ACS grism spectra were taken. Analysis of this ACS data is in progress. With the refurbished NICMOS, we obtained final reference images of the host of SN1998eq, which we had previously studied in G0-8088, and these images will allow us to complete that analysis. {\bf GO-8585}: In GO 8585 we observed six Type Ia supernovae with HST using WFPC. The supernovae were discovered in ground based searches at the CTIO 4-m, CFHT and Subaru telescopes. We obtained both U- and B-band restframe photometry (using either F814W or F850LP depending on the redshift) for each supernova for a period of 2 months. %Analysis of this data will be completed when the final reference images are % available, % scheduled for spring 2003. Analysis of this data is presented in the PhD thesis of J.~Raux (Univ. of Paris, 2003), presented at the January 2004 AAS meeting. A publication is in progress. {\bf GO-8313}: The objective of this project, which has now been completed with the publication (Sul03) mentioned above, was to obtain snapshot unfiltered STIS images of distant galaxies of known redshift which have hosted supernovae (SNe) of Type Ia found by the SCP, 20 of which are used in the Hubble diagram of 42 type Ia SNe (Perlmutter \etal\ 1999). % In Sullivan et al. (2003, Sul03) we present these new results % on the Hubble diagram of SNe~Ia as a function of host galaxy % morphology that demonstrates that host galaxy extinction is unlikely % to systematically dim distant SN~Ia in a manner that would produce a %spurious cosmological constant. The internal extinction implied is small, even for late-type systems ($A_{B} < 0.3)$, and the cosmological parameters derived from those SNe~Ia hosted by (presumed) dust-free early-type galaxies are consistent with our previous determination of a non-zero $\Lambda$. The brightness scatter about the Hubble line for SNe~Ia in these early-type hosts is also significantly smaller than for the SNe~Ia in late-type galaxies. This result was based on HST STIS ``snapshot'' images and Keck spectroscopy of SNe spanning the range $0.3 < z < 0.8$. {\bf GO-8346}: We had the unique opportunity of following up SN2000fr, which had been discovered {\it 14 days prior} to maximum light in its restframe. Because this supernova at z=0.54 was discovered so early we were able to obtain excellent light curves from HST in F555W, F675W and F814W spanning the period from one week prior to maximum light to 6 weeks after Several spectra of the supernova were taken at VLT and Keck along with NIR photometry at VLT. % To date, this is still the best observed % high-redshift supernova and preliminary results were presented in % Nobili, S. \etal\, 2001, AAS, 199,1611N. {\bf DD-8088}: WFPC2 and NICMOS (cycle 7) observations were obtained for SN1998eq at $z=1.20$ %(a record-breaking redshift for a %spectroscopically confirmed Type~Ia supernova; (Aldering, \etal, 1998,IAUC,7046). The preliminary photometry is consistent with the previous results for $\Omega_M,\Omega_\Lambda$. With the final NICMOS image of the galaxy without the supernova obtained, this analysis can now be completed. {\bf GO-7850} and balance of {\bf GO-7336} and {\bf DD-7590}: WFPC2 and NICMOS observations were obtained for 11 Type Ia supernovae in the redshift range 0.36---0.86. These observations, including final references where necessary, are now complete, and the results were published in Knop, R., \etal\, 2003 as mentioned above. % The cosmological results from these SNe % are in close agreement with results from the first supernova % results (Per99) that gave direct evidence for a % cosmological constant. % The lightcurves provided by WFPC2 for these supernovae % were excellent; at the higher redshifts, these lightcurves provide a % % % substantially % better measurement of the calibrated supernova magnitude than those for % comparable supernovae observed only from the ground. The color information provided by NICMOS (Burns, S., \etal, 2001,AAS,199.1610B), was only possible with HST. % The improvement of % the confidence limits on the cosmological parameters $\Omega_M$ and % $\Omega_\Lambda$ are as good as we had previously predicted. {\bf GO-7336} and {\bf DD-7590}: Perlmutter \etal, 1998, Nature, 391, 51 reported the results of our HST and ground-based imaging and Keck spectroscopic observations of SN1997ap. %then the {\it highest redshift} ($z = 0.83$) {\it %spectroscopically confirmed} Type~Ia supernova. The HST portion is based on a total of 4 orbits. Also from this program, HST observations of two $z = 0.83$ SNe~Ia are included in the analysis in Per99 which reports on the results from our HST and ground-based imaging and Keck spectroscopic observations of 42 type~Ia supernovae with $0.18 < z < 0.86$. The paper rules out a flat $\Omega_M = 1$ universe and presents very strong evidence for a positive cosmological constant. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% \end{document} % End of proposal. Do not delete this line. % Everything after this command is ignored.