Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/~webdocs/STScINewsletter/2002/spring_02.pdf Äàòà èçìåíåíèÿ: Mon Jan 19 23:37:30 2009 Äàòà èíäåêñèðîâàíèÿ: Mon Apr 6 00:27:09 2009 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: mercury program

VOL



19



ISSUE



02



S T S I

Spa ce T elescope Science Institute

Three Hubble Treasury Programs Selected in Cycle 11
Ages from Near-UV Spectra of Globulars and Galaxies
Ruth C. Peterson, PI, peterson@ucolick.org, and B.W. Carney, B. Dorman, E.M. Green, W. Landsman, J. Liebert, R.W. O'Connell, and R.T. Rood, Co-Is

The Hubble Treasury The AC S Contribution Program on Eta Carinae to G O O D S
Kris Davidson, PI, kd@astro.umn.edu, Theodore R. Gull, Kazunori Ishibashi, John Hillier, Roberta Humphreys, and the Eta Car Team1 Mauro Giavalisco, PI, mauro@stsci.edu, Mark Dickinson, and the GOODS Team

T

SPRING

he ultimate goal of Hubble Treasury program GO-9455 is characterizing the age and metallicity of spatially unresolved stellar systems older than a few billion years. Heroic efforts using optical integrated-light spectra have aimed to establish the age of globular clusters and galaxies from the temperature of main- sequence turnoff stars and their metallicity from the strengths of spectral absorption lines. Our program will provide the tools to do this in the near ultraviolet (UV), where the contribution from red giants is nearly negligible, but where line absorption must be Continued painstakingly modeled. page 3 Our near-UV STIS (Space



2002

O

ur Hubble Treasury project to observe Eta Carinae has behind it a web of motives and applications. Though primarily focused on a mysterious spectroscopic event expected to occur in mid-2003, the data will become resources for several branches of stellar and nebular astrophysics. Also, the emission line spectrum will provide reference measures of instrumental characteristics pertinent to other archived data from the Space Telescope Imaging Spectrograph (STIS). In several respects, ours is the most intensive spectroscopic project yet attempted with Hubble. We are `pushing the envelope' for both spatial and spectral resolution on a complex target. The resulting data Continued will be widely useful-- page 7 and impressive in volume.

O

ur Hubble Treasury program will take deep images of two fields at high north and south galactic latitudes using the newly installed Advanced Camera for Surveys (ACS). These data sets are part of the Great Observatories Origins Deep Survey (GOODS), which is a multi-wavelength campaign of imaging and spectroscopy on these fields using three NASA Great Observatories (Hubble, SIRTF, and Chandra), the European Space Agency 's XMM-Newton, and large ground-based telescopes. The goal of GOODS is to create a community resource for exploring the distant universe and studying the evolution of galaxies-- a grand, unsolved problem of astronomy that is profoundly connected to cosmology. Using these Continued ACS images, astronomers page 8



A rejuvinated Hubble after its release from space shuttle Columbia on flight day nine of Servicing Mission 3b. ESC Number S109E5706.


DIRECTOR'S PERSPECTIVE

Small Differences
Steven Beckwith, svwb@stsci.edu

ritz Strobl of Austria won the men's downhill event at the 2002 Winter Olympics in 1:39.13. Kjetil Andre Aamodt of Norway posted 1:39.78 to come in fourth. His performance lagged Strobl's by just 0.7%, but that was enough to knock him out of the medals. Fifteenth place went to Franco Cavegn (Switzerland) with a time of 1:40.81, 1.7% off the best mark, and 1.4% away from a medal. At 85 miles per hour, the 0.65 seconds separating gold and fourth place corresponds to 25 meters, an unmistakable margin if the races were run side-byside, yet almost imperceptible as a percentage of performance. A small burr on a ski's edge, a slight flare on a turn, too much airtime over a bump, and a racer could plummet from first to eighteenth. Only 1/5 of the skiers between Cavegn and Strobl would medal, about the same as the fraction of Hubble proposals that are granted telescope time. Electronic timing eliminates controversy from the downhill race. Not so for women's figure skating. Sarah Hughes won the gold with a stunning free skate that brought her from fourth to first, dropped the leader, Michelle Kwan, to third, and knocked Sasha Cohen out of the medals. Nine judges watched the athletes go through their routines and graded them on their performance. The winners were decided by a rating system that is confusing to the untrained eye. A slight `double foot ' when landing a triple Axel, insufficient stretch during a spin, or a poor extension on a wide spiral merits a deduction from the perfect mark. The final score in figure skating is decided in two parts: a mark for technical merit, the product of the degree of difficulty of the routine and the perfection of its performance, and a mark for artistic presentation, a measure of how much the judges liked the combination of jumps, spins, and footwork chosen by the skater. There is considerable latitude in assigning this second mark: witness the controversy over the pairs competition between the Russian and Canadian couples. As with downhill skiing, the separation between the winner and fourth place is tiny. The rating of proposals for Hubble time closely parallels the judging of figure skating. Proposals are rated first on technical merit and then on the scientific equivalent of artistic performance. A stumble on technical merit, like a fall in figure skating, is usually enough to knock the proposal out of the competition for a time award. Most proposals pass muster on technical merit, and the medalists are selected on the scientific equivalent of artistic performance: is the question really interesting? Is the technique clever? Most importantly, do the proposers demonstrate a clear understanding of current norms in scientific thinking, showing they will advance knowledge in predictable ways that their peers can understand? This latter demonstration is a lot like choosing music and choreography to suit the judges. What works for one Telescope Allocation Committee (TAC) may fail for another. Just as in Olympic events, the separation between the best proposal and the cutoff is exceedingly small. Almost all members of the TAC and the subject panels have expressed frustration that they differentiate between proposals on unquantifiable subtleties. We cannot make sound judgments about the likelihood of great science near the margin between successful and unsuccessful proposals. Nevertheless, just as there are only 3 medals for which the top 15 ice skaters compete, there is only enough Hubble time to award to about 1 of every 5 proposals. The TAC must find reasons to make those judgments, even though the reasons are small differences in perception. Just as there are protests about Olympic judging, there are protests about the decisions made by the TAC. After every cycle, I receive correspondence from unlucky proposers who insist that their proposals were rated unfairly. Usually, poorly worded comments in the notification letters stimulate the protests--comments suggesting the TAC misunderstood the technical merit or did not value what is truly chic in science. Indeed, the TAC often fails to `get ' the proposal, usually because the message is unclear or unpersuasive. We assume that the burden is on the proposer to write cogently for nonspecialists. Not all proposers feel that way. My job is to explain why the decision will not be reversed and why an appeal is almost always futile: because we cannot run the TAC process twice.



2




DIRECTOR'S PERSPECTIVE
In truth, the problem is one of small differences. We get about three times as many interesting, technically meritorious proposals as we can approve. There is nothing wrong with the 2/3 of these proposals that do not get time. A brilliant proposal may get time in one cycle and be turned down in the next only because the panel uses slightly different criteria for `artistic presentation.' As with Olympic judging, there is some consistency from year to year, with the known favorites being looked upon favorably and some corporate memory of long term programs. There is, nevertheless, a large element of chance; past performance is not a good indicator of future promise, a concern voiced bitterly by a number of luminaries this cycle, who believed the very fact that they did not get time indicated a problem with the process. Michelle Kwan and Picabo Street probably feel the same way. One of the harder parts of my job is to reply to people who cannot understand why their proposals are turned down. I take no pleasure in these replies. Luck is a factor in any selection process with such a high rejection ratio. I generally sympathize with the protest, but there is rarely a practical way to change the outcome. There is one obvious solution to my dilemma. We need three Hubble Space Telescopes. Three Hubbles would produce three times as many great discoveries as the current one does. The ideas are there. Our ability to predict which ideas are best is modest or worse. If chance is at work, the increase in science productivity will be proportional to the resources we devote up to the limit of meritorious ideas. So, the next time you suffer at the hands of your peers and feel inclined to write a letter of protest to me, put your energy to more productive use by writing your government representatives. Tell them that we need three Hubble Space Telescopes. Please don't tell me your proposal really deserved time. I already know it was deserving. It was edged out of the medal competition by small differences.

Globulars & Galaxies from page 1

Telescope Imaging Spectrograph) spectral observations will substantially extend the temperature and abundance range of stars and globular clusters that currently serve as age/metallicity templates. The observations, outlined in the tables (next page), will be publicly available immediately. Our theoretical grids will foster quantitative comparisons. This endeavor will constrain cosmology as well as stellar and galaxy formation. For example, the Extremely Red Objects discovered by Hubble and large ground-based telescopes have redshifts z=1 or higher, so their near-UV spectra are redshifted into the optical. The ages now being estimated for them (e.g., Spinrad et al. 1997, ApJ, 484, 581) imply that stellar populations of near- solar metallicity were first generated when the universe was quite young. Our work will also benefit stellar research. Because hot horizontal branch stars and blue stragglers influence near-UV spectra, we will empirically assess their numbers from a ground-based membership survey of two old open clusters and judge from radial-velocity variations what role close binaries play in their origin and evolution. We will directly evaluate the effect of enhanced light-element abundances on single-star evolution by calculating new theoretical isochrones with and without such enhancements. Our high-quality, near-UV spectra of nearby stars, such as Procyon and the Hyades, will allow refined characterizations of these fundamental standards. We will redetermine the parameters of each standard star from optical echelle spectra and judge chromospheric effects from near- and far-UV emission. We will then calculate each star 's near-UV spectrum and compare it with observation. The results will be documented in an electronic and a paper atlas identifying many absorption lines. Most of our program effort is dedicated to the creation of a comprehensive grid of stellar and composite spectra calculated from first principles. A researcher may then extract spectra whose predetermined parameters densely cover a wide range. Matching these theoretical templates to observations will provide quantitative determinations of age/metallicity and of the systematic uncertainties due to reddening and other parameter choices. The first of the three steps of our program is to revise the list of near-UV absorption-line parameters used in the grid calculations. We will make empirical assignments of transition probabilities and approximate identifications for thousands of lines still missing from the calculations by comparing highquality observed spectra of stars with similar line strengths but different temperatures. Peterson, Dorman, and Rood (2001, ApJ, 559, 372) show how this process leads to the successful reproduction of near-UV spectra of stars of metallicity one-third solar ([Fe/H]=-0.5). However, at higher metallicities the match deteriorates as ever-weaker Continued unidentified transitions intrude. To remedy this, in Cycles 11 and 12 we will page 4



3




Globulars & Galaxies from page 3

obtain echelle spectra with signal-to-noise ratio near 40 of several (mostly) strong-lined standards. These standards are listed in Table 1. Other Cycle 12 targets, listed in Table 2, will extend the spectral types of near-UV templates. They include four field blue horizontal-branch stars and the `super-metalrich' red giant µ Leo, and stars near and blueward of the turnoff in the old open clusters M67 and NGC 6791, the latter with a metallicity two- or three-times solar (Peterson & Green 1998, ApJ, 502, L39).

Table 1.
HD #
1461 26736 27406a 28033 61421 103095 140283 157466 184499
a

Stellar Targets for Mid-UV Line-List Improvements Cycle 11 Orbits: 29; Cycle 12 Orbits (first three stars): 16
B-V
0.69 0.61 0.53 0.51 0.40 0.75 0.49 0.47 0.58

V
6.46 8.08 7.46 7.35 0.34 6.43 7.22 6.89 6.62

Type
G0V G5 GOV F8V F51V-V G8Vp sdF3 F8V GOV

T

eff

E14OM E230H E230H E230H E230H [Fe/H] 1425å 2263å 2513å 2762å 3012å
+0.43 +0.15 +0.15 +0.15 +0.00 ­1.45 ­2.60 ­0.44 ­0.60 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

5930 5720 6000 6460 6700 5000 5750 5935 5700

Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes

This star will be replaced, because it is indistinguishable in the optical from HD 28033.

Table 2.
Star ID

Stellar Targets for Empirical Mid-UV Templates Cycle 12 Orbits: 22
V
10.29 8.72 12.11 12.34 13.24 3.88 6.99 17.43 16.53

B-V
­0.17 0.04 0.46 0.42 ­0.29 1.22 0.14 0.89 0.80

Type
sdB BHB strglr strglr sdB K2111 BHB TO strglr

Teff
27000 8900 6750 6500 29000 4500 7550 ~5900 ~6250

[Fe/H] Grating cen
~0 ­1.42 0.0 0.0 ~0 +0.3 ­1.55 +0.4 +0.4 E230H 2513å , 2762å , 3012å E230M 2707å G230L 2376å G230L 2376å E230M 2707å E230H 3012å E230M 2707å G230L 2376å To be observed with KR 8222

HD 4539 HD 74721 M67 5997b M67 S 1195b PG 0941+280 HD 85503 HD 161817 NGC6791 KR8222b NGC6791 KR8343b
b

Subject to confirmation of cluster membership.

The second step is to use the revised line list to calculate the grid. It will cover metallicities from one-tenth to four-times solar, will embrace various levels of light-element abundance enhancements, and will include red giants, turnoff and main- sequence stars, blue stragglers, and horizontal-branch stars. From these calculated spectra, we will compute diagnostics such as the indices of Fanelli et al. (1990, ApJ, 364, 272) as a function of stellar temperature, gravity, and abundance. From new stellar isochrones we will derive weights to combine the individual spectra of stars across the HertzsprungRussell diagram, producing composite spectra representing the near-UV integrated light of old stellar populations. Stellar and composite grids, diagnostics, and the line list itself will all be made publicly available as they are completed.



4




The third step is to test the grid's application to integrated-light Cycle 13 near-UV spectra of several moderately to extremely metal-rich globular clusters in the nearby Andromeda galaxy, listed in Table 3. Because these globulars can be spatially resolved into stars, at least in part, the validity of the deduced age(s) and stellar mix can be checked directly.

Table 3.
ID
GI G170 G174 G177 G280

Globular Cluster Targets in M31 Cycle 13 Orbits: 43, with G23OLB at 2375å
V
13.75 16.25 16.26 16.06 14.15

A1tID
MayII Bo109 Bo112 Bo115 Bo225

B-V
0.83 1.02 1.07 0.99 0.95

[Fe/H] RV
­1.08 ­0.31 +0.29 ­0.15 ­0.70 ­332 ­613 ­286 ­497 ­164

N

orb

Notes
Oblate in disk near center in disk [Fe/H] range

5 10 10 10 8

Figure 1 (next page) illustrates that the match between stellar calculations and the observations of old stellar systems may be substantially improved by using spectra from multiple models spanning a range of parameters, rather than that of a single star (as Spinrad et al. had to do). The black line in all four panels is the spectrum obtained by Hubble's Faint Object Spectrograph (FOS) for the Andromeda globular G1 (Ponder et al., 1998, AJ, 116, 2297). The colored lines are calculated spectra. In the calculations, [Fe/H]=-0.60 was assumed, and the magnesium abundance was raised above this by a factor of two in all but the case plotted in the next-to-bottom panel. The two bottom panels show calculations made for a single stellar model, that of a cool turnoff star. In the second-from-top panel, fluxes from two models were coadded--those of a hotter turnoff star Continued and a cooler star on the main sequence. The best match, seen in the top panel, page 6 adds a model of a cool blue horizontal-branch star.

Hubble Calibration Workshop October 17-18, 2002
Harry Ferguson, ferguson@stsci.edu

T

he list is long: charge-transfer efficiency, pedestal effects, point- spread functions, throughputs, scattered light, line- spread functions, cosmic-rays... These are instrumental effects and potential sources of systematic error. Whether we like it or not, tomorrow 's textbooks will hold astronomical knowledge that depends on today 's understanding of our imperfect observing equipment. Furthermore, pushing the forefronts of science often means pushing the instruments to their limits, where all kinds of calibration `gotchas' may be hiding. So, please come help set the course of astronomical history at the fourth Hubble Calibration Workshop! The workshop will feature reports from the commissioning of the Advanced Camera for Surveys (ACS) and the re-commissioning of the Near Infrared Camera and Multi-Object Spectrometer (NICMOS). Experts will present new calibrations and advances in the understanding of the Space Telescope Imaging Spectrograph (STIS), the Wide Field Planetary Camera 2 (WFPC2), the Fine Guidance Sensor (FGS), the Faint Object Spectrograph (FOS), and possibly the Goddard High Resolution Spectrograph (GHRS) and the Faint Object Camera (FOC). We intend the workshop to foster sharing of information and techniques between observers, instrument developers, and instrument support teams. The published proceedings will provide a valuable source of information for Hubble observers. There will be sessions of talks, posters, and time for demonstrations or splinter groups on various topics. The meeting will be held at the Space Telescope Science Institute immediately following the Astronomical Data Analysis Software and Systems conference being held in downtown Baltimore earlier the same week. Further information on the Hubble Calibration Workshop, including registration forms and instructions for presentations, will appear soon on the Institute's Hubble website, http://www.stsci.edu/hst/.



5




Globulars & Galaxies from page 5

2300

2400

2500

2600

2700

2800

2900

3000

3100

Figure 1. The plot compares four theoretical stellar spectra (in color) to the observed near-UV spectrum of the M31 globular cluster G1 (black), offset vertically for visibility. The complexity of the modeling increases from bottom to top, as the match improves. The bottom two calculations are based on one stellar model, the next on two models, and the top on three. The model parameters adopted are listed above and to the left of each spectrum, including the effective temperature (in K), the letter C indicating chromospheric heating, the log of the surface gravity, [Fe/H], the microturbulent velocity (in km/s), and the logarithm of the magnesium abundance enhancement shown immediately below. In the top two plots, the weights adopted for each model are listed; they are proportional to the numbers of stars represented by the model times the stellar radius squared.

The figure suggests three important consequences of using a range of stellar models. First, the effect of the magnesium-to-iron ratio can be explicitly evaluated. Second, the temperature of turnoff stars can be derived more accurately, because the cooler main- sequence (and subgiant and giant) stars can be included. Finally, the role of blue horizontal-branch stars can be assessed. All these determinations will be done from theoretical composite spectra encompassing a broader range of stellar parameters, to be coadded with weights assigned from isochrones. The results for blue horizontal-branch stars could be verified empirically from direct Hubble imaging in the blue. The results may impact the young ages inferred from strong Balmer lines in some Andromeda clusters (e.g. Burstein et al. 1984, ApJ, 287, 586). Peterson will shortly submit a paper to the Astrophysical Journal that describes a preliminary effort along these lines.



6




Eta Carinae from page 1

A combination of circumstances has made Eta Car one of Hubble's most scientifically productive targets. It is the most luminous evolved star that can be easily observed. It has a history of titanic outbursts and exhibits an extraordinarily dense wind. In addition to the star, Eta Car has a surrounding `Homunculus' nebula of ejected material, in which a STIS slit spectrogram finds at least five distinct types of emission-line spectra--three of them unlike any other known object. Because it is spatially complex, spectrally complex, and bright enough for us to obtain excellent data rapidly, the Homunculus has famously demonstrated Hubble's capabilities.2 Our Hubble Treasury observations with STIS will use about 30 grating-tilt combinations to cover the entire CCD wavelength range, from 170 to 1000 nm, several times. The Eta Car spectrum is so rich that we dare not omit any wavelength region! Many of the problems associated with Eta Car are unsolved at a surprisingly basic level, including unfamiliar aspects of stellar structure, stellar winds, diffuse gas dynamics, dust formation and destruction, and exotic nebular excitation processes. For this reason, each data set on Eta Car, imaging or spectroscopic, has application to a variety of topics. Recently, the discovery of a puzzling secular variation added another dimension to Eta Car 's roster of enigmatic characteristics. In 1996 Augusto Damineli recognized that certain brief spectroscopic episodes, occasionally reported in the past, recurred at 5.5-year intervals. He predicted that the next episode would occur near the end of 1997. Alerted X-ray astronomers monitored a tremulous rise in the hard X-ray flux from Eta Car, which crashed to near zero in mid-November of 1997. While groundbased spectroscopy showed complex changes in the following weeks, we were unable to obtain STIS observations until the first day of 1998. Thus began the project to explore the entire 5.5 year period, which has obtained STIS observations in each subsequent year. There is no good explanation for Eta Car 's 5.5-year cycle, although some have suggested that a companion star with that orbital period is the ultimate cause. Our spatially resolved observations of the next event, in mid-2003, should assist interpretations--and perhaps even discover the cause--by obtaining unique information on the physical state of Eta Car and its ejecta. We plan to make repeated STIS observations from April to August 2003. We will supplement these observations with WFPC2 and ACS imaging and additional STIS spectroscopy before and after the event. Our 2003 observations may be the last comprehensive spectroscopy of an Eta Car event for a very long time. The survival of Hubble's long- slit spectroscopic capability until 2008 is quite uncertain, and no other instrument with comparable high-resolution and UV capabilities is even on the horizon. Our core data product will be a series of processed STIS spectrograms covering the UV to nearinfrared wavelength range with a spatial extent of at least 10 arcsec. Each spectral image will consist of a roughly 30,000 x 500 pixel FITS file. In addition to the new observations, we plan to include the data obtained earlier in the 5.5-year cycle on a total of about 10 occasions. Each of these giant spectrograms will contain a number of useful components, including the stellar wind, several distinct types of very rich, non-routine emission line spectra, and the spectrum of the star reflected by dust in the Homunculus, which `sees' this non- spherical star from a range of directions. We expect that the radiatively excited emission lines will be useful for diverse research topics, such as UV spectra of AGN. Experienced STIS users will appreciate how challenging our data processing task will be. Largely motivated by earlier observations of Eta Car, the STIS instrument team has improved techniques, such as distortion correction, to enable high spatial resolution. They have improved the wavelength calibration using Eta Car 's ejecta, which exhibit more than 2,000 identified, narrow emission lines. Our data products may improve the analysis of STIS observations of other objects. In addition to our `core' data products, we plan to provide supplementary STIS observations, including CCD data in limited wavelength intervals, MAMA (Multi-Anode Microchannel Array) echelle data in the UV, and new WFPC2 and ACS images. Recognizing that future users of our data sets will want practical results with minimal fuss, we plan to make them as user-friendly as possible. We hope to provide convenient tools for viewing and extracting and an appropriate meta-database.
1 Jon Morse, Fred Hamann, Sveneric Johansson, Nolan Walborn, Nathan Smith, Kerstin Weis, Henrik Hartman, Mike Corcoran, Augusto Damineli, Otmar Stahl, and Manuel Bautista. 2 For information and references see Davidson, K., 2000, in "Cosmic Explosions," AIP Conf. 522, ed. by S. Holt & W. Zhang, p. 421, and many papers in "Eta Carinae and Other Mysterious Stars," ASP Conf. 242, ed. by T.R. Gull et al., 2001.

HST image of Eta Carinae. STScI-PRC96-23a - June 10, 1996 Jon Morse (U Colorado) and NASA



7




GOODS from page 1

will trace the emergence of the Hubble sequence and study the evolving relationships between the physical and morphological characteristics of galaxies. GOODS data will enable a wide variety of topical investigations, including star formation at high redshift, active galactic nuclei (AGN), the extragalactic background light, type Ia supernovae (SNe Ia) to probe the cosmological geometry, gravitational lensing, and the discovery of galaxies with redshifts up to z~7. The GOODS ACS program will cover two 10 x 16 arcmin regions centered on the Hubble Deep Field North (HDF-N) and the Chandra Deep Field South (CDF-S). The sky area will be larger than most previous surveys using the Wide Field Planetary Camera 2 (WFPC2)--33 times larger than the combined area of the HDF-N and HDF-S, 4 times larger than the combined HDF Flanking Fields, and 2.5 times larger than the Groth Strip Survey. Only the Medium Deep Survey and other Hubble parallel programs covered larger sky areas, and they did so with less sensitivity, fewer filters, and non-contiguous fields. We have chosen two fields to ensure against variance due to line-of- sight clustering effects and to enable follow-up programs by astronomers and observatories worldwide. We will take ACS images in four broad, non-overlapping filters: F435W (B), F606W (V), F775W (i), and F850LP (z), with exposure times of 3, 2.5, 2.5, and 5 orbits, respectively. We will image each field in five visits separated by about 45 days to search for high redshift supernovae. The extended- source sensitivities of the combined images will be only 0.5-0.8 magnitudes shallower than the WFPC2 HDF observations. The target fields are the most data-rich and best- studied deep- survey areas on the sky. In addition to optical and near-infrared imaging, the fields have been subjects of extensive spectroscopic, radio and sub-mm surveys, and the deepest X-ray observations from Chandra and XMM-Newton. The GOODS SIRTF Legacy program (M. Dickinson, PI) will make the deepest observations of these fields at 3.6 to 24 microns. We are also carrying out optical and near-infrared imaging and spectroscopy using facilities at the European Southern Observatory, National Optical Astronomy Observatory, Gemini, and Keck. The HDFs became a scientific phenomenon by attracting many researchers with varied interests and approaches. Indeed, the HDFs catalyzed the deepest observations of all sorts, at many wavelengths, and by the most powerful telescope facilities in space and on the ground. Nevertheless, the HDFs have limitations. First, they are very small fields--5 square arcmin each--probing small co-moving volumes. Second, their wavelength coverage has an important gap in the mid- and far-infrared; ISO and SCUBA observations of the HDF at 7, 15 and 850 microns detect only a small number of the most luminous, dust-obscured objects and do so at wavelengths that miss the peak of dust re-emission. As a result, HDF studies of AGN and galaxies at z>1 are missing both near-infrared, rest-frame light, which traces total stellar mass, and the mid- and far-infrared light, where most of the bolometric luminosity from star formation and dust-obscured active nuclei must emerge. GOODS will extend the lessons and address the shortcomings of the HDFs. The GOODS ACS observations will cover much larger areas, and the GOODS SIRTF data at 3.6 to 24 microns will fill the `infrared gap.' Additionally, Guaranteed Time Observers will acquire SIRTF data down to 170 microns, albeit at brighter flux limits. The GOODS ACS observations will image large samples of galaxies and AGN at 0.5
Figure 1. Layout of the ACS observations for GOODS. The grid of white boxes shows the tiling of ACS fields at one telescope orientation superimposed on the Chandra (outer greyscale) and SIRTF (inner greyscale) exposure maps. We will revisit the fields approximately every 45 days to enable a search for high-redshift, type Ia supernovae. The center inset schematically shows how the ACS field of view rotates and tiles at different epochs of our Hubble Treasury observations.



8




also cover the poorly explored redshift range 1
Progress on WFC3
John MacKenty, mackenty@stsci.edu

T

he WFC3 Science Oversight Committee (SOC) views Wide Field Camera 3 (WFC3) as Hubble's `panchromatic camera.' Scheduled to replace the venerable Wide Field Planetary Camera 2 (WFPC2) in 2004, WFC3 has two channels: an ultraviolet-visible channel (WFC3/UVIS) for imaging a wide field of view (FOV) from 200 to 1000 nm and an infrared channel (WFC3/IR) for imaging from 820 to 1700 nm. WFC3 will ensure Hubble can continue to take superb pictures until the end of the mission, expected in 2010. The SOC--consisting entirely of volunteers selected by NASA1--provides broad scientific advice to the WFC3 development project at Goddard Space Flight Center (GSFC). The SOC developed key scientific objectives for the WFC3 and now strives to ensure the necessary capabilities in WFC3 by providing useful studies, reviews, and recommendations to the project. Prospective users of WFC3 are encouraged to contact any of the SOC scientists for information on WFC3. Exploiting recent gains in detector technology, WFC3/UVIS provides the same 4096 x 4096-pixel CCD array format as the Wide Field Camera (WFC) in the Advance Camera for Surveys (ACS). WFC3/UVIS has the same FOV as the original Wide Field/Planetary Camera (WF/PC) and the current WFPC2, which is somewhat smaller than the ACS/WFC FOV. Built by Marconi Applied Technology, the WFC3 flight CCDs have high QE, excellent near UV sensitivity, read noise below 2.5 electrons, good resistance to damage by cosmic radiation, and near-perfect cosmetic qualities. The projected pixel size is 39 milliarcsec. The WFC3/UVIS offers astronomers a choice of 42 filters, 5 quad-filters, and 1 UV grism. The WFC3/IR employs a 1024 x 1024 pixel mercury-cadmium-telluride (HgCdTe) detector made by Rockwell Scientific. Called `Hawaii-1R,' it is a variant of their well-known `Hawaii-1' device. Rockwell modified the composition of the detector material to place the long wavelength cutoff at 1700 nm, which sufficiently reduces dark current to be zodiacal-background limited in the H band. It also permits thermo-electric cooling. To track system drift and improve calibration, Hawaii-1R incorporates 5 rows of `reference pixels' along each edge. It provides 16 times as many pixels as a NICMOS detector over WFC3/IR's 123 x 139 arcsec FOV, which is 6.6 times larger than the NICMOS Camera 3 FOV. WFC3/IR is equipped with 14 filters and two grisms covering broad and medium bands plus key narrow line features.
1 Bruce Balick, Howard E. Bond, Daniela Calzetti, C. Marcella Carollo, Michael J. Disney, Michael A . Dopita, Jay A . Frogel, Donald N.B. Hall, Jon A . Holtzman, Gerard Luppino, Patrick J. McCarthy, Francesco Paresce, Robert W. O 'Connell (Chair), Abhijit Saha, Joseph I. Silk, John T. Trauger, Alistair R. Walker, Brad C. Whitmore, Rogier A . Windhorst, and Erick T. Young. Edward Cheng (Instrument Scientist) and John MacKenty (Deputy Instrument Scientist) are ex officio members.

Continued page 10



9




Early in the design phase, when WFC3 consisted only of WFC3/UVIS, the HST Second Decade committee recommended the addition of WFC3/IR. The SOC and other advisory committees endorsed this recommendation, and NASA adopted it. NASA is building WFC3 under a stringent cost cap, with strong inheritance from earlier investments. It is reusing designs and components and applying knowledge and know-how already in the Hubble program. A GSFC-led `integrated product team' is building WFC3. This team includes participants in previous Hubble science instrument developments: Institute staff members are providing scientific and ground- system support; experts from Ball Aerospace are applying their experience with optics, electronics, and systems engineering; Swales Aerospace is developing the optical bench and thermal control system; the Jet Propulsion Laboratory (JPL) is procuring and testing the UVIS Figure 1. An isometric view of the WFC3 instrument. filters; and GSFC is characterizing detectors and infrared filters, The outer enclosure and the radiator (bottom) are re-used WF/PC items. The interior optical bench providing integration and test, and managing the project overall. (lavender) is a new structure that contains all the The construction of WFC3 is proceeding on schedule with all optics and detector subsystems. major components in hand for assembly. Spring 2002 will see the installation of the flight optics and mechanisms onto the optical bench at Ball Aerospace. The electronics boxes are currently in assembly and with their completion, the bench assembly will ship to GSFC in late summer. Meanwhile, GSFC has been carefully testing and characterizing the flight detectors. After the installation of the bench into the enclosure and the attachment of the electronics and thermal systems, WFC3 will be extensively tested and calibrated and then delivered to the Hubble project in spring 2003. The design of WFC3 provides for simple exchange of detector assemblies, even late in the program if necessary. This feature gains schedule flexibility (and cost control)--plus the opportunity to fly the best possible detectors. NASA plans two flight detector packages for each channel, to provide a spare in the event of last minute problems. In March 2002, Ball Aerospace is well along in assembling a pair of Marconi CCD detectors for the first UVIS detector package. A second pair awaits the final selection between three outstanding CCD devices. The CCD detector assemblies are direct copies of the ACS assemblies with only minor modifications to ease rapid change-out. The IR assembly is a new design derived from the CCD assembly. Ball Aerospace is presently building a qualification and test unit containing an engineering- grade Rockwell IR detector and will soon start the first flight- detector build. In parallel with the hardware Figure 2. WFC3's optical bench being installed at Ball Aerospace into its alignment fixture. Until development and testing, scientists summer 2002, WFC3's optical bench will reside at Ball Aerospace as its mechanisms and optics are integrated and tested. and engineers from the Institute and GSFC are modifying the operational procedures and software needed to fly WFC3 and bring it into service for GOs. From the observer 's perspective, the WFC3 may look quite familiar. The operational interface is based strongly on the ACS design for the UVIS channel and the NICMOS design for the IR channel. The data processing pipelines are also re-writes of these instruments' CALACS and CALNIC software. Expected enhancements include an on-chip binning capability for the UVIS CCDs (2x and 3x) to better exploit the low background in the near ultraviolet. Another enhancement is sub-array readout of the IR detector to enable short exposures. Interested users should look for the WFC3 mini-handbook later this summer or visit the WFC3 website at http://wfc3.gsfc.nasa.gov for more information.
WFC3 from page 9



10




Cosmic Origins Spectrograph
Tony Keyes, keyes@stsci.edu

T

he Cosmic Origins Spectrograph, COS, will be installed in Hubble during the fifth servicing mission, which is presently expected in early 2004. COS will have separate detectors for the far ultraviolet (FUV: 1150-1775 å) and near-ultraviolet (NUV: 1700-3200 å) spectral ranges. The instrument is highly optimized to perform low-background, moderate-resolution (R~20,000) point- source spectroscopy with broad wavelength coverage and the highest possible throughput in the FUV. (COS will be the most sensitive FUV spectrograph ever flown on Hubble.) COS will operate primarily in time-tag mode. Event lists will have 32-millisecond time resolution and--for the FUV only--will include pulse-height information. Two cross delay line detector segments, each with a 16k x 1k pixel format, cover approximately 300 å of spectral range per exposure at R=20,000. This large detector format, combined with a unique `single reflection' design increases discovery efficiency 4-15 times over equivalent STIS modes. The COS NUV channel design and Multi-Anode Microchannel Array (MAMA) detector provide capabilities similar to the L and M modes on the Space Telescope Imaging Spectrograph (STIS) but with anticipated lower background and up to 2 times increase in discovery efficiency. (See Table 1.)

Table 1.

Summary of available COS observing modes
FUV NUV
25 x 25 1k x 1k 25 x 25 75 x 75 G185M, G225M, G285M G230L 1700-3200 millimeters pixels microns microns 1st-order gratings 1st-order gratings ångstroms

Active Detector Area Digitized Format Pixel Size Resolution Element Size R=20,000 Spectroscopy R=2000 Spectroscopy Spectral Range

Two 85 x 10 Two 16k x 1k 6 x 25 38 x 300 G130M, G160M G140L 1150-1775

The faint-object spectroscopic sensitivity and routine application of time-resolved observing methods to faint sources are unique COS capabilities. Investigations making use of these features will include the helium II Gunn-Petersen effect, the intergalactic medium deuterium-to-hydrogen ratio, the Lymanalpha forest, large - scale structure, the molecular chemistry of the cold interstellar medium, star formation, and occultation studies of planetary atmospheres and cometary tails. Principal Investigator James Green at the University of Colorado leads the COS Investigation Definition Team (IDT). The IDT manages the instrument hardware development project. O. H. W. Siegmund of the Experimental Astrophysics Group at the University of California, Berkeley, leads the FUV detector program. Ball Aerospace integrates the instrument. All gratings and optics are in-house at Ball and meet or exceed design specifications. The flight FUV detector has been chosen and tested under vacuum conditions. The STIS flight- spare NUV MAMA is the COS NUV flight detector. The Institute is currently developing the scheduling system modifications to operate COS. We have completed all science-exposure and target-acquisition modes and are working on calibration exposure modes, such as for wavelength calibration and flat fields. We are also working on the COS Exposure Time Calculator, which users will access through the Astronomer 's Proposal Tool, APT. We have designed, reviewed, and are now building the COS calibration pipeline. We will distribute the COS mini-handbook with the upcoming Hubble Cycle 12 Call for Proposals. This document will provide quantitative descriptions of COS capabilities, measured detector efficiencies, anticipated observatory-plus-instrument throughputs, and comparisons with equivalent STIS capabilities. Our website provides COS information updates: http://www.stsci.edu/instruments/cos/.



11




Specview: A New Tool for Visualizing and Modeling Spectra
Ivo Busko, busko@stsci.edu

W

e are pleased to announce the release of Specview, which is an interactive, quick-look tool for users to manipulate, analyze, and display spectra. We wrote Specview in the Java programming language to exploit its powerful, interactive, graphical user interfaces. Specview ingests all the complex spectral formats that the Hubble spectrographs create and which the Institute delivers to users in FITS format. It also accepts data from other instruments when presented in a simple, text-based format. After ingestion, Specview displays the spectra and retains them in an internal buffer for the user to manipulate, analyze, and visualize further. For example, the user could combine separate data sets into a single spectrum, convert between alternative physical units, or fit spectral models to the data. Specview 's ability to convert physical units readily is essential to supporting multiple instruments. To achieve this flexibility, Specview distinguishes between data units, internal units, and display units. Depending on circumstances, the software converts units automatically or the user does so manually. Specview supports the same spectral units as the Institute's Synphot software. When multiple display windows are open, cursor movements in each window are correlated--also enabled by the ready units conversion. We have build the engine for spectral model fitting around a model manager, which keeps a list of spectral components--objects that implement analytic functions, such as Gaussians, Lorentzians, power laws, and black bodies--in an internal library with 15 functional types. The user builds a model interactively, by selecting component types from the library and adding them to the model at will. Users knowledgeable in Java can add their own custom spectral component types to the library on the fly. Spectral components in turn are defined by parameters, such as line width, black body temperature, and power law index. The model fitting process consists of finding the set of parameter values that minimize the chi- square statistic between the current model spectrum and the raw data. The minimization engine provides a choice of two algorithms: simplex (amoeba) and Levenberg-Marquardt. The fitting process is entirely interactive, which means that the user can exert a good deal of control over individual spectral components, their parameters, and convergence behavior. Because these fitting algorithms are very general--they do not make any restrictive assumptions about the model, such as linearity--and iterative, the user must provide a reasonable first- guess model spectrum to converge. The interactive tools in the model manager make building this first-guess spectrum easier. Since spectral models can be saved, the user can spread the data analysis process over multiple sessions. Specview is available as a standalone program for Solaris, Linux, and Windows at http://specview.stsci.edu. In the future, we will include Specview in the data browser tools of StarView (http://starview.stsci.edu) and Archive Scrapbook (http://archive.stsci.edu/scrap book.html). A future data-processing engine will allow additional operations, such as spectrum arithmetic, filtering, and splicing.

New Tool for Preparing Phase 1 Proposals

T

he Institute has enhanced the Astronomer 's Proposal Tool (APT) to support Hubble Phase 1 proposal preparation for Cycle 12. Users will be able to develop and analyze their proposals with new capabilities provided by APT. We strive to offer the user increasingly intuitive, visual, and interactive means of proposal preparation. We encourage users to try using APT and send us feedback, so that we can continue to improve the product. (http://apt.stsci.edu/).
The APT in Phase 2 mode running the Visual Target Tuner.



12




Next Generation Space Telescope Transitions to a New Phase
Roelof de Jong, dejong@stsci.edu, John Hutchings, Gerard Kriss, Micheal Regan, and Peter Stockman

O

n 17 December 2001, the NGST program formally began the transition from feasibility studies to the detailed design phase (Phase B). The milestone was the Mission Definition Review (MDR), entailing a three-day meeting held at the Institute and the Goddard Space Flight Center (GSFC). The MDR is part of the process that confirms NASA's preparedness to commit funds and award contracts for the initial development of a mission. Dennis Dillman chaired the MDR panel from GSFC. Because of the importance and scope of the NGST Program, NASA Headquarters (HQ) supplemented the MDR panel with an Independent Review Team (IRT) chaired by Jean Olivier, the former Chief Engineer for Hubble and Deputy Project Manager for Chandra. John Mather, the Senior Project Scientist for NGST, presented the scientific case for the mission as developed by the Ad Hoc Science Working Group (ASWG). He explained and justified the associated mission requirements. Bernie Seery, the NGST Program Manager, provided an overview of the entire project, from advanced technologies to international partnering. The remainder of the presentations to the MDR covered each of the major program elements in detail, including restricted executive summaries of the mission architectures proposed by the two possible Prime Contractors. The overall response of the two review committees has been positive. Nevertheless, reflecting NASA's emphasis on mission success, the review teams recommended that the NGST Project reconsider some of its assumptions in the design of the mission and its plans for allocating development responsibilities. With the selection of the Prime Contractor nearing completion and the selection the NIRCam team underway, this is a good opportunity to revisit these issues as well as the scope and schedule for the NGST program. An international partnership meeting will be held in May and other meetings have been planned to address the other MDR/IRT concerns. Following the MDR, Seery expressed his confidence in the readiness of the program to move into Phase B and that "our vision of NGST is solid." NASA's Langley Research Center has started an independent cost estimate of the NGST development. Meanwhile, the President 's budget submission to Congress includes the funding for the next three years of development. By this summer, the NGST program will have established all the major development partners, including the NGST Science Working Group.

NIRSpec Developments
Many programs in the NGST Design Reference Mission (DRM) require the Near Infrared Spectrograph (NIRSpec). The DRM is a core set of science programs to test the NGST design by representing typical requirements of future NGST observers. The DRM reflects that observers will want to use NIRSpec to investigate galaxies at all redshifts, to study their formation, clustering, and chemical abundances, and to observe their kinematics and star formation. They will also want to use NIRSpec to study active galactic nuclei, young stellar clusters, and the initial mass function of stars. The ASWG recommended that NIRSpec be able to obtain spectra of more than 100 objects simultaneously in a 9- square-arcmin field of view. The baseline design has micro-electro-mechanical systems (MEMS) providing dynamic aperture masks to select targets over this full field of view. The spectra will cover the 0.6 to 5 micron wavelength range with selectable resolving powers of ~100 and ~1000. The NIRSpec aperture and detector pixels will have ~100 milliarcsec projected size, which is well matched to the size of high redshift galaxies. The NIRSpec pixels are larger than those in the NIRCam to provide a larger field and improve sensitivity. The detectors will be either indium antimonide (InSb) or mercury-cadmium-telluride (HgCdTe), depending on technology studies currently underway. The European Space Agency (ESA) will be providing the NIRSpec instrument as part of their contribution to NGST. NASA will provide the detectors and the MEMS aperture masks. In November 2001, ESA reviewed the results of four independent studies of concepts for the NIRSpec. Teams from Alcatel (Cannes)/Laboratoire d'Astrophysique de Marseille and Astrium (Ottobrun)/Laboratoire d'Astrophysique de Marseille presented two concepts based on MEMS devices. Both of these concepts were consistent with the recommendations Continued of the ASWG report. These two teams have now begun definition studies to page 14 develop more detailed designs, which they will present in the fall of 2002.



13




New Phase from page 13

Figure 1. A scanning microscope image of micro- shutters, which is the selected MEMS concept for NIRSpec. The shutters are 100 microns center-to-center and about 90 microns wide. All shutters are slightly magnetized and are opened by scanning a permanent magnet over the array. Selected apertures are held open electrostatically, by applying a voltage difference between the shutter and an electrode on the wall. After the magnet has passed by, the resilience of the hinges flips the remaining shutters closed. (Note that these shutters, unlike the flight devices, do not have shields to prevent light from passing around the edges.)

Because MEMS technology is still developmental, ESA is pursuing backup concepts for NIRSpec in parallel with the MEMS -based concepts. A team from Astrium (Ottobrun)/Laboratoire Astrophysique de Marseille presented a design for an integral-field spectrograph. A team from Astrium (Toulouse)/Centre Suisse d'Electronique et de Microtechnique (CSEM) (Neuchatel) presented several alternative slit-mask concepts. While the mechanical slit-mask concepts would not provide the desired 100 objects per 9- square-arcmin field, they use mature technology and offer a multi-object capability that is qualitatively superior to a simple long- slit spectrograph. Three separate groups have been developing the enabling MEMS technology by different approaches and techniques. Each group is aiming toward arrays of up to 2,000 x 2,000 micro -mirrors or micro shutters, each individually controllable. Two concepts are based on movable mirrors, similar to those used in modern computer- screen projectors. The Institute's John MacKenty is developing CMOS controlled aluminum micro-mirror arrays in collaboration with GSFC. Sandia National Laboratories are developing a micro-mirror concept based on polysilicon. Harvey Moseley of GSFC leads the development of an alternative concept, using `micro- shutters' rather than mirrors. In order to concentrate resources and to develop the best MEMS concept as quickly as practical, the NGST project convened an independent, international review board in December 2001 to select the most promising development efforts. The review board selected Mosley 's micro- shutter concept as the sole MEMS concept to continue in the NGST development path. This technology is illustrated in Figure 1. The micro- shutter concept is completely transmissive and provides inherently better contrast. It also alleviates concerns in the micro -mirror approaches about focal-ratio degradation due to out- of-plane diffraction and curved facets. In the summer of 2003, NASA and ESA will decide whether a MEMS -based NIRSpec concept will be developed for flight.

The Fine Guidance System for NGST
Under project management of the Canadian Space Agency (CSA), EMS Technologies and the Herzberg Institute of Astrophysics (Project Scientist John Hutchings) are currently designing the NGST Fine Guidance System (FGS). Unlike earlier concepts that used NIRCam for guiding, the current design calls for an FGS separate from the NGST science instruments. This separation optimizes the speed and precision of guide- star centroiding. Because the detailed FGS performance requirements depend on the telescope design, final FGS specification await the selection of the NGST Prime Contractor. The current baseline requires the FGS to report guide- star centroids to the telescope control system with a precision of 3 milliarcsec at a rate



14




of 20 Hz. The baseline requires a 95% probability of finding a usable guide star in the FGS field of view anywhere in the sky, even at the low star density regions near the galactic poles. The FGS must also be fully redundant, providing its own backup system. After investigating several FGS concepts, the current baseline calls for two separate, fully redundant FGS units in the NGST focal plane. Each unit will re-image an 8.4- square-arcmin field on a pair of 2K x 2K detector arrays with a 60 milliarcsec pixel size. The NGST image quality permits using an unfiltered camera to maximize the guide signal. The FGS detectors will use one of the two near-infrared detector technologies currently being developed for NGST, as discussed in the winter 2002 Newsletter. The two technologies are being tested for FGS requirements, paying special attention to detector differences relevant for the FGS, such as pixel size, read noise, read-out speed and intra-pixel response for moving-target tracking. The FGS team has optical designs to cover the two detector options as well as changes in the field of view if needed for guide star accessibility. To ensure adequate guiding signal and to achieve the all-sky 95% success rate, the team has investigated star densities Figure 2. The current design concept for the NGST Fine and fluxes in the near infrared, and it is working to improve our Guidance System. knowledge of stars in the galactic halo. In the coming months, the team will update the FGS design with new information about detector performance, required centroiding precision and update rate, and the telescope collecting area and optical performance. It will also study how best to place the FGS units in the focal plane for guiding and how to package them among the other focal plane instruments. To stay up -to - date with all the NGST developments, we recommend you regularly visit our website: http://www.stsci.edu/ngst/.

News from the Multimission Archive at STScI (MAST)
Paolo Padovani, on behalf of the MAST team, padovani@stsci.edu

T

he Hubble data archive now contains about 7.6 terabytes of data in about 258,000 science data sets. The archive ingests an average of 3.5 gigabytes per day. Lately, researchers have been retrieving data from the archive at a rate almost 6 times higher, about 20 gigabytes per day.

Cycle 11: Hubble Treasury, Large, and Legacy Archival Programs
At MAST, we are making plans to archive sets of contributed data coming from the Hubble Treasury, Archive Legacy, and Large programs selected in Cycle 11. We are creating a new class of files in the archive called `contributed,' for these data, which will take the form of reduced images, object catalogs, tables of measurements, and other products, including fitted galaxy profiles and synthetic stellar spectra. All these contributed data will be in FITS format files, which can be sought and retrieved in MAST in the same way as Hubble data files. For example, a search of the archive on the term `Eta Carinae' will return both the unprocessed Hubble images and the files contributed by the Hubble Treasury team selected to observe that star in Cycle 11. MAST will provide some description of the individual programs and their contributed files, and it will provide links to the proposal teams' World Wide Web (WWW) sites, where they will describe their programs and contributed data in Continued more detail. Hubble Treasury Program data will have no proprietary period. page 16 (See accompanying articles on the selected Hubble Treasury programs.)



15




MAST from page 15

Co-plotting IUE Spectra
We have added a new feature to the page returned by the International Ultraviolet Explorer (IUE) search interface. This feature enables the user to compare multiple spectra by co-plotting them on the same axes. For example, the user can examine observations of different targets together or compare observations of the same target taken at different times, as illustrated in Figure 1. The IUE search page is http://archive.stsci.edu/cgi-bin/iue. Eventually, the co-plotting feature will be added to the other search interfaces.

Flux (ergs cm2/s/å)

0 1200

5x10

-13

10

-12

1.5x10

-12

1300

1400

1500 1600 Wavelength(å)

1700

1800

1900

Figure 1. A series of 9 IUE spectra of the planetary nebula Abell 35 obtained in March 1992. The plot shows one discrepant spectrum, namely SWP 44120. GO comments in the FITS header note that tracking was temporarily lost during this one observation, so that the actual exposure time is unknown.

The co-plotting script extracts the fluxes and wavelengths from input files selected on the form returned from a search and plots them as a GIF-format image displayed on a web browser form. The form offers the user options to redraw the page with different scaling factors. Currently, up to 15 spectra can be co-plotted, with each spectrum rendered with a different color and descriptive information shown below the plot. In the winter 2002 Newsletter, we described the new tool for co-plotting in another context, as an enhancement to the MAST Spectral/Imaging Scrapbook (http://archive.stsci.edu/scrapbook.html). The scrapbook is a WWW tool that permits the user to peruse representative spectra or sky images from mission data stored in the MAST archives. The addition of the co-plotting tool to our search interfaces opens up new ways to exploit the growing MAST.

Sloan Digital Sky Survey (SDSS) Quasar Catalog
The SDSS Quasar Catalog is now available at http://archive.stsci.edu/sdss/quasars/. This catalog consists of 3,814 objects, 3,000 of which discovered by the SDSS and contained in the initial public release. Each object in the SDSS Quasar Catalog has at least one emission line with a full width at half maximum larger than 1,000 km/s, a luminosity brighter than M(i*)=-23, and a highly reliable redshift in the 0.15 to 5.03 range. The sky area covered by the catalog is 494 square degrees. A multicolor selection technique found most cataloged objects in the SDSS commissioning data. The SDSS Quasar Catalog is now preferred; it supersedes the quasar catalog in the SDSS early data release. Nevertheless, the sample is not homogeneous and is not intended for statistical analysis. From the SDSS Quasar Catalog website, Archive users can retrieve three related units of information: the catalog itself in ASCII format, a gzipped (compressed) tar file containing calibrated spectra from 3800 to 9200 ångstroms of all objects in FITS format, and the published paper (Schneider et al., 2002, AJ, 123, 567) from the NASA Astrophysics Data System. We developed a web-based search interface to enable archive users to search the quasar catalog and retrieve individual spectra. (http://archive.stsci.edu/cgi-bin/search/sdss_edr_quasar). We have made both a `simple' and an `advanced' version of the interface available.



16




MAST to Archive Kepler Data
Kepler is a Discovery Program mission designed to detect terrestrial planets around stars in the Sun's neighborhood. In December 2001, NASA selected Kepler for development, with launch planned for 2007. The Kepler instrument will detect transits by planets across the disc of the host star. Planet transits cause a fractional reduction in stellar brightness - of 5 to 40 x 10 5 and last for 2 to 16 hours. Investigators can calculate the sizes of the orbit and the planet sizes from the period and depth of the transit, respectively. Kepler will monitor the brightnesses of 100,000 stars simultaneously and continuously for 4 years, with a cadence of 15 minutes. The targets will be A-K dwarf (main- sequence) stars brighter than 14th magnitude. At their option, NASA may fund a participating- scientist program to allow observing an additional 3000 objects that are not of interest for planet searches but lie within Kepler 's 110- square-degree field of view. NASA may also support an archival data analysis program. The Institute will archive Kepler 's approximately 5 terabytes of image-level data and provide basic data calibration. The primary data product--about 100 gigabytes in volume--will consist of extracted, very high precision differential time- series photometry for each of the target stars. More details on the Kepler mission can be found at http://www.kepler.arc.nasa.gov/.

Astrometric Planet Detection using Hubble's FGS
Ed Nelan, nelan@stsci.edu, and Melissa McGrath

O

ne of the most exciting discoveries in recent years is the detection of what may be planets around other sun-like stars. Today, over 80 stars show telltale wobbles in the Doppler shifts of their spectra caused by the tugs of unseen companions. But do these companions have planet-like masses or are they heavier--perhaps brown dwarfs or even low-mass stars? Addressing this important question helps us understand how the new discoveries relate to our Solar System. As other planetary systems are identified and characterized, we will place our own planet, Earth, in context and better appreciate how common--or rare--life may be in the Galaxy. The companion mass is somewhat ambiguous for Doppler shift detections because the inclination angle (i) of the orbit cannot be determined from these observations. For example, if we view a system edge-on, the maximum of the periodically varying radial velocity is the true speed of the star, from which we can compute the companion's mass as a fraction of the known stellar mass. However, the Doppler shift caused by a massive companion in an inclined orbit can duplicate that of a smaller mass aligned edge-on. Therefore, without knowing a system's inclination, we can only place a lower limit on the companion's mass, Mmin = Mtruesin(i). To date, we know the orbital inclination for only one extrasolar planet, the companion to the star HD 209458--it is effectively 90 deg (edge-on)! Using high-precision photometry, astronomers have observed the minute reduction in starlight as the companion transits the disk of the star.1 In this case, the minimum mass--0.69 +/- 0.05 MJupiter--is the true mass. Furthermore, the amount of light cut off by the companion provides a measure of the companion's size, which is 1.35 +/- 0.06 Jupiter radii. HD 209458b is a planet! But what can we learn about the other candidates with less fortuitously aligned orbits? If orbital inclinations of the 80-plus systems are randomly distributed, then, on average, Mtrue = (4/) Mmin. In a statistical sense, the true companion mass is not likely to be much greater than the measured minimum mass. For example, if a companion with Mmin = 1 MJupiter is actually a low-mass star (Mtrue > 80 MJupiter), the inclination angle must be less than 0.7 deg, which is very unlikely. Nevertheless, selection effects may skew the distribution of inclinations toward low values, in which case the companions might include some low-mass stars or brown dwarfs. The true masses of these companions-- and their interpretation--remain important and controversial scientific topics. The skeptics of calling the radial-velocity companions `planets' present three lines of evidence. First, the distribution and correlation of their orbital eccentricities and periods are statistically indistinguishable from binary star systems. Second, some of the host stars seem too bright for their spectral type and distance, like unresolved binary stars. Third, measurements by Hipparcos suggest astrometric wobbles for some of these stars that require the companions to be low-mass stars.
Carbonneau, D., Brown, T., Latham, D., Mayor, M., ApJ, 529, L45-48, 1999. "Detection of Planetary Transits Across a Sun-Like Star."
1

Continued page 18



17




Planet Detection from page 17

(Unlike radial velocity studies, astrometry measures the orbital inclination and determines the true companion mass.) In the case of one star, Cor Bor, a ground-based astrometric instrument (the Multichannel Astrometric Photometer (MAP) at the Allegheny Observatory) found a wobble consistent with the Hipparcos measurement, indicating that the companion is a low-mass star, i.e., a non-planet. Radial velocity observations of another star, 1 Cancri, found a companion with Mmin = 0.88 MJupiter in a circular orbit with a period of 14.65 days. On the other hand, Hipparcos data suggest that 1 Cancri has a reflex motion with a radius of 1.15 milliarcseconds (mas), implying that its companion is an M dwarf with the mass of 126 Jupiters. The measurements made by Hipparcos and MAP are near the detection limits of these instruments. Clearly, a more sensitive instrument is needed to confirm or refute their results. Until the launch of the Space Interferometry Mission (SIM) at the end of the decade, the only such instrument available is Hubble's Fine Guidance Sensor (FGS). In 2001, we used the FGS to observe 1 Cancri astrometrically. Because the reflex motion of a star with a low-mass companion is tiny compared to its proper motion and parallax, it normally takes observations over a whole year or more to isolate the small wobble. Nevertheless, the short period and circular orbit of 1 Cancri's companion permitted us to use a novel observing strategy that could detect any reflex motion larger than about 0.3 mas in just one month. (Detecting a motion of 0.3 mas is like measuring a shift of 1 inch from 10,000 miles away!) We carried out a set of observations on 1 Cancri in the one-month period centered on the time of maximum parallax factor in right ascension, May 1, 2001. We observed at times symmetric about this date and phased with the companion's orbital period, spanning two full orbits. By comparing 1 Cancri's position relative to other stars in the field, and by using only data collected at the same parallax factor (equal times before and after May 1), we could measure the star 's proper motion without needing to determine its parallax. We did this for three pairs of epochs: +/- 14, +/- 10.5, and +/- 3.5 days offset from May 1. A measurable reflex motion would manifest itself as a systematic, periodic variation of 1 Cancri's proper motion determined from the paired observations. The quality of our astrometric data was excellent. We were able to determine the proper motion of 1 Cancri relative to other stars in its vicinity with an accuracy of about 0.02 mas/day. We found no evidence of any perturbation to the star 's position, as illustrated in Figure 1. In our recently published results2, we place a 3 upper limit to the companion mass at Mtrue < 30 MJupiter, well below a stellar mass and within the brown dwarf range.

Figure 1a. The measured position of 1 Cancri relative to field stars with the proper motion and parallax removed. The numbers next to the symbols indicate the epoch of the observations, which were made every 87 hours (1/4 of the companion's period). Figure 1b. The predicted positions of 1 Cancri for a reflex motion of 1 mas radius (the Hipparcos result). These would have been easily seen in the FGS data (compare to figure 1a).

The FGS is being used to measure the wobble of three other stars with suspected planetary companions. This includes GL876 (PI Forveille), the nearest and only M dwarf with 2 companions. The two other stars are Upsilon Andromedae, which is thought to host 3 companions, and Epsilon Eridani, which has a companion in a 7-year orbit (PI Benedict). Operating at an astrometric accuracy of a few-tenths mas, Hubble's FGS will make useful studies of both long- and short-period low-mass companions for years to come.
2 McGrath, M.A ., Nelan, E., Black, D.C., Gatewood, G., Noll, K., Schultz, A ., Lubow, S., Han, I., Stepinski, T.F., and Targett, T., ApJ, 564, L127-130, 2002. "An Upper Limit to the Mass of the Radial Velocity Companion to 1 Cancri."



18




A Surprise within a Surprise: Hot Gas around High-Velocity Clouds and a Corona around the Milky Way
Kenneth Sembach, sembach@stsci.edu ne of the fun things about being an astronomer is the finite possibility of stumbling on something unexpected while treading beaten paths toward the questions posed in our observing proposals. It is the new twist or turn in the road, the unplanned chance to find something by exploring just a bit further, that more often than not brings an extra smile at the end of the day. This is the brief tale of one such smile. When the Far Ultraviolet Spectroscopic Explorer (FUSE) satellite was launched in late June of 1999, we had a carefully crafted plan to study the properties of hot gas in the halo of the Milky Way galaxy using quasars and active galactic nuclei as background sources for our far-ultraviolet observations of O+5 absorption. (O+5 is an excellent diagnostic of collisionally ionized gas at temperatures of a few hundred thousand Kelvins.) The observing plan stipulated that we would observe approximately 10-20 sight lines along complete paths through the Galactic halo, and since our initial time allocation for such observations was modest, the target list was limited to some of the brightest known extragalactic sources of far-ultraviolet light. The primary goal of the program was to determine the distribution and properties of the hot gas within a few kiloparsecs of the Galactic disk, with a secondary goal of determining whether there was any hot gas associated with the high velocity clouds (HVCs) detected previously through their H I 21-cm emission or through their ultraviolet absorption recorded by the Hubble Space Telescope. We expected that some HVCs, particularly those within a few kiloparsecs of the Galactic disk, would show O+5 absorption, while more distant HVCs would not. The reasoning behind this seemed sound--the nearby HVCs have kinematics that can be reasonably well described if they are subject to the same processes, such as supernovae, which circulate and heat the gas in Galactic halo, while more distant clouds would presumably be removed from such processes. We set to work, but a few months after launch it was clear that our observing plan would change-- for the better--as a result of the difficulties encountered in maintaining the co-alignment of the four FUSE spectrographs. To accomplish its objective of finding low-redshift Lyman-limit systems for D/H studies, the FUSE team increased substantially the observing times needed to obtain spectra for about 100 AGNs and QSOs. This change had immediate ramifications for our Milky Way hot gas program since the O+5 lines occur at 1032 å and 1038 å , in the more stable lithium-fluoride channels where the effective area of FUSE is highest. What at first seemed like a stumbling block for FUSE mission operations quickly turned into an unplanned bonanza of data, which would not only prove valuable for its original purpose but would also turn what would have been marginal O+5 measures into an excellent data set for studies of high velocity clouds and the Galactic halo. Furthermore, since most of the D/H sight lines were chosen without regard to the expected properties of hot gas along the sight lines, the resulting sample of sight lines was scattered more widely across the sky than it might otherwise have been. Together with my colleagues Blair Savage, Bart Wakker, Philipp Richter, and Marilyn Meade at the University of Wisconsin, I watched the data come in and the sample of HVCs grow. As the sight lines were observed, slowly at first and then at a frenetic pace, we realized that the number of sight lines exhibiting O+5 absorption at high velocities was greater than we had previously expected. (`High - velocity ' is defined loosely here as |vLSR| 100 km s 1, with a few exceptions covering slightly lower +5 velocities.) We found O absorption in several well-known HVCs seen in 21-cm emission. We found it in directions where there was no detectable 21-cm emission at high velocities. We found it in the Magellanic Stream and in Complex C, a region of high velocity H I covering nearly 200 square degrees of the northern Galactic sky. When we culled the sample for the 100 best observations, we found high velocity O+5 in 61 of them. Our first surprise was that so many high velocity clouds contained O+5. Most of the information about high velocity clouds has come from H I 21- cm observations of the neutral gas, but here was clear evidence that ionized gas is also important and common. To create O+5 requires energies in excess of 114 keV, which generally means that the oxygen must be collisionally ionized. To create O+5 by photoionization is difficult; there simply aren't enough high energy photons escaping from the outer layers of hot stars to produce the amount of O+5 observed, and those photons that do escape are quickly absorbed by the nearby interstellar medium. Furthermore, photoionization of O+5 requires very low densities, even for photons from the integrated light of the extragalactic ultraviolet background, and this in turn requires very long path lengths (>100 kpc) to produce the typical Continued O+5 column densities observed. Thus, collisional ionization is definitely the favored page 20 +5 mechanism for producing the O . But what is the source of the collisions?

O



19




This question led to the second surprise. To approach the answer, consider the O+5 observed in Complex C and the Magellanic Stream. In both cases the internal velocity gradients are too low to produce O+5 by colliding clouds of neutral gas together. (A typical shock speed of ~170 km s-1 is required to produce significant amounts of O+5.) Furthermore, both regions of gas are located well within a few hundred kiloparsecs of the Galaxy, so photoionization is not viable. Thus we were led to the surprising conclusion that the clouds must be interacting with an external medium--a corona around the Milky Way! What must be the properties of this corona? It must be sufficiently diffuse (n 0.0001 cm-3) to escape detection by other means (e.g., X-ray emission) and to prevent substantial decay of the Magellanic Stream orbit on timescales of a few hundred million years. It must also be hot (T >106 K) to avoid detection in the same O+5 lines used to find the high velocity clouds. Finally, since the Magellanic Stream is located at distances of tens of kiloparsecs, this corona around the Milky Way must be large--possibly extending out to the Magellanic Clouds or beyond. As the high velocity clouds plow through, interfaces form between the cool/warm gas of the clouds and the hot gas of the corona. The O+5 observed most likely occurs at these boundaries as the hot and warm gases come in contact, mix, and collisionally ionize the boundary layer. Figure 1. A schematic showing high velocity clouds interacting with the In the 1950s, Lyman Spitzer postulated a halo Galactic corona. of hot gas around the Milky Way, but his halo is much smaller in extent than the corona revealed by the high velocity O+5 measurements. Three Examples of High Velocity O VI Absorption Spitzer 's halo was revealed by absorption line measurements of C+3 and N+4 with Hubble in 12 the 1990s. It is roughly characterized by a plane 10 8 parallel layer with a scale height of just a few 6 kiloparsecs, compared to tens of kiloparsecs 4 2 inferred for the more extensive corona. Star0 formation activity in the Galactic disk feeds the 1020 1025 1030 1035 halo with hot gas, while the corona is probably a 12 remnant from the formation of the Galaxy. (At 10 8 the coronal temperatures and densities, the 6 cooling time for the gas is comparable to the age 4 of the Galaxy.) Models of galaxy formation predict 2 0 the presence of large amounts of hot gas near 1020 1025 1030 1035 galaxies. It will be interesting to see how the 4 corona we found relates to that `cosmic web' of hot 3 gas so prevalent in our thinking of galaxy formation. 2 The observations of clouds falling into the 1 corona indicate that the Milky Way continues to accrete material. Many of the infalling high 0 1020 1025 1030 1035 velocity clouds seen in O+5 are still a mystery. They may be fragments of smaller galaxies, such Wavelength(å) as the Sagittarius dwarf, gravitationally torn +5 Figure 2. Three examples of O (O VI) absorption in the Magellanic apart by the Milky Way, or they may be gas left Stream (top), the Milky Way halo (middle), and Complex C (bottom). over from the formation of the Galaxy and the Additional interstellar medium lines and intergalactic medium (IGM) lines corona. As Hubble observations of some of are indicated. In the top panel, the bar above the line label indicates the these clouds become available, we expect that locations of interstellar H2 lines. they will continue to produce surprises.
A Surprise from page 19



Flux (erg cm-2 s-1 å-1)

20




Interview: Megan Donahue
Megan, please tell us how your interest in technical subjects--science particularly--arose and grew. What were your sources of encouragement, and what obstacles did you have to overcome? Did astronomy stand out as an early focus? What were the critical points in your astronomical career? stronomy and physics engaged me as far back as I can remember. My first `crime' was stealing a little magnet from my kindergarten class. I was caught almost immediately because I couldn't resist showing my father how to magnetize a needle. When I researched careers, I was always drawn to `astronomer ', but I couldn't imagine how one possibly made a living at it. In those days, one requirement was that an astronomer had to be fluent in French, German, or Russian. Coming from Nebraska, the likelihood of being fluent in anything but English was remote, especially given my talent in languages. Resolved to be practical as an undergraduate, I chose MIT and completed a physics major. Even though I did an undergraduate thesis in X-ray astronomy, I was convinced by my peers (who were mostly computer engineers bound for job security and stacks of cash) that I was doomed to flip burgers alongside the history majors from up the river if I committed to astronomy in graduate school. But the admissions process committed me anyway: I wasn't admitted to a physics program of comparable prestige, so off I went to the University of Colorado at Boulder. My early science experiences gradually accumulated into an `AHA! I gotta be a scientist ' kind of revelation. I liked strange things like ham radios and antennas. I liked the meeting of physics concepts with experimentation and loud noises and explosions (after the age of ten). I don't know why, but when occasionally experimentation reveals physical laws, the collision always surprises and delights me at some level. My greatest challenge was trying to convince myself I had the right stuff, and to keep on plugging in the face of dire employment predictions. I knew that as long as I liked what I was doing right then, what happened in the future would matter less. You worked for IBM for a summer after your undergraduate degree, which gave you a view of research in an industrial context. The Institute is another environment that differs from a traditional university setting for astronomical research. What was distinctive about work at IBM, and what perspectives on the Institute did you gain as a result? I learned that the PhDs have all the fun when it comes to research. I didn't have the background then to take off on my own research but at IBM, the PhDs were the ones who defined their research paths-- in context with the company goals. We had access to the most amazing state-of-the-art equipment and computers. I didn't like the corporate environment, with the emphasis on security. ("Security is a Good Feeling" had replaced "Think " as the company motto all through the halls). And a parking lot with 30,000 cars in it is something to behold. When people call the Institute `corporate', I have to laugh. True, the Institute has a structure unlike a university and a distinctive goal, to provide the best possible access to the Hubble Space Telescope. But beyond having an organizational mission and a semblance of corporate structure, the Institute isn't corporate. We don't answer to stockholders; we don't have to pop out a new product on the market every 6 months; we don't suspect the Japanese of espionage; and a large part of my work is doing research that very likely will not lead to new products and inventions. Your research is focused on clusters of galaxies as evolving structures and settings for physical processes and phenomena on a very large scale. How did this focus come about? How will the future research capabilities of the Next Generation Space Telescope (NGST) and the National Virtual Observatory (NVO) play a role in your Continued research? What breakthroughs do you anticipate in our understanding page 22 of galaxy clusters in the next decade?

A



21




Interview from page 21

I started working on X-ray clusters as an undergraduate with Claude Canizares and Tom Markert at MIT. X-ray astronomy wasn't a big field in the late 80s, so I did my graduate work in theory of the intergalactic medium and some ground-based observations of cluster galaxies. X-ray telescopes give us access to a parameter space that reveals the physics of what is happening in the intergalactic gas in clusters. I enjoy making the connection between theory and observations, a connection that is fairly tight in X-ray studies, due to the ability to extract spectra and images from the same data. I didn't make the connection between clusters and cosmology until well into my postdoctoral studies. When all the substructure studies were claiming to be the cosmologically interesting aspects of clusters, I began losing interest. Nevertheless, I re-engaged when the cluster temperatures of the distant clusters became the cosmologically interesting property. Temperatures are much easier and more satisfying to quantify than substructure. Archival data has always played a role in my research. Clusters and other astronomical objects usually manifest themselves at multiple wavelengths. Assembling clues about the nature of clusters and galaxies requires plumbing the archives not only for X-ray data, but also for radio, infrared, ultraviolet, and even optical data. Putting the pieces of a puzzle together into a coherent picture of what might really be going on in that far away place has always been a happy activity for me. NGST will present some very exciting capabilities for studying clusters--particularly clusters at very high redshifts. NGST will enable multi- object spectroscopy, which could measure the velocity dispersion of a cluster of galaxies at redshift z=2 or more! The most distant X-ray clusters known today are around z=1.2-1.4, although clusters may exist around radio galaxies at higher redshifts. Nevertheless, it is very difficult now, even at z=1.3, to obtain enough redshifts to derive a velocity dispersion to tell you what the mass of the cluster is. NGST will change that. It will also have the capability to trace the 3-D structure of the universe by studying filaments of galaxies out to a very high redshift. Getting out to z>1.5 means that we'll be seeing structure formation at the time it is presumably happening. We expect structures to be growing more rapidly at that epoch of time--but expectations are no substitute for seeing if we're right. You are raising two young children with your husband, Mark Voit, who is also an astronomer at the Institute. How do you two manage it? I think most parents might agree with the following statement: you do what you gotta do. The main thing we try to do is keep the priorities in order and at a minimal number. So maybe a room stands in need of wallpaper and paint for a few years; a fragment of sample fabric might hang instead of a living room curtain for 5 years (to the immense chagrin of my mother in law). I used to be involved in a track club and volleyball competitions, but those activities are incompatible with active 4 and 7 year olds. We are quite involved with orienteering, one of the main reasons is that both of us can compete because start times for the races are set individually, and the kids love playing in the woods. We've developed a routine where one of us drops off the children at daycare and the other picks them up. Both of us are night owls, so having remote access to the computer system after hours has been key when additional work is needed. We do have to plan for crunch times when late hours at the Institute might be necessary, because it is extremely difficult for both of us to have a crunch time at once. It has been lifesaving to have terrific and reliable daycare. The one thing I never worry about in daily life is whether or not my children had sufficient love, stimulation, and security during their day. My seven year old, Michaela, is also refreshingly articulate about her needs, so I do not worry about her wasting away in some program that does not suit her. It also has been extremely critical that both Mark and I share the responsibilities at home. It is likely that Mark even over-compensates, after witnessing my efforts in pregnancy, delivery, and childcare. He does most of the cooking and food shopping. My daughter is not sure that I know how to do those things. In the last 5 years, we've had a regular cleaning service. If there's any one thing I could recommend to another scientist feeling overwhelmed by household chores, it is to take advantage of the fact that we're at least paid well enough to pay someone else to get some of the housework done. You are a serious runner, as are Mark and several other Institute staff members. How organized is your running, and is there an Institute `banner ' under which you sometimes compete? Is there a `zen' of running that you find supportive of your intellectual activities? I used to be involved with a running group, and I miss the comradeship. I also used to run with Mark which was a great bonding experience, but now I mostly run alone, taking turns on short runs during the week and longer runs on the weekend. I've run two marathons, Pittsburgh in 1999 and Boston in 2000. (I started Chicago in 2001, but I had to step off due to injury.) Marathons take more planning than simply running for the occasional 10K or 5K, but I liked the discipline required to follow a 16-week plan and the pay-off in vastly increased capacity to run long distances. The marathon provides the bigger challenge, physically and mentally, and in the end rewarded me with a brightly lit, acutely remembered experience of achievement.



22




There is a `zen' to running. My sister in law and her husband are Buddhists and practice meditation on a regular basis. My lizard brain is far too hoppy to sit and be, but, yes, distracted by a regular physical activity, I can feel perhaps a little of that `zone'. Running enhances my capacity to think. I'll think about work for about the first 30 minutes, but after that I start to zero in on the running itself. The longer I run, the more I'm thinking about what is happening to my body, the light of the twilight-illuminated clouds, the birds wheeling over Loch Raven Reservoir. And those are precious solitary times for me. There's no other time that I can simply claim for my own. I've managed to fit it in during pregnancy, during observing trips, during travel to meetings. I've never regretted finding the time. I think the fact that Mark is an extremely talented and dedicated runner makes this choice very easy. What is the most head- spinning revelation you can make about yourself? OK, Barbara, here it is. I was the bass guitarist for an all-female heavy metal band called Persia. We made recordings in Boston. We practiced in the same studio as the band eXtreme, whose lead singer (Gary Cherone) became even more famous as the lead singer of Van Halen for a while. I also remember picking through the remains of Aerosmith's studio and finding one of their own songbooks in it. I had bleached hair that was occasionally blue, and I liked the lace and leather combination. Boulder mellowed me out a lot; there I played in a more conventional garage/grungy rock band (think REM- style) called No Profit, and they were serious about the name, I don't think we ever broke even on a gig. I still own my guitars, but my kids hate their sound. Now that we've fixed up the basement, I suspect the family will move my rig and my guitars downstairs and restore the living room to its proper respectability.

The FOC Story
Duccio Macchetto, macchetto@stsci.edu

Background

T

he Hubble Space Telescope concept had its origins in the writings of Hermann Oberth in the 1920s and Lyman Spitzer in the 1940s. They had suggested that astronomy could benefit greatly from a telescope that viewed the universe from above Earth's atmosphere. In the early 1960s, interest increased in astronomy as a scientific discipline to be pursued from space, and momentum grew for development of a large orbiting telescope. In 1962, NASA asked the Space Science Board of the National Academy of Sciences to study and recommend future astronomy payloads. In 1966, the Space Science Board recommended that NASA develop a large space telescope. In the autumn of 1971, NASA began to do serious feasibility studies of a 3-m aperture telescope called the Large Space Telescope (LST). The study results were favorable, and preliminary design was initiated in 1972. NASA Headquarters (HQ) and Project personnel had approached ESA in late 1973 with a request to consider participation in LST in several different areas: spacecraft hardware, scientific instruments, and science operations. At the time, NASA and ESA already had an ongoing partnership in the International Ultraviolet Explorer (IUE) project. Within ESA , I was the IUE Project Scientist and Manager. In view of my expertise in having designed the IUE telescope and spectrograph and because of my good contacts with our NASA colleagues, I was asked to lead the ESA team that defined the scope of the collaboration. Over a two-year period, our joint NASA/ESA team discussed a large number of options, concentrating on the potential provision of a scientific instrument to be placed in the telescope's focal plane. We considered a number of instruments, including a two - dimensional spectrograph and an imaging camera optimized for the ultraviolet. A series of discussions within the ESA Astrophysics Working Group and with NASA narrowed this variety down to the Faint Object Camera (FOC). The decision in favor of the FOC was prompted in part by the fact that it required a detector imaging system that could work in a so-called `photon-counting mode' to exploit the LST's imaging capabilities fully. At that time, Europe had a lead in this area; scientists at the University College London had developed the only photon-counting imaging system in routine use for optical astronomy. As the LST design grew in maturity, the problems of building and launching such a large telescope emerged. Housing the spacecraft inside the Shuttle was proving difficult, and the projected weight was already at the limit of the Shuttle's lift capabilities. In addition, the cost was growing at an alarming rate. In 1974, Congress set two conditions for continuing LST: interContinued national collaboration and lower cost. In 1975, NASA descoped the telescope, page 24 reducing the mirror size to 2.4 m, and changed its name to Space Telescope (ST).



23




FOC from page 23

The NASA/ESA joint working group drafted a Memorandum of Understanding (MOU) for cooperation on ST. The ESA contributions would consist of the FOC, the solar arrays, and a continuing contribution to ST operations. In return, European astronomers would secure 15% of the ST observing time during the planned ten years of operations. In 1976, NASA requested that a team of scientists and engineers be allowed to visit Europe to review its ability to provide the Faint Object Camera. This review team visited eight establishments in Europe (ESA and ESA contractors) over a ten-day period in June 1976 and subsequently submitted a positive report. Later in 1977, Congress gave the approval to build and operate the ST observatory. In March 1977, NASA released the Announcement of Opportunity (AO) for instruments and investigations from the U.S. After ESA's Science Programme Committee (SPC) gave the go-ahead for European participation, the NASA-ESA MOU was finally signed in October 1977. At that time, NASA expected to launch ST in October 1983. Due to a series of financial and technical difficulties on the U.S. side, NASA would put back the launch no fewer than five times, resulting in a three-year delay overall. In January 1983, NASA changed the name to "Hubble Space Telescope" (HST) in honor of Edwin P. Hubble. When the Shuttle `Challenger ' disaster occurred in January 1986, NASA was making final preparations to launch Hubble in October of that year. NASA finally launched Hubble in April 1990.

The FOC Instrument
The FOC was conceived as an imaging instrument with performance complementary to that of the Wide Field and Planetary Camera (WF/PC). Specifically, we designed it to take fullest advantage of the telescope's angular resolution. This called for sensitivity at ultraviolet wavelengths, where the pointspread function (PSF) is smallest, and for a small pixel size, to ensure critical sampling of the PSF according to the Nyquist criterion. The resulting FOC pixel size was 0.022 arcsec, which also was compatible with the expected jitter performance of the spacecraft (~0.007 arcsec). The limited number of these small pixels in our detector--512 x 512 in a typical configuration--meant the field of view (FOV) was rather small. To obviate this limitation to some degree and to increase the overall instrument reliability, we decided to include a second, identical detector in an optical channel with half the magnification--and therefore twice the FOV.

Figure 1. The Faint Object Camera.

The final FOC design consisted of two, independent optical channels, each with its own entrance aperture in the telescope's focal plane. The Hubble telescope has a focal ratio of f/24. The FOC's internal optics increased the effective focal ratio to f/48 in one channel and f/96 or f/288 in the other. Hubble's highest resolution was achieved via the FOC's `f/96' channel. The most favorable setup yielded a 0.022 arcsec pixel size and an 11 x 22 arcsec FOV. To match the PSF at the shortest ultraviolet wavelengths, a three-fold zoom in this channel could increase the focal ratio to f/288, which reduced the pixel size to 0.0072 arcsec and the FOV to just 3.6 x 7.3 arcsec.



24




The f/96 (or f/288) channel offered a choice of 44 different filters, polarizers, and prisms. In addition, this channel had two narrow occulting fingers projecting in from the edge of the field. Either finger could be placed on a bright image to enable the study of a fainter object nearby. The f/48 channel could be used in either of two key ways: direct imaging and long- slit spectroscopy. For direct imaging, observers typically used an angular resolution of 0.043 arcsec over a 22 x 44 arcsec FOV, although several other combinations of resolution and field of view were possible. Fourteen filters and prisms were available in this mode. Nevertheless, most observers used the f/48 channel for long- slit spectroscopy. We added this mode during the design phase when it was clear that neither the Faint Object Spectrograph (FOS) nor the Goddard High Resolution Spectrograph (GHRS) would have this important capability. The FOC detector consisted of an electronic image intensifier, which operated at 35,000 volts, then the highest voltage ever flown in a spacecraft. It intensified the light of individual detected photons more than 100,000 times. The resulting signal was imaged on an electronic camera, and a solid- state memory counted and recorded the location of each detected photon. The limited recording speed restricted the choice of targets to objects sufficiently faint so as not to saturate the electronics. On the other hand, the instrument was sensitive and noise free. In blue light it could detect objects as faint as 30th magnitude in less than 10 hours of observation.

The Science Team
To advise ESA and to lead the scientific design of the FOC, we conducted a competition in Europe to select a science team, which reported to me as the ESA Project Scientist. Henk van der Hulst (Leiden), who chaired the team, was one of ESA's founding fathers as well as a world-class astronomer. The other members of our team were: Rudi Albrecht (Vienna), Cesare Barbieri (Padova), Alec Boksenberg (University College London), Jean-Michel Deharveng (Marseille), Mike Disney (Cardiff), Theo Kamperman (Utrecht), Craig Mackay (Cambridge), Gerdt Weigelt (NÝrnberg), and Ray Wilson (European Southern Observatory, ESO). In addition, NASA selected two U.S. members of this core FOC team, Ivan King (Berkeley) and Phil Crane (ESO). In exchange, both van der Hulst and I were members of the key NASA advisory committee, the HST Science Working Group, from its beginning in 1976 to its disbanding after the first servicing mission.

Figure 2. The FOC Investigation Definition Team. First row, left to right: Blades, Albrecht, Barbieri, Boksenberg, Mackay, Macchetto, King, Nota, and Paresce. Second row: Jakobsen, Disney, Deharveng, Sparks, Jedrzejewski, Crane, and Greenfield.

When I moved to the Institute in 1983, I hired several ESA staff members to work specifically on the FOC. They became part of the science team and participated fully in the discussions and planning for the Guaranteed Time Observer (GTO) scientific observations. They included Francesco Paresce, who later moved to ESO, and others still at the Institute: Chris Blades, Antonella Nota, and Bill Sparks. We selected a new Project Scientist, Peter Jakobsen, to keep close contact with the ESA Project team in ESTEC, ESA's technical center in the Netherlands. He also became a full member of our team. Many others worked with and for the FOC team at the Institute, notably Robert Jedrzejewski and Perry Greenfield. The full list would probably include a large fraction of the Institute staff today. I want to extend my thanks to all those who participated Continued and supported the work of the FOC throughout its many years, on the ground page 26 before launch and later in orbit.



25




FOC from page 25

Industry and Project Team
Top industrial teams in Europe built the FOC. British Aerospace built the photon-counting detector. Matra built the optics. Dornier Systems was the prime contractor for the whole assembly. Many other European firms contributed important subsystems. Of the many contractor personnel, I want to mention that Manfred Miebach, who worked for the FOC industrial team at Dornier before becoming an ESA staff member at the Institute, has been associated with the FOC and Hubble almost as long as I have! The ESA project team was led by Jan Burger and Robin Laurance, who were always ready to listen and respond to the requirements of the science team, while keeping the cost and schedule under control. Taking inflation into account, ESA built the FOC within 10% of the originally approved cost.

Spherical Aberration
Soon after launch, we realized the Hubble mirror was flawed with spherical aberration. We were devastated. Whereas we had designed the FOC to take advantage of the full resolution of the telescope, spherical aberration meant that about 80% of the scientific objectives of the FOC were not achievable. To compensate, we redesigned our science program to observe objects with less structure and contrast or that were not located in crowded fields. Nevertheless, it was clear that the FOC had suffered a crippling blow. We participated in the process that eventually led to the design of COSTAR and worked with the COSTAR team to optimize the design for the FOC, which was by far the most demanding instrument in terms of accuracy and alignment. While COSTAR recovered most of the capabilities lost to spherical aberration, it came at a cost. The spatial resolutions of both the f/48 and f/96 channels were increased by the re-imaging optics and the already small FOVs were correspondingly reduced. This penalty was particularly hard on the long- slit spectrograph mode, where the minute 0.1 arcsec original aperture became a 0.07 arcsec in size, which threw away most of the light of a point source without a real improvement in spatial resolution. Furthermore, it proved to be impossible to correct the optical beam for the f/288 mode, which had to be abandoned, thus sacrificing the highest resolution observations that we had planned. Nevertheless, we were very happy when COSTAR was deployed and worked perfectly! It meant that we--and the community--could recover a substantial fraction of our science.

FOC Science
Without attempting a general review of FOC science, it is safe to say that three observational capabilities contributed most strongly to its scientific success: the high spatial resolution, the polarizers, and the long- slit spectrograph.

Figure 3. FOC observations of the jet in the galaxy M87 showing motions at apparent velocities several times the speed of light. The actual velocities of these features, presumably caused by energetic electrons, is very close to the speed of light indicating an extremely energetic source.



26




Even with spherical aberration, FOC observations fully resolved Pluto and Charon. COSTAR enabled detailed observations of their surfaces. The FOC measured the lower main sequence of Galactic globular clusters, such as 47 Tuc and Omega Cen. These observations require both high spatial resolution, to isolate individual stars, and high photometric precision. Similar observations were conducted on globular clusters in nearby galaxies, such as the Magellanic Clouds and Andromeda. At greater distances, we used the FOC to study the jet of M87 in detail, using both imaging and polarimetric techniques. Multiyear observations allowed us to detect and measure the proper motion of features in the jet, some of which appear to move at several times the speed of light. With these results, we developed a detailed physical model for the jet. Our team invested a great deal of time investigating all known optical jets and discovered a number of new ones. These studies increased our general understanding of physical processes in active galaxies. FOC studies of the central regions of active galactic nuclei led to precise measurements of the masses of the central black holes. Using the long- slit spectrograph, we made a precision measurement of the mass of the black hole in the center of M87--3 billion solar masses. FOC polarization observations of active galaxies, such as NGC 4151, NGC 1068, and Cen A , revealed features of their central regions not detectable with ground-based telescopes. These results have profoundly modified our understanding of such galaxies.

The Future
In early March 2002, as astronauts remove the FOC from Hubble to make space for the Advanced Camera for Surveys, I reflect on the long collaboration in ultraviolet/optical astronomy between NASA and ESA , which started with IUE and has had Hubble as its centerpiece for almost 20 years. This record of collaboration is a model for the future. Having had the privilege to be very closely involved with both IUE and Hubble, I look forward to NASA and ESA forming a similar close and fruitful collaboration on the Next Generation Space Telescope and perhaps in a future large optical/ultraviolet space telescope project.

Contact STScI:
website http://www.stsci.edu SThe Institute'savailable atis:help@stsci.edu or 800-544-8125. Assistance is TInternational callers can use 1-410-338-1082. SFor current Hubble users, program information is available I at: http://presto.stsci.edu/public/propinfo.html. The current members of the Space Telescope Users Committee (STUC) are: Prof. George Miley, STUC Chair, Sterrewacht Leiden, miley@strw.leidenuniv.nl Prof. David Axon, University of Hertfordshire Prof. Marc Davis, University Of California Dr. James Dunlop, Royal Observatory Edinburgh, Dr. Martin Elvis, Harvard-Smithsonian CfA Dr. Debra Elmegreen,Vassar College Prof. Holland Ford, Johns Hopkins University Dr. Suzanne Hawley, University of Washington Dr. Chris Impey, University of Arizona Dr. John Kormendy, University of Texas at Austin Dr. Karen Meech, Institute for Astronomy Honolulu Dr. Peter Nugent, Lawrence Berkeley Laboratory Dr. David Sanders, Institute for Astronomy Honolulu Dr. Karl Stapelfeldt, JPL Dr. John Stocke, University of Colorado Dr. Lisa Storrie-Lombardi, Caltech The STScI Newsletter is edited by Robert Brown, rbrown@stsci.edu, who invites comments and suggestions. Technical: Christian Lallo Editorial: Karen Falk Design: Christine Klicka To record a change of address or to request receipt of the Newsletter, please contact Nancy Fulton (fulton@stsci.edu).

ST- E CF Newsletter

T

he Space Telescope - European Coordinating Facility publishes a newsletter which, although aimed principally at European Space Te lescope users, contains articles of general interest to the HST community. If you wish to be included in the mailing list, please contact the editor and state your affiliation and specific involvement in the Space Te lescope Project. Richard Hook (Editor) Space Telescope European Coordinating Facility Karl Schwarzschild Str. 2 D-85748 Garching bei MÝnchen Germany E-Mail: rhook@eso.org



27




Contents:
Hubble Treasury Program Ages from Near-UV Spectra of Globulars & Galaxies . . . .1 The ACS Contribution to GOODS . . . . . . . . . . . . . . . . . . .1 The Hubble Treasury Program on Eta Carinae . . . . . . . . . .1 Institute News Director 's Perspective . . . . . . . . . . . . . . Hubble Calibration Workshop . . . . . . . . . Progress on WFC3 . . . . . . . . . . . . . . . . Cosmic Origins Spectrograph . . . . . . . . . New Tool for Preparing Phase 1 Proposals Specview: A New Tool for Visualizing and Modeling Spectra . . . . . . NGST Transitions to a New Phase . . . . . MAST News . . . . . . . . . . . . . . . . . . . . .

Calendar
May Symposium, " The Astrophysics of Life," at STScI Institute Interim Visiting Committee meeting at STScI End of Servicing Mission Orbital Verification Space Telescope Institute Council meeting at ESTEC Start of Cycle 11 May May May June July 6-9 9-10 13-14 1

. . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. . . . .

. .2 . .5 . .9 .11 .12

. . . . . . . . . . . . .12 . . . . . . . . . . . . .13 . . . . . . . . . . . . .15

Institute Science Astrometric Planet Detection using Hubble's FGS . . . . . . .17 A Surprise within a Surprise: Hot Gas around HighVelocity Clouds and a Corona around the Milky Way . . . . .19 Institute People Interview: Megan Donahue . . . . . . . . . . . . . . . . . . . . . . .21 Hubble History The FOC Story . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23 Contact STScI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .27 Calendar . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28

AC SP E TELESCOPE SCIENCE INSTITUTE

NON PROFIT U.S. POSTAGE

PAID
PERMIT NO. 8928 BALTIMORE, MD

3700 San Martin Drive, Baltimore, Maryland 21218

STScI N-02-02