7341( 4) - 02/18/99 14:48 - [ 1] PROPOSAL FOR HUBBLE SPACE TELESCOPE OBSERVATIONS ST ScI Use Only ID: 7341 Version: 4 Check-in Date: 18-Feb-1999 14:48:05 1.Proposal Title: Absolute Proper Motions of Nearby Dwarf Spheroidal Galaxies ------------------------------------------------------------------------------------ 2. Proposal For 3. Cycle GO 7 ------------------------------------------------------------------------------------ 4. Investigators Contact? PI: Edward Olszewski University of Arizona CoI: Hugh Harris US Naval Observatory N CoI: Mario Mateo Univ of Michigan N CoI: Dante Minniti Lawrence Livermore National Labs N CoI: David Monet US Naval Observatory N CoI: Heather Morrison Case Western Reserve Univ N CoI: Carlton Pryor Rutgers University N CoI: Chris Tinney Anglo-Australian Observatory N ------------------------------------------------------------------------------------ 5. Abstract We propose to measure precise absolute proper motions for four dwarf spheroidal satellites of the Milky Way using spectroscopically-confirmed background QSOs to define a zero- velocity reference frame. Two epochs separated by 2 yrs will yield systemic tangential velocities of UMi, Car, Scl, (and For) to +/- 78 kms\ (+/- 130 kms). These are worst-case velocity precisions and they are likely to be 2--4* smaller. Our long-term goal is to reduce them by an additional factor of several by obtaining data over the lifetime of WFPC2. With 2-3 QSOs per galaxy, we will still be confident of our motions with only 2 epochs. We will test whether the halo contains a small number of massive streams containing several dwarf galaxies, or whether the individual halo dwarfs are traveling along independent orbits. HST is essential to achieving the high precisions needed to conclusively compare the projected orbital motions of the individual galaxies; even with our conservative uncertainties, we are competitive with the best ground-based efforts with only a 2 year baseline. We will also use our results to improve our estimate of the mass of the Galaxy interior to ~ 100 kpc. We believe that our project will show that astrometry has been a much ignored resource and power of HST. If HST performs as well as we suspect it can, it will be possible to measure the internal motions of stars in the dwarf spheroidals and the proper motions of all of the Local Group members over a timespan of 5 -- 10 years. ------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 2] Observations Description ------------------------ We address in some detail the necessary requirements to achieve precision astrometry with HST. We will show that each potential stumbling block has been dealt with at STSCI. We also point out that Meylan Etal\ (Cycle 5 program 5912) and Becklin (NICMOS GTO program NC01), among others, are doing similar astrometry with HST. We have had extensive conversations about the technical issues with R. Gilliland of STSCI, who finds ``no show stoppers.'' CCD astrometry from the ground has shown that three fundamental issues must be addressed to achieve high-precision relative astrometry. If these issues can be successfully addressed, it has been shown that relative positions to better than 1 percent can easily be obtained, provided that the required photon counting limits (about 10, 000 photons) can be reached. Proper motions and parallaxes done with CCDs at the US Naval Observatory show that centroiding to better than 0.01 pixels is possible in general from the ground. Results at the Naval Observatory also show that the principle limitation is the atmosphere, not the CCD. o The first requirement is that the astrometry must be carried out in a differential mode. Traditional photographic astrometry has always been limited by the need for a detailed knowledge of field distortions - if your field is not flat, and you cannot guarantee a priori where your astrometric stars will be placed in that field, it is essential that it be possible to flatten the field to high precision. On the ground, CCDs obviate this problem by allowing one to always place the target objects back in the same place on the CCD to within a few pixels. Under these conditions, field distortions cancel out, and high relative precisions can be reached. For HST observations we need to place a fiducial object (the background QSOs) back on the same place of the PC1 CCD at each epoch to within a few pixels. Moreover, all observations must be made with an identical roll angle. This angle is not pre-determined, but will become fixed once the first-epoch observations are made. Although the associated overheads are high, this requirement is very stringent. The field distortion of the PC1 is not known to sufficient precision for 0.4 milli-arcsec epoch^-1 positions to be determined unless true differential observations are obtained. R. Gilliland says that we can expect +/-2 pixels RMS positioning with roll angle positioning to a few arcmin-- this is adequate for our purposes. o The second requirement is that the PSF must be well sampled. It has become apparent in recent years that CCDs do not have uniform sensitivity across a pixel. In the worst cases, the sensitivity may vary by up to 5 percent across a single pixel. This has serious implications for a program designed to measure positions to an accuracy of better than 1 percent of the size of a pixel with critically- sampled images. Because HST is undersampled even with the Planetary Camera, we must deal with this problem. Meylan Etal\ (GO 5912) have designed a solution that requires 10 individual exposures (of 160 sec each) per epoch. The telescope will be offset according to a pattern involving 0.5 and 0.7 pixel offsets. A similar strategy will allow us to effectively oversample the PC images as required to carry out high-precision astrometry. Gilliland was able to achieve 0.024 pixel centroiding with longer (more cosmic rays) undithered data. We will remove cosmic rays by taking two exposures at each spatial position. o The last requirement is for a distant reference frame. This has been solved in the present case by centering the PC on previously-identified background quasars. Based on our experience with other high- latitude HST observations, it is also very likely that we shall have 3-5 galaxies in the PC field with sharp cores. These too can be used to help define the zero point. Note that the exposure times are designed to obtain 3000-5000 counts (data numbers) per main-sequence turnoff star in the target galaxies. This will ensure that the quasar images will have the > 10^5 counts necessary to obtain the requisite centroid position. All of the reference QSOs are located within about one core radius of their respective foreground dSph galaxy (typically < 15'). We have scaled counts of ground-based and HST images to determine that we shall observe 50-300 stars brighter than 0.5 mag of the turnoff points of Scl, UMi and Car within the field of view of the PC (the worst case is Car). Such stars will all have >= 3000 total counts. For Fornax, we shall obtain only about 1000 counts 0.5 mag below the turnoff point, but for more than 1000 stars. Since the ensemble of dSph stars will be used to define the mean proper motion of the galaxies, we shall readily achieve sufficiently high total counts to obtain precise mean positions for the stars in each field at each epoch. vspace*0.3 in To summarize, if we obtain the same roll angle for each epoch for a given field, attain adequate S/N in the QSO and the sum of the stars in the galaxies, and we obtain our data using the dithering pattern described above, we should easily achieve precisions of 100th of a pixel. This translates into a 1-sigma error of 0.23 mas/yr in the resulting proper motions if we make two observations separated by two years. We stress here that we may also achieve 2- 4 times better precision based on our experience with ground-based astrometric measurements. We have requested one additional observation for each UMi field to empirically measure our measurement precision in Cycle 7. These two orbits should involve the full dithering pattern described above and be obtained at the same roll angle. The QSOs will be offset by about 10 pixels (~ 0.5 arcsec) to simulate a new visit at a much later epoch and to test whether we differentially remove the distortion pattern to the expected level. vspace*0.3 in To minimize telescope time, we shall observe with only one filter, F606W. This filter provides the best throughput of any HST filter (apart from some blockers and polarizing filters) and so provides the best S/N for a given exposure time. (While STIS does have much wider filters and a larger PC-like field, we are concerned that the combination of the mirror-repositioning problems and the lack of experience with STIS will hurt our project. We also are not photon starved and are limited by saturating the quasars.) The bandpass is also well-matched to the spectral distribution of the turnoff stars in the galaxies. The 160-second individual exposure times do not saturate any of the QSOs yet provide adequate counts for the turnoff once the multiple exposures are combined. The timeline is 9 minutes for guidestar, 2 min for ``alignment'', 1 minute for filter change, 1 minute readout for each of 8 160 sec exposures, 3 minutes total time per exposure (160 sec + 14 sec = 3 minutes in HST land), and 1 minute times 4 dithers= 51 minutes. (We take double exposures, effectively CR -splitting, at each spatial position.) Orbit 2 allows 9 160 sec exposures. Carina can have 9 such exposures in the first orbit, too. Ursa Minor, in the CVZ, can have 17 in the first orbit, and 19 in the second orbit. More QSO's are known behind our target galaxies than we propose to observe with HST. We have chosen to observe only two fields in Scl, UMi, and Car to achieve some gain in precision at small cost in telescope time. In addition, one of the Scl fields has two QSOs in it. We have also requested three fields for Fornax to mitigate the larger distance of that galaxy (130 kpc) compared to the other three (70 kpc). Real Time Justification ----------------------- The two Ursa Minor fields are in the CVZ. We will add more observations to our dithering patterns here thus improving our ability to derive coordinates of stars. Once initial observations are made for each field, we require that the roll angles for each subsequent visit be identical to the first-epoch roll angle. We also request a special dithering pattern during each visit consisting of many separate observations per field (more in UMi because of CVZ). The background quasars were discovered previously using radio, X-ray and optical techniques. Two QSO's are known near the center of UMi (Olszewski), while seven are known behind Car, and four each behind Scl and For (Tinney). The quasars have been confirmed spectroscopically; their redshifts range from 0.3 - 3.1. No future ground-based observations are needed in support of observations of these four galaxies. However, future searches may be carried out to search for quasars near other dSph galaxies or outer-halo clusters for future HST astrometric observations. Calibration Justification ------------------------- Additional Comments ------------------- ------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 3] Data Distribution Paper Products: Media: DAT Blocking Factor: 10 Ship To: PI_Address Ship Via: Email: ------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 4] TARGET LIST a) Fixed Targets ------------------------------------------------------------------------------------------------------------------------------------ Tar| Target | Target | Target |Coord | Radial | Flux data No | Name | Description | Position |Eqnx | Vel. | ------------------------------------------------------------------------------------------------------------------------------------ 1 QSO-UMI- Galaxy, Dwarf RA=15H 08M 40.41S +/- 0.1S, J2000 V = 18.6 1508+67 Spheroidal, QSO, Quasar DEC=+67D 17' 47.5" +/- 1.0" B-V = 0.2 2 FOR-QSO-0240 Galaxy, Dwarf RA=02H 40M 07.7S +/- 0.1S, DEC=- J2000 V = 19.1 -34 Spheroidal, QSO, Quasar 34D 34' 20.0" +/- 1.0" B-V = 0.2 Comments: two QSOs 5 arcsec apart 3 CAR-QSO-0641 Galaxy, Dwarf RA=06H 41M 51.6S +/- 0.1S, DEC=- J2000 V = 20.8 -51 Spheroidal, QSO, Quasar 50D 57' 13" +/- 1.0" B-V = 0.2 4 SCL-QSO-0100 Galaxy, Dwarf RA=01H 00M 25.3S +/- 0.1S, DEC=- J2000 V = 20.4 -3341 Spheroidal, QSO, Quasar 33D 41' 07.0" +/- 1.0" B-V = 0.2 7341( 4) - 02/18/99 14:48 - [ 5] Visit: 01 Visit Priority: Visit Requirements: CVZ On Hold Comments: Additional Comments: Exposures ------------------------------------------------------------------------------------------------------------------------------------ Exposure| Target |Instr | Oper. | Aper |Spectral|Central| Optional |Num| Time | Special Number | Name |Config| Mode |or FOV |Element |Waveln.| Parameters |Exp| | Requirements ------------------------------------------------------------------------------------------------------------------------------------ 1 QSO-UMI-150 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO,SCAN-READ=ALL 3 160. S SPATIAL SCAN 1 1 8+67 ------------------------------------------------------------------------------------------------------------------------------------ 3 QSO-UMI-150 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO,SCAN-READ=ALL 2 160. S SPATIAL SCAN 1 3 ; POS 8+67 TARG +0.231, -0.231 ------------------------------------------------------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 6] Visit: 02 Visit Priority: Visit Requirements: On Hold Comments: Additional Comments: Exposures ------------------------------------------------------------------------------------------------------------------------------------ Exposure| Target |Instr | Oper. | Aper |Spectral|Central| Optional |Num| Time | Special Number | Name |Config| Mode |or FOV |Element |Waveln.| Parameters |Exp| | Requirements ------------------------------------------------------------------------------------------------------------------------------------ 6 FOR-QSO-024 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO,SCAN-READ=ALL 2 160. S SPATIAL SCAN 1 0-34 ------------------------------------------------------------------------------------------------------------------------------------ 7 FOR-QSO-024 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO 2 160. S 0-34 ------------------------------------------------------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 7] Visit: 03 Visit Priority: Visit Requirements: On Hold Comments: Additional Comments: Exposures ------------------------------------------------------------------------------------------------------------------------------------ Exposure| Target |Instr | Oper. | Aper |Spectral|Central| Optional |Num| Time | Special Number | Name |Config| Mode |or FOV |Element |Waveln.| Parameters |Exp| | Requirements ------------------------------------------------------------------------------------------------------------------------------------ 9 CAR-QSO-064 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO,SCAN-READ=ALL 2 160. S SPATIAL SCAN 1 1-51 ------------------------------------------------------------------------------------------------------------------------------------ 10 CAR-QSO-064 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO 2 160. S 1-51 ------------------------------------------------------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 8] Visit: 04 Visit Priority: Visit Requirements: On Hold Comments: Additional Comments: Exposures ------------------------------------------------------------------------------------------------------------------------------------ Exposure| Target |Instr | Oper. | Aper |Spectral|Central| Optional |Num| Time | Special Number | Name |Config| Mode |or FOV |Element |Waveln.| Parameters |Exp| | Requirements ------------------------------------------------------------------------------------------------------------------------------------ 12 SCL-QSO-010 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO,SCAN-READ=ALL 2 160. S SPATIAL SCAN 1 0-3341 ------------------------------------------------------------------------------------------------------------------------------------ 13 SCL-QSO-010 WFPC2 IMAGE PC1 F606W CR-SPLIT=NO 2 160. S 0-3341 ------------------------------------------------------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 9] Scan Paramters Form ------------------------------------------------------------------------------------------------------------------------------------ Data ID: 1 Exposure Logsheet lines: 01.001,01.003,02.006, FGS Scan: D Dwell Only: dwell points/line: 2 Seconds per dwell: 1.00 Scan width (arc-secs): 0.0805 Scan length (arc-secs): 0.0683 Angle between sides (degrees) 135.00 Number of lines: 4 Scan rate (arc-sec/sec): 0.0000 PA of first scan line (degrees) 90.000 Scan frame (CEL or S/C): S/C Length Offset (arc-sec): 0.0403 Width Offset (arc-sec): 0.0342 ------------------------------------------------------------------------------------------------------------------------------------ 7341( 4) - 02/18/99 14:48 - [ 10] Summary Form for Proposal 7341 Item Used in this proposal ------------------------------------------------------------------------------------------------------------------------------------ Apertures PC1 ------------------------------------------------------------------------------------------------------------------------------------ Configurations WFPC2 ------------------------------------------------------------------------------------------------------------------------------------ Opmodes IMAGE ------------------------------------------------------------------------------------------------------------------------------------ Optional Parameters CR-SPLIT=NO SCAN-READ=ALL ------------------------------------------------------------------------------------------------------------------------------------ Proposal Category GO ------------------------------------------------------------------------------------------------------------------------------------ Special Requirements CVZ SPATIAL SCAN 1 1 SPATIAL SCAN 1 3 ; POS TARG +0.231, -0.231 SPATIAL SCAN 1 ------------------------------------------------------------------------------------------------------------------------------------ Spectral Elements F606W ------------------------------------------------------------------------------------------------------------------------------------ Target Names QSO-UMI-1508+67 FOR-QSO-0240-34 CAR-QSO-0641-51 SCL-QSO-0100-3341 ------------------------------------------------------------------------------------------------------------------------------------