Page 1 PROPOSAL FOR HUBBLE SPACE TELESCOPE OBSERVATIONS ST ScI Use Only ID 4144c Report Date: 09-May-96:19:16 Version: ********** Check-in Date: ********** 1.Proposal Title: LINKING HIPPARCOS TO THE EXTRAGALACTIC REFERENCE FRAME PART 5 OF 6, NEWC, CYCLE 3,CONTINUATION OF 2565 ------------------------------------------------------------------------------------ 2. Scientific Category 3. Proposal For 4. Proposal Type 5. Continuation ID STELLAR POPULATIONS GO 2565 Sub Category ASTROMETRY ------------------------------------------------------------------------------------ 6. Principal Investigator Institution Country Telephone Dr Paul D Hemenway UNIVERSITY OF TEXAS AT AUSTI USA (512)471-3431 RSSCI ------------------------------------------------------------------------------------ 7. Abstract Determination of a non-rotating Reference Frame is crucial to progress in many areas, including: Galactic motions, local (Oort's A and B) and global (R0) parameters derived from them, solar system motion discrepancies (Planet X); and in conjunction with the VLBI radio reference frame, the registration of radio and optical images at an accuracy well below the resolution limit of HST images (0.06 arcsec). The goal of the Program is to tie the HIPPARCOS and Extra- galactic Reference Frames together at the 0.0005 arcsec and 0.0005 arcsec/year level. The HST data will allow a deter- mination of the brightness distribution in the stellar and extragalactic objects observed and time dependent changes therein at the 0.001 arcsec/year level. The Program requires targets distributed over the whole sky to define a rigid Reference Frame. GTO observations will provide initial first epoch data and preliminary proper motions. The observations will consist of relative positions of Extra- galactic objects (EGOs) and HIPPARCOS stars, measured with the FGSs, or with the FGSs and PC together in "transit circle mode". The combination of HST and HIPPARCOS observations will provide ------------------------------------------------------------------------------------ 8. Scientific Key Words: REFERENCE FRAMES, HIPPARCOS, QUASARS ------------------------------------------------------------------------------------ 9. Est obs time (hours) pri: 80.00 par: 25.00 10. Num targs pri: 175 par: 30 ------------------------------------------------------------------------------------ 11. Instruments requested: FGS WF/PC ------------------------------------------------------------------------------------ ------------------------------------------------------------------------------------ Page 2 I. GENERAL FORM Proposal 4144c PI: Dr Paul D Hemenway Proposal Title: LINKING HIPPARCOS TO THE EXTRAGALACTIC REFERENCE FRAME PART 5 OF 6, NEWC, CYCLE 3,CONTINUATION OF 2565 ------------------------------------------------------------------------------------ 1. Proposers: Proposers Institution Country ESA ------------------------------------------------------------------------------------ Dr Paul D Hemenway UNIVERSITY OF TEXAS AT AUSTIN USA Dr Noel Argue THE OBSERVATORIES ENGLAND Dr Raynor L Duncombe UNIVERSITY OF TEXAS AT AUSTIN USA Dr James Hughes U.S. NAVAL OBSERVATORY USA Dr David Jauncey C.S.I.R.O. AUSTRALIA Dr Kenneth Johnston U.S. NAVAL RESEARCH LAB USA Dr Jean Kovalevsky C.E.R.G.A. FRANCE Dr Jean Lestrade BUREAU DE LONGITUDE FRANCE Dr Michael Perryman E.S.T.E.C. HOLLAND Dr Robert Preston JET PROPULSION LAB USA Dr Catherine Turon OBSERVATOIRE DE MEUDON FRANCE Dr Christian Devegt HAMBURGER STERNWARTE FGR Dr Hans Walter ANSTRONOMISCHE RECHENINSTITUT FGR Dr Graeme White C.S.I.R.O. AUSTRALIA Dr Jerome Kristian CALTECH USA Prof. Byron Tapley UNIV OF TEXAS AT AUSTIN USA Page 3 ------------------------------------------------------------------------------------ 2. Scientific Justification. The astronomical uses of an accurate coordinate reference frame are manifest. The reference frame provides the kinematic basis for all dynamical studies of celestial objects. The HIPPARCOS satellite will determine positions, proper motions, and parallaxes for about 100000 stars evenly distributed over the sky (Kovalevsky, _Space Science Rev._, 39,41,1984). A residual time dependent rotation of the reference frame would introduce a systematic effect in the system of proper motions that biases all the astronomical uses that can be made of the system in kinematical and dynamical studies of the Galaxy. (Galactic differential rotation and its variation with distance, asymmetry of motions with respect to the galactic plane, kinematics of halo stars and other stellar populations, group motions, and determining the birth places of stars, are a few examples of galactic studies which are adversely affected by a residual rotation of the reference frame.) This proposal is to reduce this undesirable bias by determining the rotation of the HIPPARCOS frame with respect to extragalactic objects. The method is to measure the proper motion of some stars in the HIPPARCOS system with respect to extragalactic objects (EGOs) using the Hubble Space Telescope (HST). HST can provide the information needed to link the HIPPARCOS proper motion system to the fixed reference system defined by the EGOs. The object of the overall program is to obtain the observations necessary to determine this link to an accuracy of the order of the systematic regional errors that may exist in the final HIPPARCOS catalogue. The nominal HIPPARCOS accuracy for stars brighter than 10.5 is expected to be 0.002 arcsec in position and parallax, and 0.002 arcsec/year in proper motion. HIPPARCOS will observe objects between 10.5 and 12 magnitude with reduced accuracy. The basic HIPPARCOS data will be accurate large angle separations of stars (chords on the celestial sphere). The system solution will yield a "solid body" celestial coordinate system an order of magnitude more accurate systematically than the individual random errors of the individual stars. Caused by unknown systematic biases of instrumental origin, they have not been properly assessed, but they should not exceed 0.1 milliarcsec in position, or 0.1 milliarcsec/year in proper motion. However, the nature of the HIPPARCOS reference system is such that it has a double indeterminacy: an arbitrary "solid body" rotation proportional to time in its proper motion system, and an arbitrary solid body rotational offset from any "standard" (e.g. inertial or quasi-inertial) coordinate system, in position. R. L. Duncombe and P. D. Hemenway plan to use about 40 hours of GTO time to observe HIPPARCOS stars near extragalactic objects. The stars have been screened from about 300 stars by a large program of speckle interferometry specifically for this purpose. (C.f. Hemenway, et al., 1985 ESA special Publication SP-234, p.281; Argue, et al., _Mon. Not. RAS_, 206,669, and 216,447.) Drs. Franz and McAlister have done the northern hemisphere speckle observations (private communcation to PDH). Simulations by Froeschle and Kovalevsky (_Astron. Astrophys._, 116, 89, 1982) show that for observations evenly distribued over the sky, the elements of the rotation matrix would be obtained with an accuracy of 0.003 arcsec/year rms for 40 pairs and 0.0023 arcsec/year for 70 pairs. We have identified 175 stars around 91 EGOs that are observable astrometrically with HST. This distribution should allow the determination of the rotation matrix to an accuracy of the order of 0.0016 arcsec/year. However, these figures represent measurements over the lifetime of the HIPPARCOS mission, 2.5 years and the subsequent accuracy for the tie to EGOs at the end of the mission. A longer time base will reduce the error propor- tionally to the time base for the next several years after the HIPPARCOS mission is completed. The determination of the components of the rotation matrices depends critically on the distribution being uniform over the sphere. A loss of the sources south of -30 degrees declination increases the rms of the solution by a factor of 1.4, for example. Therefore, the first priority is to make sure that the EGOs that are observed are well distributed over the celestial sphere. Although this result is very accurate, it does not reach the expected accuracy of the HIPPARCOS system by a factor of 2 or 3. By reobserving all the HIPPARCOS star-EGO pairs over the course of the HST lifetime, the error of the individual motions is reduced proportionally to the time, and the overall solution should reach the 0.0008 arcsec/year level. If, in addition, about 15 radio stars are observed by VLBI giving a precision of proper motion of about 0.002 arcsec/year, then the expected global accuracy will become: a) 0.0025 arcsec/year if this proposal is not accepted, b) 0.0016 arcsec/yr if the GTO plus this proposal is accepted, c) 0.0012 arcsec/year if the full combined list is observed, and d) 0.0005 arcsec/yr if all observations of all types over the HST lifetime are observed. Some points which need stressing: a) Two HST time scales are important, one common to the HIPPARCOS epoch should start AS SOON AS POSSIBLE AFTER LAUNCH to reach the full potential of the tie between the VLBI frame and the HIPPARCOS frame, which requires observations of the same objects at the same epochs. The early epoch observations will also give the longest time base for the proper motion determinations, b) HIPPARCOS can only observe down to 12th magnitude with reduced accuracy. It cannot observe the Exragalactic objects directly, c) The program and object selection have been the subjects of intensive efforts by many people over the last few years, including the Working Group on Radio/Optical Identifications of IAU Commission 24, Subgroup 2130 of the HIPPARCOS Input Catalogue Consortium, a working group of one of the HIPPARCOS reduction consortia (F.A.S.T.), the speckle observers headed by Franz and McAlister in the north and Argue and Morgan in the south, and the southern hemisphere radio observers headed by Jauncey and White, who have identified the southern sources in this proposal through a major effort in the last two years, d) Jauncey, White, and Preston are working to make all the radio sources in this proposal part of the VLBI reference frame, including establishing the Australian VLBI capability, e) A set of 414 candidate stars near EGOs was included in the HIPPARCOS Input Catalogue in 1982 to make sure that the stars would be observed by HIPPARCOS. The speckle programs comprised these stars and the stars near the southern EGOs subsequently indentified by Jauncey and White. The objectives of this program are: 1) Provide an accurate tie of the HIPPARCOS Coordinate Reference System (positions and proper motions) to the best available extragalactic (e.g. VLBI) system, at the epoch of HIPPARCOS, 2) Tie the HIPPARCOS Reference System of proper motions to a (non-rotating) extragalactic system of optical objects, 3) Thereby establishing the connection between the Radio and Optical reference frames to the millarcsecond level (allowing the registration of HST images to radio images (VLA, VLBI, VLBA) to an accuracy well below the reolution limit of HST), anywhere in the sky, 4) Obtain single-epoch information on the brightness distri- bution of EGOs and stars, and 5) Obtain time dependent studies of the resolved and unresolved structure of (4) at the 0.001 arcsec/year level of accuracy. We have a brief statement from the HIPPARCOS Science Team supporting the importance of this work to the HIPPARCOS project, and its usefulness for both HIPPARCOS and HST astrometric data. The stement is included as an appendix to the proposal. APPENDIX: The following statement was communicated at the end of December, 1985: The eleven members of the HIPPARCOS Science Team, the scientific representatives who advise ESA directly on the scientific conduct, planning and organisation of the HIPPARCOS project, have expressed their full support for this collaborative programme. For the reasons set out in the proposal, the HIPPARCOS Science Team has confirmed the importance of the propsed observations for the interpretation and exploitation of both HIPPARCOS and Space Telescope astrometric data. Professor Adrian Blaauw, chairman of the ESA HIPPARCOS proposal selection committee; and Professor Walter Fricke*, chairman of the HIPPARCOS Input Catalogue Consortium steering committee, have also given their support to the present proposal. *Deceased 28 March 1988. Page 4 ------------------------------------------------------------------------------------ 3. Description of proposed observations. The basic datum required is the relative separation angle of an EGO and a HIPPARCOS star. One would hope to rely on the knowledge of the FGS calibrations to simply measure one object and then the other. However, because of thermal and mechanical instabilities during "exposures", initially we will follow the GTO observing strategy of measuring two objects several times with some intermediate background stars as a stability check. Should the FGS measurements prove to be as stable as expected, then the number of check stars and the frequency of measurement may be reduced by a large factor (up to about 5). Since a single measurement of a single pair is expected to average about 20 minutes, overall slewing time is expected to be a major factor in actual time consumption. Based on OV/SV/GTO observations, we will learn how to optimize the use of HST for the objectives of this program. For EGOs brighter than 17.0, the observations will consist of obtaining FGS reading on: 1) the EGO, 2) the HIPPARCOS star, and 3-5) 3 background stars (when available) in the pickle. The following sequence will be observed: 1,3-5,2,1,2,5-3,1. When a star is brighter than V=9.0, the neutral density filter will be inserted. Whenever possible, another HIPPARCOS star will be included as one of the "background" stars. For EGOs fainter than 16.5, a Planetary Camera frame (P8) will be obtained for the EGO while the guiding FGS data are recorded. A PC frame will then be obtained for the HIPPARCOS star using a narrow filter; recording the FGS data for the guide stars. THE SAME GUIDE STARS MUST BE USED FOR EACH TARGET. The posi- tional objects' centroids give the separation. A third exposure will be taken through a red filter. The color data will be analysed for astrophysical information about the objects observed. ------------------------------------------------------------------------------------ 4. Justification of need for HST observations. Simulations have shown that the expected regional systematic effects within the HIPPARCOS catalogue should be ~0.0001 arcsec. The expected individual HIPPARCOS position and motion errors are expected to be 0.002 arcsec and 0.002 arcsec/year. In tying the HIPPARCOS Reference Frame to EGOs, the error of the frame will depend on the errors of observations of the individual objects used. Two methods of high accuracy observations have been proposed. First is the VLBI observation of radio stars in HIPPARCOS directly with respect to EGOs. VLBI sensitivity limits the stars to a few RS CVn stars, with an accuracy of 0.002 arcsec. Some of these stars show radio- optical offsets at the 0.002 arcsec level. The second method is the one proposed here. The FGS measure- ments have an accuracy commensurate with both HIPPARCOS and radio measurements, again 0.002 arcsec. Other proposed methods rely on photographic astrometry of thousands of stars, galaxies, and QSOs, with measuring accuracies in the 0.1 to 0.01 arcsec range. These methods suffer from large systematic effects. No information about the optical structure of the EGOs or the stars can be obtained at teh 0.04 to 0.002 arcsec level. Such information is obtained as a matter of course during HST observations. The proposed HST observations are the only optical means which can approach the expected accuracy of the individual HIPPARCOS star measurements, and these observations are REQUIRED, to achieve a reduction to a non-rotating coordinate frame at the level of the internal systematic accuracy of HIPPARCOS. Page 5 ------------------------------------------------------------------------------------ 5. Description of special scheduling requirements. The FGS observations are expected to take 51.2 sec/setting= (1/40 sec/sample)*(32 samples/integration)*(2^6 integrations/ setting) seconds/setting for the 11 settings/"observation" described in (3) above, due to the need to average over the "thermal hunt" cycle of the FGSs. The overhead per observation is expected to be twice the "time on target". Thus we have allotted 20 minutes for a single "average" observation with the FGSs. Similarly, PC exposure times are expected to be of the order of a few minutes for the EGOs (13.0