Authors: Yoshikiko Mizumoto, Yoshihiro Chikada, George Kosugi, Eiji Nishihara, Tadafumi Takata, Michitoshi Yoshida, Yasuhide Ishihara, Hiroshi Yanaka, Yasuhiro Morita, Hiroyuki Nakamoto
We are developing an analysis system DASH (Distributed Analysis Software Hierarchy) for efficient data processing of SUBARU telescope. SUBARU telescope will be able to produce a large amount of data. The predicted data production rate of 50 TB/year requires both high computational power and huge size of mass storage. It also demands a new astronomical data reduction and analysis system, which cooperates with SUBARU observation software system and the data archival system. As the first step of the development, we made a prototype of DASH for trial of the new computer technology.
There are a lot of workstations, Supercomputers, and mass data archives (capacity of 150 TB) connected in the high-speed network in SUBARU observatory. DASH is designed for this computer environment; considering role sharing on a distributed heterogeneous computer system. We adopted CORBA as a distributed object environment and Java for the user interface in the prototype of DASH.
The astronomical data reduction packages widely used have the full range of applications in spite of the same applications already being available in other systems. We intend to use the applications from other packages such as IRAF or Eclipse within DASH. The DASH prototype can use some applications from IRAF and Eclipse using wrapping technique.
Moreover, we implemented an analysis procedure object (APO) for the distributed system. APO has the following features; (1) data management function of observed data, temporary data, processed data, and analysis engines, (2) self-descriptive of record and the present state of operation of APO, (3) easy creation of a pipeline, (4) executable as a pipeline processing. We are planning to extend APO into Agent in the next step.
In the presentation, details of the Dash prototype will be presented including the design concept.
Authors: Toshiyuki Sasaki, George Kosugi, Junichi Noumaru, Tadahumi Takata, Yoshihiko Mizumoto, Ryusuke Ogasawara, Yoshihiro Chikada, Wataru Tanaka, and Jun A. Kawai
Subaru Observation Control System is reviewed over the overall structure, which is composed of several subsystems such as Telescope Control subsystem, Observation supervisor subsystem, Data acquisition subsystem, Data archival subsystem. Each subsystem consists of several processes to carry out observation operation in cooperating with other processes by passing control messages and by exchanging their status data.
User interface of the control system is described in aspects of operation control, quick-look of data, and making an observation journal. All acquired data is registered in database together with related data such as status and log data of the telescope and instruments. Observers and their observation proposal are registered to the control system as a NIS+ user and NIS+ group. User access to the control system is managed according to their registered operation level. Samples of graphic user interfaces will also be presented.
Keywords: Subaru Telescope, Observation control, User interface
Authors: Junichi Noumaru, Yoshihiko Mizumoto, Toshiyuki Sasaki, George Kosugi, Jun A. Kawai, Yasuhiro Morita, Itsushi Akiyama, Yasutoshi Kusama, Shigeaki Iwai
Subaru Observation Control System defines their hardware interface to instruments as the ports of Ethernet and Fibre Channel. Every instruments should be connected with either or both of the LANs. LANs which Subaru Telescope has are shown with their role. Regarding the data transfer to Hilo base, the first concern is that no data must be lost during transfer process, whatever troubles may happen on hardware or the network. In the hardware, we provide RAID, tape library at the summit and another RAID at the base facility. As the other measure in software, we have the data file management by Subaru Observation Software System (SOSS), which will be discussed in the presentation.
The schedule observation engine of the SOSS issues "abstract commands" which are common to instruments and which contents are abstract like "Expose" "Get ready for observation". The "abstract commands" are translated into commands which each instrument can understand. Every instruments has it own command structure and it varies by instruments. This translation is done by a function of schedule observation engine with referring preset "instrument table", dictionary for instrument commands. A few instruments may have fixed their commands and instrument tables at the time of the conference. Then, I can talk about the command translation in the schedule observation engine. Since the content of this talk is in the different field from the first topic, I may ask to divide the presentations into two.
Keywords: Subaru, Software, Network, Data Flow, Instrument Commands
Authors: Ryusuke OGASAWARA, Yoshihiro CHIKADA, Yoshihiko MIZUMOTO, George KOSUGI, Toshiyuki SASAKI, Jun'ichi NOUMARU, Tadafumi TAKATA, Kenji KAWARAI
The relationship between computational power and size of storage system is the essential point to utilizes the system for data acquisition and data analysis in the astronomy, and also is interesting from the viewpoint of computer science. Big science produces large a mount of data, e.g., Subaru Telescope of NAO-J has wide field in prime focus area, say 16 cm in diameter, and will be able to produce more than 50 GB of images especially in IR wavelength, per one night, i.e., about 15 TB per year at the maximum data production.
Recent technology teaches us the cost of computation, i.e., using CPU, is less expensive than the cost of storage, and it will be rather advantageous to use CPU each time to get astrophysical information from those large amount of raw data, in comparison with the method in which one stores intermediate results into secondary storage in each phase of data reduction and analysis. Based on this strategy, we have constructed a distributed hierarchical storage system, which consists of: 150 TB of Digital Tape Beta-CAM Format Tape Libraries, 2.4 Tera Bytes of Fast Magnetic Arrayed Disks, paralleled scalar data servers and Vector Parallel computing servers. The total CPU power of the system is approximately 20,000SpecInt92 in scalar processing and 50 GFlops for vector computing. This system is mainly used for off-line and detailed analysis of observed data, along with on-line data storing onto the Tape Libraries with suitably assigned tag for future access to data stored in the tape library those data for off-lined analysis.
The system provide a homogeneous storage environment for each scientists, and data analysis system will be implemented to use scalar processors and vector processors regarding to the properties of the tasks to be executed under a control of distributed object-oriented operating system. Beside, the management of Subaru telescope as a public observatory is an observatory open to the astronomical community suffered from distributed location, i.e., Mitaka headquarters in Japan and Hilo facility should cooperate not only in management but also in analysing data and system management but also software development and debugging data along with software development or debugging. So, we will prepare a virtual headquarters in between these two institutes, i.e., just in the middle of the pacific ocean between Japan Main Island and Hawaii Big island.
The details of hardware and these software environment will be presented in the meeting.
Keywords: subaru, storage hierarchy, data management, distributed computation, virtual office
Authors: George Kosugi, Toshiyuki Sasaki, Yoshihiko Mizumoto, Tadafumi Takata, Jun A. Kawai, Yasuhide Ishihara
Observation dataset has an important role in Subaru Observation Software System (SOSS) in order to connect the observation control system with the data analysis system. An observation dataset includes abstract commands of getting both a science object data and its calibration data indispensable to calibration. Acquisition conditions of each calibration data are also defined in the observation dataset. For example, the acquisition timing of each calibration data is specified relative to the acquisition timing of the corresponding science object. Optimization and re-arrangement of the observation schedule, and recommendation of the next observation command in the manual operation mode are taken place during observation by using the observation dataset. The observation dataset is used for automated data analysis, such as pipeline processing, in the data analysis system in the Hawaii Facility at the foot of Mauna Kea. The feedback of the control parameters and real-time quality assessment of the acquired data to the observation scheduling will be achieved using the supercomputer system in the Hawaii Facility in a few years.
Keywords: Subaru, Software, Dataset, Scheduling, Data analysis
Authors: Tadafumi Takata, Ryusuke Ogasawara, Kenji Kawarai, Tadahiro Yamamoto
Subaru Telescope is one of the most largest optical and infrared telescope in the world and will produce the flood of observational data. The data will be sent down directly via ATM-link from the summit to our base facility at Hilo. Therefore Subaru Data Archive system must be very large, but can serve requested data to observers and archive users as fast as possible, and also have the function of on-line registration of observational data information into Data Base Manager without any delay, lack and inconsistency with observation at the summit.
A part of such functions can be implemented by using "Multi-master Replication" of ORACLE7.3, in which two independent ORACLE server refer each other and data registration is available even in the case of down in one server by switching target database server. The data search engine is based on MOKA2 (2nd Version of Mitaka Okayama Kiso Archival system), which is now in operation in Japan, and some additional function such as on-line user certification and linkage with data analysis system.
Large telescopes, while on the other hand, need very large (equal to very deep) catalog data for effective observation such as in the case of guide star searching, target selection and so on. The data base system therefore must do very fast serve of any catalog data in the selected area. By using the "partition view" function of ORACLE7.3, we've got so good performance of searching data from USNO-A Catalog, which includes about 500 million stars.
Keywords: data archive, database system, Large Telescope
Authors: Yoshihiro Machida, Jun Nishikawa, Koich Sato, Toshio Fukushima, Masanori Yoshizawa, Yukihiro Honma
MIRA Project is a series of optical-infrared stellar interferometers by National Astronomical Observatory of Japan. We call the first step MIRA(Mitaka optical-InfraRed Array)-I project, and now, we aim to get stellar fringes and skills of fringe tracking. MIRA-I is a prototype for demonstration, and followed by the next projects MIRA-I.2, MIRA-II, MIRA-SG, and MIRA-III.
The MIRA-I's instrument is located at National Astronomical Observatory in Mitaka, Tokyo. It consists of two element telescopes with 25 cm aperture placed on a 4m N-S baseline. The aperture size is variable from 6 cm to full aperture according to the seeing and observing wavelength. We use star light around 800 nm wavelength for observations, a He-Ne laser for alignment, and a white-light source for determine the zero point of delay line.
Element telescopes are Coude type. From the telescope, the two beams are led to the delay line system on an optical bench. Between the two optical path length are compensated the geometrical delay by 2m coarse delay line, earth rotation delay by 200 mm fine delay line, and atmospheric turbulence delay by 100-micron PZT actuator. Stellar tilt motion is compensated with high sensitivity tip/tilt servo system, using photon counting APD and digital filter. Furthermore, for the fringe tracking, we are developing a fringe tracking method with delay modulation and servo system.
Experimentally, we obtained artificial white light fringes at Dec. 1995. And now, the instruments are being developed to try detecting stellar fringes.
Keywords: interferometer, optical interferometer, high-resolution imaging, astrometry
Authors: Koichi Sato, Jun Nishikawa, Masanori Yoshizawa, Toshio Fukushima, Yoshihiro Machida and Yukihiro Honma
We are promoting the construction of the ground-based optical and infrared interferometers, called MIRA Projects(I, II, III) at the National Astronomical Observatory of Japan. We are now making experiments of stellar fringe detection and fringe tracking at Mitaka (MIRA-I). MIRA-I system consists of 25 cm coude telescopes with a 4m N-S baseline. MIRA-I.2 is an extended version of MIRA-I. The outlines of MIRA-II and MIRA-III are as follows: MIRA-II(seven 30 cm siderostats, 680m baseline, at Mitaka), MIRA-III (thirteen 1.5m siderostats, 1.4 km baseline, at Mauna Kea). Development of MIRA-I.2 system is discussed in this paper.
The MIRA-I.2 system consists of siderostats, beam reducer, vacuum delay lines, baseline metrology system, tip-tilt mirrors, interferometer optics, and fringe detector. Two siderostats, of which aperture of the flat mirror is 300 mm, are placed apart by 4 meters, in the north-south direction. Beam reducer is a Cassegrain optics with the paraboloidal primary and secondary mirrors of D1=200 mm, D2=30 mm, respectively. Metrology system with laser interferometers is set up to stabilize the baseline length for astrometry. Piezo-driven tip-tilt mirrors are equipped for the correction of image motions due to the air turbulence. Delay lines are placed in a vacuum tube. Observation wavelengths range from 400 nm through 2500 nm. By the developments of MIRA-I.2, it is aimed to establish the basic techniques of astrometry and future projects, especially of the MIRA-II.
Key words: optical interferometer, astrometry, high resolution imaging
Authors: Yukihiro Honma, Jun Nishikawa, Yoshihiro Machida, Koichi Sato, Takashi Kasuga,
A delay line system for long base line optical interferometers is under development in MIRA project. This system should compensate variation of the optical path length due to the diurnal motion, atmospheric turbulence and other delay errors. In our delay line system, a 100É m PZT actuator is going to be used as a fast delay line, a micro-stepping motor driven cart as a fine delay line, and a cable-dragged cart as a long delay line. The test structure of the fine delay line consists of a micro-stepping motor cart and a cat's eye cart. It is the most important for the fine delay line to be driven at a steady speed to a theoretical value of the delay rate. Particularly, high frequency jitter of driving speed must be prevented so that fringe movement may not be less than Å}É/10 in the integration time. To accomplish this capability, we studied mainly a geared motor and the structure of drive. As a consequence, our motor cart had a good performance without special control components. We also made tests of isolation between motor cart and cat's eye cart or between cat's eye cart and cat's eye assembly with rubbers. It may be possible to reduce high frequency component of delay errors sufficiently by using rubbers without losing ease of control. According to these experiences, we have started designing a prototype of the vacuum fine delay line for MIRA-II project having a 680 m base line.
Keywords: interferometer, optical interferometer, delay line
Authors: J. Nishikawa, K. Sato, T. Fukushima, M. Yoshizawa, Y. Machida, and Y. Honma
After MIRA-I and MIRA-I.2 experiments, we hope to enter into the long-baseline interferometer world. The next Mitaka optical and infrared array project is called MIRA-II. It consists of four fixed telescopes as an astrometric array and three movable ones out of sixteen stations for imaging. They are placed in a sideways T-configuration with three 128m arms and extended lines from the north end and the south getting the longest baseline of 680m. Each of the telescopes is a 30 cm siderostat added with a 20 cm beam reducer. Fine delaylines will be 20 m long to make 10 min integration and long delaylines about 200 m to observe objects at z=30 deg. The astrometric array aims at the 1 mas astrometry using the most precise baseline metrology system and the whole array the 0.2 mas resolution imaging with the 680m baseline.
MIRA-III is a proposal of Mauna Kea optical/IR array having 1.4 km baseline with 1.5m telescopes. Its shape is a modified Y-configuration. It also aims at precise astrometry including many quasars as well as high resolution imaging of fainter stellar objects than MIRA-I.2.
MIRA-SG, a proposal in the near future of Mauna Kea optical/IR array connecting Subaru with GEMINI, is one of the largest interferometer with 800m baseline by 8m telescopes. It become possible by using optical fibers fed from each cassegrain focus with an adaptive optics system. Delaylines and detectors may be put in the control building of Subaru Telescope. Of course Keck telescope and other large telescopes on Mauna Kea are also candidates to connect with Subaru.
Keywords: interferometer, optical interferometer, high-resolution imaging, astrometry
Authors: K.-I. Morita and S. Katagiri
We present a maximum-entropy algorithm using the bispectrum observed with interferometers. It is iterative and based on the steepest-ascent method. Imaging simulations show that this algorithm can reconstruct more reliable images of weak, extended sources than can the conventional self-calibration method.
Authors: A. Tokunaga, J. Bell, K. Hodapp, J. Hora, P. Onaka, J. Rayner, L. Robertson, T. Young, N. Kobayashi, M. Weber, D. Warren
A 1-5 micron Infrared Camera and Spectrograph (IRCS) is described. The IRCS is be a facility instrument for the 8.3-m Subaru Telescope. The main specifications of this instrument are as follows. Spectrograph section: resolving power is 20,000 with two pixel sampling a slit width of 0.15 arcsec. The spectrograph utilizes an echelle grating and a cross-disperser grating to provide large spectral coverage in a single exposure. Camera section: serves as slit viewer and as a camera. Two pixel scales are available, 0.022 arcsec/pixel and 0.060 arcsec/pixel. Grisms providing 400-1400 resolving power will be available. Each section will utilize an ALADDIN II 1024x1024 InSb array. The instrument specifications are optimized for 2.2 microns using the tip-tilt secondary and an adaptive optics system. A cryogenic beamsplitter near the focal plane of the telescope sends visible light to a wave-front sensor. Completion date is Jan. 1999. Design and technical details will be described.
Keywords: infrared, instrument, Subaru, camera, spectrograph
Authors: Motohide Tamura, Hiroshi Suto, Hideki Takami, Koji Murakawa, Yoichi Itoh, Noboru Ebizuka, Norihide Takeyama, Katsuyuki Chikami, Norio Kaifu
CIAO is a stellar coronagraph imager now under development for use on the Subaru 8 meter telescope. The purpose of this instrument is to obtain diffraction limited (0.05 arcsec at 2 micron) images of faint objects in close vicinity of bright objects. For achieving both high spatial resolution and high dynamic range, the instrument is used with the Subaru Cassegrain adaptive optics and designed to have a sophisticated coronagraphic capability.
CIAO is optimized for use at near-infrared wavelengths (1 - 5 micron) where the adaptive optics works most efficiently and the relative effect of scattering by telescope and instrument optics is smaller. There are three optical modes for observations and a pupil imaging mode for optical alignments, all of which are best optimized at J and K bands. The optics elements are all refractive and composed of smallest number of surfaces as possible for reducing the scattering effect. The size and shape of the occulting masks and Lyot apodizing stops are determined from computer simulations. A number of sizes and shapes of these components are selective. Besides the standard broad band imaging, a number of narrow band imaging as well as slit spectroscopy with grism is supported with/without coronagraph. Polarimetry is also available for all observing modes. CIAO will employ a 1024x1024 InSb array as a detector. Great care is taken to design the "tension-strap supported" cryostat which minimizes the flexure within the cryostat.
Performance of CIAO is extensively evaluated with computer simulations which take into account both primary mirror errors and atmospheric turbulence. CIAO will not only be able to employ much smaller occulting masks but also to achieve higher dynamic range than the previous stellar coronagraph instruments.
Keywords: Infrared - Coronagraph - Adaptive Optics
Authors: Mamoru Doi, Hisanori Furusawa, Fumiaki Nakata, Sadanori Okamura, Maki Sekiguchi, Kazuhiro Shimasaku, and Norihide Takeyama
We describe the design and performance of a dichroic-mirror camera (hereafter DMC) which can take 15 narrow-band images simultaneously. We separate the wavelength range of 390-950 nm into 15 narrow bands with 14 dichroic filters. The detector of DMC is a mosaic CCD camera which has 15 CCDs (TI TC-215). When we put it to the MAGNUM 2-m (F/9) telescope being built at Haleakala, Hawaii, the field of view becomes about 4.5 arcmin in diameter. The design of optics shows that we can get image size of about 0.13 arcsec r.m.s. or better (without atmosphere), though we use only two different kind of lenses (the camera lens and the collimator). Simulations using spectra of galaxies and QSOs show that DMC can get good signal-to-noise (S/N = about 10/band/object) images of galaxies (I_AB = 22) and QSOs (I_AB = 23) for 30 min - 1 hour exposure.
Future prospects of dichroic-mirror system are also discussed. For an 8-m class telescope with a field of view of 15 arcmin, we can obtain 10000 narrow-band images of galaxies down to B=26 for about 1-hour exposure. To get higher resolution, DMC combined with Fabry-Perot is an interesting application. We discuss efficiency of the system comparing with those of multi-slit and multi-fiber spectrographs, and show that DMC is a very efficient instrument for faint object surveys.
Authors: Kentaro Motohara, Toshinori Maihara, Fumihide Iwamuro, Shin Oya, Masatoshi Imanishi, Hiroshi Terada, Miwa Goto, Jun'ichi Iwai, and Hirohisa Tanabe, Hiroyuki Tuskamoto, Kazuhiro Sekiguchi
We have been developing a unique near-infrared spectrograph for Subaru Telescope, OHS (OH-airglow suppression Spectrograph), which achieves highest sensitivity among the low to medium resolution infrared spectrographs for 8-10 m telescopes. OHS suppresses sky background emission by a factor of 20 to 40 in the J, and H bands with a special spectroscopic filter system. It is expected to extend the limiting magnitudes by about 1.5 to 2 mag.
CISCO is a back-end camera & spectrograph unit of OHS, based on a single 1024 x 1024 HgCdTe array (HAWAII). It is also supposed to be mounted on the Cassegrain or Nasmyth focus to be served as an independent general-use infrared camera & spectrograph. With a 2-D slit mechanism, one can select either the 2'x2' FOV imaging mode in the z, I, J, H, and K bands, or the long-slit spectroscopy mode with grisms which achieves resolving power in wavelength of 400 to 1000 depending on the slit width.
To make use of the full potential of OHS, it is necessary to minimize the dark current and readout noise of the detector, because OHS reduces the natural background to a great extent and its detectivity is no more background-limited. Therefore, we have developed a detector electronics with total readout noise less than 10(e- rms/readout) and a multi-purpose data acquisition system with which the multiple readout scheme has been tested to decrease the noise by another appreciable factor.
The results of test observations using a 1.5m telescope will also be presented.
Authors: Masatoshi Imanishi, Hiroshi Terada, Miwa Goto, and Toshinori Maihara
LEWIS (L, M-band Echelle Wide coverage Intermediate-resolution Spectrometer) is an infrared spectrograph designed primarily for spectroscopy in the 3 micron region. It is an echelle type spectrograph using the prism and grating cross-disperser. Using LEWIS, we can observe the whole L-band (2.8 micron - 4.2 micron) in one exposure with a resolving power over 1250, which makes observations very efficient. A Santa Barbara Research Center (SBRC) 256 x 256 InSb array is employed as a detector. The grating used is characterized by large groove spacing of 125 micron and is utilized at very high orders, 25th - 37th order in the L-band. A closed-cycle cooler is employed to keep the optics at 80 K, and to maintain the detector at 30 K. Up to now, some scientific observations have been made at Steward 60 inch telescope on Mount Lemmon, Steward 61 inch telescope on Mount Bigelow, and Wyoming Infrared Observatory 88 inch telescope, on Mount Jelm. The achieved throughput of the spectrograph including the quantum efficiency of the detector is about 20%. With the present detector control system, the background limited condition is achieved at 3.5 micron utilizing 8 times multi-sampling readout procedure, and a limiting magnitude of 9.0 mag is achieved for S/N = 10
Authors: Daigo Tomono and Tetsuo Nishimura
We are constructing MIRTOS, an infrared imager system to evaluate and monitor the performance of the Japanese National Large Telescope SUBARU. The system consists of two array cameras. One of the camera is for near-infrared (NIR) and the other for mid-infrared (MIR). They capture images simultaneously at the rate fast enough to freeze the seeing. Simultaneous NIR images are useful not only for evaluation of the image quality of the telescope but for Two-wavelength Shift-and-Add that enhances angular resolution of MIR images. The system also has a telescope emissivity mapper mode that images the telescope entrance pupil in MIR. For the NIR channel, a SBRC InSb array with 256 pixels-square is used. Pixel scale is lambda/2D = 0.028 arcsec/pixel that is optimized to detect position of the brightest speckle in atmospheric disturbed images at 2.2 microns with enough field of view for point like reference sources. For the MIR channel, a SBRC Si:As IBC array with 320 by 240 pixels is used with pixel scale of lambda/3D = 0.068 arcsec/pixel that takes enough samples to make diffraction limited images at 8 microns. In the emissivity mapper mode, a temperature tunable black body can be inserted just outside the dewar window as an absolute calibration source of the telescope emission. MIR Optics images the secondary mirror and the surrounding sky. In this paper, we will present the idea and preliminary result of Two-Wavelength Shift-and-Add, detailed design of MIRTOS, and the latest status of construction of the system.
Keywords: SUBARU Telescope, near-infrared and mid-infrared, Two-Wavelength Shift-and-Add, imager, emissivity mapper
Author: Norio Kaifu
Authors: Noboru Itoh, Yasushi Horiuti, Kouki Asari, Manabu Sawa (Mitsubishi Electric Corporation), Masahiko Hayashi
The SUBARU Telescope introduces a tip-tilt and chopping mechanism for a IR secondary mirror, which is of ULE light weight type, 1.3m in diameter and 180 kg in weight. Performance targets of the tip-tilt and chopping mechanism are 30 Hz control band width with 0.01'' resolution for tip-tilt and 30'' amplitude with frequency of 5 Hz for chopping. To achieve these targets, a system combining 6-point dynamic drive mechanism and 15-point passive support mechanism against for gravity are developed and compact actuators of electric magnet type for the drive mechanism are employed. Tests with a dummy mirror shows that the performance targets are achieved. This paper describes the design and test results of the tip-tilt and chopping mechanism.
Key words : Chopping secondary, Tip-tilt, Tracking
Authors: Saeko S. Hayashi, Yukiko Kamata, Tomio Kanzawa, Akihiko Miyashita, Masao Nakagiri, Tetsuo Nishimura, Takeshi Noguchi, Kiichi Okita, Norio Oshima, Goro Sasaki, Yasuo Torii, Masami Yutani, and Tsuyoshi Ishikawa
One of the major problems to retain the efficiency of a telescope is to maintain high reflectivity in wide wavelengths and low emissivity in the infrared term of the telescope optics. For coating large mirrors, we employ conventional evaporation scheme, in the expectation of uniform coverage of the film. We will report installation and the performance verification of the coating facility of Subaru telescope. This facility consists of a washing tower for stripping the old coating from the primary mirror, an evaporation coating chamber, two trolleys for the primary mirror, and a scissors-like primary mirror lifter.
Tests with large coating chamber at Mauna Kea as well as with smaller chamber at Mitaka will be discussed. To supply a large number of filaments with uniform quality, practical solution is to pre-wet the filaments and keep them in a controlled manner before the evaporation. In the initial test, Aluminum film over the large area exceeded the number targeted for the thickness and yet the uniformity turned out to be better than the specification. Reflectivity of the fresh surface was over 90% at visible wavelength. In September 1997, we will re-aluminize 1.6m Infrared Simulator at Mitaka for the first time to apply pre-wetted filaments. The result should confirm our strategy for coating the first secondary of 1.33 m diameter in late autumn of 1997 and eventually the 8.3 m primary mirror in early 1998. Still uncertain are the contamination in various areas especially in the washing process and in the preparation of filaments.
Keywords: coating, stripping, emissivity, reflectivity, evaporation coating
Authors: Yukiko Kamata, Saeko S. Hayashi, Takeshi Noguchi, Tomio Kanzawa, Goro Sasaki, Yasuo Torii, Masami Yutani, and Tsuyoshi Ishikawa
We have conducted a series of coating experiments using the newly installed 1.6m evaporation chamber at the Advanced Technology Center(ATC) of the National Astronomical Observatory. The main task of this chamber is to re-aluminize the 1.6m mirror of the Infrared Simulator at the ATC. The design concept of the 1.6m chamber is basically the same with the 8.3m coating facility for Subaru Telescope. Therefore we could utilize this chamber to evaluate fundamental performance of the larger chamber. The extensive coating experiments were done in the spring and autumn of 1996, and is planned in the autumn of 1997.
Reduction of the number of the filaments has lead to the increase in their size, which gives difficulty in annealing process. Attempts are being made to secure the sufficient metal loads on the filaments. Then the filaments are fired to measure the spray pattern of a single filament exposure, or the uniformity pattern resulted from the full setup. Using small slide glasses, the important parameters of the resultant reflecting film that are the thickness, the uniformity of the thickness, and spectroscopic reflectance are measured. The absolute value of the reflectivity is estimated to be around 91 % immediately after opening the chamber. In order to cover a wide range of observing wavelengths for the Infrared Simulator, and eventually for the optical-IR Subaru Telescope, it is necessary to seek for a higher evaporation rate with these chambers.
Keywords: mirror coating, evaporation coating chamber, coating uniformity, spectroscopic reflecting performance
Authors: Yasuo Torii, Saeko S. Hayashi, and Masahiro Toda
Since Subaru Telescope adopted a flushing-type enclosure, we considered that fine cinders at the Mauna Kea summit are blown into the enclosure by wind and get stuck on the optical elements, particularly on the primary mirror as it will be looking up the sky for long hours during its operation. Contamination on the primary mirror surface decreases its reflectivity and increases its emissivity and scattering. All these will degrade the telescope efficiency over the wide range of the wavelengths, which is what we would like to prevent as much as possible. Dry ice(Carbon dioxide snow) in-situ cleaning system is one of the candidates for cleaning large area, and we have conducted experiments to understand why the dry ice cleaning method is better than others such as dry air or nitrogen gas blowing technique. We used 3.5 inch test mirrors to apply various cleaning schemes and measured the resultant performance. The result clearly confirmed the better performance of the dry ice cleaning. We have further investigated to refine the parameters of this method, such as the shape of the nozzle, distance between the mirror, blowing time interval, and direction with respect to the mirror. We will report the results and implication of these experiments that lead to the design concept of the dry ice in-situ cleaning system for Subaru Telescope.
Keywords: in-situ cleaning, dry ice cleaning, contamination, scattering, large optics
Authors: Satoshi Miyazaki, Maki Sekiguchi, Katsumi Imi, Norio Okada, Fumiaki Nakata
Large format and high QE CCDs are one of the most critical components in modern astronomical instruments. The format of 2048X4096 (13.5~15 micron pixel) CCD with three side buttable edge is supposed to be ``industrial standard'' and some of the devices are becoming available in astronomical communities. Because such a large device is still premature, careful characterizations are crucial. We have developed an evaluation system consisting of a dewar which holds two CCDs described above, analog electronics and QE measurement bench. The characterization result of Hamamatsu photonics's device, which is developed by joint efforts with NAOJ recently, will be presented mainly in the talk as well as the result of MIT/LL devices. Second topic of the talk will cover the mosaicing strategy of the three side buttable CCDs. We will epoxy a piece of block underneath the CCD package which has two alignment pins. The block absorbs the difference of package of CCD when we mount the CCDs on a cold plate. We insert three pieces of thin metal sheet with different thickness between the package and block to compensate the tilt of the CCDs. Alignment of CCD pixels to the line between the pins will be done by pushing CCDs on side before the epoxy is cured. Finally we will report the status of Subaru Prime Focus Camera (SuprimeCam), which is a wide and deep field instrument --- very unique among 8-m class telescopes. The camera holds ten 2K x 4K pixel CCDs and covers 30' x 24' under the fast F/2 beam. Because of this fast optics, the flatness of the CCD plane needs to be better than +/- 10 microns. The mounting and mosaicing technique described above is developed and employed to satisfy this requirement.
Authors: Kunio Noguchi, Hiroyasu Ando, Hideyuki Izumiura, Satoshi Kawanomoto, Wataru Tanaka, and Wako Aoki
This paper presents a brief description of a high dispersion spectrograph (HDS) now under construction towards the first-light phase of Subaru Telescope. HDS is an echelle spectrograph with grating as a post-disperser. It is located at a nasmyth focus. The collimated beam size is 272 mm, and the echelle is a 1Å~2 mosaic, 300 mm by 840 mm in total size, of 31.6 gr/mm, R-2.8 echelles. The overall 'throughput' (resolution Å~ slit width product) achieved is 38,000 arcseconds. HDS has two cross-dispersers which are optimized for blue- and red-wavelength regions, respectively. Each cross-disperser is a 2Å~1 mosaic, 630 mm by 410 mm in size. A plane mirror is selected instead of a cross-disperser for getting a single-order spectrum with use of narrow band filters.
The camera is of catadioptric type system, consisting of three correctors and a mirror. It is a large (600 mm aperture) f/1.0, all-spherical system. Fused-silica lenses are used for correctors. It spans the entire chromatic range from 0.30 to 2.0 microns. It delivers 10-micron images on average within a flat 60 mm-diameter focal plane area in the wavelength range from 0.3 to 1.0 microns, without refocusing. Refocusing is required in the range from 1.0 to 2.0 microns. This image quality corresponds to typical limiting spectral resolutions well above 300,000 though the resolution will generally be limited to less than this by the entrance slit width and finite pixel sizes.
The detector will be a 1Å~2 mosaic of 2k Å~ 4k CCDs with 15-micron pixels. It is a thinned, backside-illuminated, CCD, which spans the spectral region from 0.3 to 1.0 microns with very high overall quantum efficiency. Typical limiting spectral resolution will be 100,000.
Keywords: high dispersion spectrograph, echelle spectrograph, catadioptric camera, optical spectra, Subaru Telescope
Authors: Masanori Iye and Noboru Ebizuka
A Fiber Multi-Object Spectrograph (FMOS), is under design study as one of the possible second phase instrumentation for the 8m Subaru telescope. The author has been proposing to establish an additional observational mode by combining FMOS fiber positioner and the High Dispersion Spectrograph (HDS), a first phase instrument at Nasmyth focus, by means of fiber bundles with pupil slicing geometry transformers. A fiber bundle geometry transformer consists of a pupil slicer unit, a 7-fiber bundle, and an output F-ratio transform optics.
With this arrangement, an object image at the F/2 prime focus can be led onto the slit of the HDS at the F/12 Nasmyth focus. By aligning the 7 output ends of fibers laterally, one can configure a 1 arcsec circle entrance aperture at the prime focus into a square input beam of 2.2 arcsec wide and 0.33 arcsec high at the HDS entrance slit. This wide and thin input beam enables to position up to 6 (15) objects in the narrowest inter-order space of 4 (10) arcsec in the blue (red) spectral region of echelle format produced by the provided cross disperser.
This new mode of observation enabling simultaneous exposure of several objects down to 21 mag at the cost of reducing the spectral resolution to 18,000 will be extremely useful especially in studies of kinematical property and chemical abundance of objects in nearby galaxies and for instance lateral correlation analysis of quasar absorbers.
The fiber pupil slicer opens up another possibility for achieving HDS a spectral resolution as high as 150,000. This mode will gain the speed of HDS by opening the entrance aperture from 0.27 arcsec to 0.8 arcsec to attain this high spectral resolution. These functions are similar to those provided by an image slicer but the pupil slicing will give more uniform and stable illumination at the entrance slit and will perform better.
Keywords: Spectrograph, Pupil Slicer, Optical Fibers, Multiobject Spectroscopy
Authors: Noboru Ebizuka, Masanori Iye and Toshiyuki Sasaki
Some of astronomical instruments under construction for the 8m Subaru telescope (to be commissioned on Mauna Kea, Hawaii, in 1998) are planned to use grisms of several types to achieve different modes of spectroscopic observation. Faint Object Camera And Spectrograph (FOCAS), which uses Si-CCD detectors for visible light observation, is planned to use seven grisms with different dispersions and different optimized wavelengths, and Coronagraphic Imager with Adaptive Optics (CIAO), which uses an InSb array detector for near infrared observation, will employ three grisms. In order to establish a practical method to manufacture grisms of the highest angular dispersion for FOCAS and for CIAO, we have been trying to develop grisms made of high index materials.
We show that a crystalline LiNbO3 grism and a hybrid grism made of a LiNbO3 transmission grating and a ZnS, in spite of their birefringent properties, can be new and powerful dispersing elements with high refractive indices to realize high spectral resolution for optical to near infrared astronomical spectrograph with transmission optics.
Key words: Astronomical spectrograph, Grism, Ion-beam etching, Optically anisotropic crystal, Transmission grating
Authors: Hiroshi Ohtani, Tsuyoshi Ishigaki, Hiroyuki Maemura, Tadashi Hayashi, Minoru Sasaki, Kentaro Aoki, Shinobu Ozaki, Takashi Hattori, Hajime Sugai
Development of a spectrograph for area spectroscopic observations of faint extended objects in optical spectral region (Ohtani et al., SPIE Vol.2198, 229, 1994) was completed and the instrument is now in commission at the 188 cm telescope of the Okayama Astrophysical Observatory. With this instrument can be made four different kinds of spectroscopy which are switched to one another by remote operation. The four modes are imaging Fabry-Perot interferometer observation, and integral field spectroscopy (IFS), filter imaging at several narrow (or wide) bands, slit scan in the long-slit spectrograph mode. The latter two are conventional types.
Two Fabry-Perot etalons of Queensgate Instruments with wide band coatings from 400 nm to 700 nm and resolving power R = 300 (tunable narrow band filter) and 7000, are available. When the tunable filter is used, a central part of 1.5' diameter of the total field of view 4.5' is quasi-monochromatic; within this part the central wavelength of transmission does not differ more than ten percent of the bandwidth when the tunable filter is used.
Prior to development of the spectrograph, behavior of characteristics of the etalons with variations of ambient temperature was carefully examined at a laboratory. Based on the results of the experiment, temperature of the etalon in the spectrograph at the telescope is stabilized within +-0.5 C to attain high performance that drift of the transmitting wavelength does not exceed one tenth of the width of the Airy Profile during observations.
For the IFS mode, the type of the TIGER spectrograph for the CFHT is employed. In spectrograph, the microlens array is shared by dual-channel enlarging optics, one channel acquiring a target object and the other an 'uncontaminated' sky field well apart from the target. With the 188 cm telescope a 9"x15" (7x11 lenslets corresponding 1.3" square) target field and a smaller sky field 3.7' apart are observed simultaneously. The spectral resolving power from 300 to 1500 is attained by selecting a relevant grism. Total quantum efficiency of the system was measured by using standard stars and it has been found to be 6 - 10 percent at yellow - red. However, effective quantum efficiency is somewhat lower because pixels superposed by adjacent spectra are removed in data reduction.
Keywords: Optical spectrograph, Multimode spectrograph, Tridimensional spectroscopy, Imaging Fabry-Perot interferometer, Integral field spectroscopy with microlens array
Authors: Hajime Sugai, Hiroshi Ohtani, Tsuyoshi Ishigaki, Tadashi Hayashi, Shinobu Ozaki, Takashi Hattori, Minoru Sasaki, Norihide Takeyama
We are building the second version of the Kyoto tridimensional spectrograph (Ohtani et al., this symposium). This will be mounted on the MAGNUM telescope, which is a 2-m telescope under construction at Haleakala, and also on SUBARU. The spectrograph has four observational modes: Fabry-Perot imager, integral field spectrograph (IFS) with a microlens array, long-slit spectrograph, and filter-imaging modes.
The new spectrograph is significantly better than the first version in several ways. The IFS has as many as 37 x 37 microlenses with 0".39/microlens when mounted on MAGNUM. The optics is designed to be used in wide wavelength ranges from 360 nm to 900 nm. The transmission at any wavelength between 370 and 900 nm is designed to exceed 50% for the collimator + camera system, and to exceed 30% even at 360 nm. In order to achieve high efficiency at short wavelengths, we use an anti-reflection coated backside-illuminated 2K x 2K CCD. We are also planning a further improvement by using multi-layer anti-reflection coatings for lenses, in a collaboration with National Astronomical Observatory, Japan.
In order to assure good image quality under a severe weight limit of 150 kg for this instrument, we have calculated the flexure of the instrument for all telescope attitudes by finite element analysis, and succeeded in limiting the maximum flexure to 30 um. This does not degrade image quality. The movements on the CCD of the light from the center of the focal plane have also been simulated, depending on the telescope attitudes. This is important to obtain not only a good image, but also a correct flat field and wavelength calibration for IFS mode. The movements are expected to be almost within one pixel for any attitude, which is considered to be small enough.
Keywords: Tridimensional multimode spectrograph, Fabry-Perot, Microlens array, High spatial resolution, Wide wavelength range
Authors: Hideki Takami, Naruhisa Takato, Masashi Otsubo, Tomio Kanzawa, Yukiko Kamata, Koji Nakashima, and Masanori Iye
The adaptive optics system for Subaru 8.2 m telescope of the National Astronomical Observatory Japan have been developing for the Cassegrain near-infrared instruments (CIAO: coronagraph imager with adaptive optics, IRCS: infrared camera and spectrograph). The system composed of a wavefront curvature sensor with 36 subapertures employing photon-counting avalanche photodiode modules, and a bimorph deformable mirror with also 36 electrodes. We are expecting to get the Strehl ratio 0.6 and 0.4 at K-band, for objects close to bright guide stars and for R = 16 mag guide stars respectively, at the median seeing of 0.45 arcsec at Mauna Kea. The system will be in operation from 1998 as natural guide star system. We are studying about its upgrade to a laser guide star system cooperating with infrared wavefront tilt sensor for obtaining nearly full sky coverage.
The full size prototype system has constructed and has been tested attached to the 1.6 m infrared telescope at our observatory in Tokyo. The system has identical optical design, deformable mirror, loop control computer to those for Subaru system, while the wavefront sensing detectors were less-sensitive analog APD for this. We succeeded to get closed loop images with diffraction limited core at K-band. The Strehl ratio and improvement was around 0.5 and a factor 20 respectively at K-band under the average seeing of 2 arcsec during the observation. The loop speed of the system was 2 K corrections per second.
Key words: adaptive optics, curvature sensor, bimorph mirror, natural guide star, laser guide star
Authors: N. Takato et al.
Authors: Wataru Tanaka, Toshiyuki Sasaki, Takeshi Noguchi, Kiichi Okita, Kyoko Nakamura, Fumio Ito, Yoshio Katsuki, and Sachiko Ishihara
The hierarchical structure of the telescope control system is adopted for the Subaru Telescope. The system is made up of TSC ( telescope control computer ), 3 MLPs ( mid-level processors ), and many LCUs ( local control units ). All devices are connected
through local network. The TSC is connected to the OBS ( observation supervisor ) through the local network, so-called C-LAN, and operated by observers with the OBS.
The MLPs connect to the TSC through the local network, so-called M-LAN. Each discharge of the MLPs is as follows: The MLP1 controls whole telescope driving mechanisms consisting of an alt-azimuth driver, an enclosure driver, secondary and tertiary mirror drivers, instrument rotator and image rotator drivers, peripheral optics ( an auto-guider, a Shack-Hartmann wave front sensor, a slit viewer ) drivers. The MLP2 intently engages in control of active supports of the primary mirror. The MLP3 engages in the environmental control and the system maintenance.
Keywords: Subaru Telescope, Telescope Control
Authors: Takeshi Noguchi, Wataru Tanaka, Norio Kaifu, Junichi Noumaru, Kiichi Okita, Takeo Shimizu, and Noboru Itoh
On-shop test erection has been carried out at the Hitachi Zosen factory in Osaka since 1996. The purpose of this on-shop erection is to check and adjust the sizes of parts, to minimize the assembling errors, to test the telescope drive and active support system of the primary mirror, and to review the processes of final assembling.
The alt-azimuth Subaru telescope structure weighting 500 ton is supported by six hydrostatic oil-film pads and is driven successfully by direct drive. The additional tests for the instrument rotator drives, cable wrappers, and other were also carried out.
We will report the preliminary results of the control efficiency.
Keywords: Telescope control system, On-shop test erection, Subaru telescope