 |
|
Astrophysikalisches Institut Potsdam
Opt. Instrumente
c/o Martin M. Roth
An der Sternwarte 16
D-14482 Potsdam
Germany
|
|
|
PMAS, the Potsdam Multi-Aperture Spectrophotometer, is a new integral field spectrograph which has been designed and built at the Astrophysical Institute Potsdam (AIP) for use at the Calar Alto 3.5m and other telescopes. Integral field, or "3D" spectroscopy provides spectra with 2-dimensional spatial resolution for extended objects in one exposure simultaneously. Unlike to some multi-mode instruments, that incorporate so-called integral field units (IFU) in front of a long-slit spectrograph, PMAS is a dedicated 3D spectrograph, based on the fiber-coupled lensarray type of instrument (for a more detailed technical description, see PMAS homepage and references therein). To date the 3D technique has predominantly been applied to the study of the kinematics of galaxies, but it is a powerful technique for many other astrophysical applications too.
PMAS is specifically designed to address the problem of accurate background subtraction of faint point sources which are superimposed on a bright, spatially variable surface brightness distribution. Typical examples are stars or gaseous nebulae in nearby galaxies which are difficult to observe with conventional slit spectrographs because of background contamination. Mapping the continuum and emission line background in two dimensions around a background-limited point source yields a significant advantage in eliminating systematic spectrophotometric errors which are otherwise difficult or even impossible to detect.
Extragalactic stellar spectroscopy, providing important data on stellar populations and galactic evolution, has been developed on 4m-class telescopes (see e.g. Stephen Smartt's webpage Massive stars in the local group ). Progress beyond the local group is possible with 8-10m telescopes: Bresolin et al. 2001, Ap J Letters 548, L159 , see also picture of NGC3621 (scroll down from top picture and ignore traffic signs) - however, high spatial resolution and background subtraction becomes an increasing concern. "Crowded Field Spectroscopy", analogous to the advances made with CCDs for globular cluster CMDs, has become a key interest of the PMAS project, with more possible applications for similar problems, e.g. spectroscopy of SNe, quasar host galaxies, and others.
Additionally, as 3D spectroscopy is insensitive to pointing uncertainties, we plan to experiment with beam switching techniques for the observation of faint optical counterparts of X-ray sources and other targets whose positions are only known to within a relatively large error-circle, and which are thus too easily missed with slit spectrographs.
Starting on May 28, three service nights were scheduled for the commissioning of PMAS at the 3.5m Telescope cassegrain focus. The main purpose of this run was to mechanically attach the instrument to the telescope, to align the optical components, to interface the instrument control and data acquisition systems to the telescope LAN and TCS, and finally to operate the whole system under real observing conditions.
Because of some concern related to the robustness of the CaF2 lenses of the fiber spectrograph, we had decided to completely dismantle the instrument in Potsdam and ship the sensitive optics in vibration-controlled boxes, requiring to essentially re-assemble the whole instrument again on Calar Alto. Therefore, a team consisting of 7 members had to arrive a week in advance to fully integrate the PMAS hardware and software under clean room environment in the aluminizing room in the basement of the 3.5m building (Jens Paschke, Volker Plank: mechanical engineering; Thomas Hahn: electronics; Thomas Fechner: CCD, LAN; Andreas Kelz: fiber optics; Emil Popow: integration and test; Martin Roth: team leader). The team was later complemented by PhD student Thomas Becker, author of the P3D PMAS Data Reduction Software, who also provided on short notice numerous observing support software tools for this run (see e.g. comment on telescope focus below).
After 5 days of intense work and trouble shooting, the instrument was completely operational in its test configuration and ready for system tests. The provisional configuration was limited in terms of performance since only a lab prototype fiber bundle was installed, and an engineering grade SITe ST002A with a severe hot pixel defect was used instead of the final science grade CCD. Otherwise, the instrument was fully functional.
On May 28, PMAS was lifted directly from the ground floor to the dome through the openings which are reserved for the mirrors to go to the aluminizing chamber. The moment when the instrument was attached to the telescope flange was a very exciting one for all of the PMAS team members, as well as the first motions of the telescope balancing routine - all of which went smoothly and almost routinely like with any standard CAHA instrument. Also the interfacing of the 3.5m TV guider system to the instrument presented no problems, neither did the interfacing of our computers to the telescope control system and LAN.
An important goal for PMAS is the ability to optimize the system for throughput and for homogeneity in terms of response variation from spectrum to spectrum (poor flat field calibration being a major limitation for faint object spectroscopy). Therefore, the PMAS IFU design allows for the exchange of broken fibers or fibers of low efficiency on-the-fly. Correspondingly, the optimal alignment of the lensarray foreoptics with the telescope, and the adjustment of the fiber bundle were among our primary objectives for this run. The first day was spent entirely for the adjustment procedure and a test series with dome flats in order to make sure that the micropupils were aligned with the fiber bundle.
On May 29, First Light was obtained with a series of short exposures, focussing on a bright star. A 1200 gr/mm grating was used (blazed at 500nm), which results in a dispersion of 0.45 A/pixel. A focus series script, connecting through the EPICS realtime database to the focussing mechanism of the telescope, was taking automatically exposures and measured the FWHM of star images from the resulting series of CCD frames. Such tools are standard for observers familiar with direct CCD imaging, but nonetheless non-trivial in the case of 3D spectroscopy: one first has to trace and extract spectra from the CCD frame, form a data cube, and create an image as a monochromatic slice from the cube (or the total from a stack of such slices) which is where finally a gaussian can be fitted to obtain a corresponding FWHM value. Fig. 2 shows a photograph of the original series, taken directly from the X-Terminal screen.
These observations, as well as the following ones on several standard stars and more instrumental tests, went extremely well. As a very encouraging outcome of these first on-line results we have realized that it is indeed possible to obtain high spatial resolution spectroscopy at the 3.5m Telescope. We have consistently measured seeing close to 1 arcsec FWHM throughout both nights with PMAS, which was in agreement with the CAHA seeing monitor readings. Moreover, there was essentially no time lost due to technical problems during this night and the following one.
As another test concerning spatially resolved spectroscopy, we have observed the galactic planetary nebula NGC6210, in order to compare with an HST image which is shown in the left panel (credit: STScI-PR98-36). Red: emission in [O II], green: [O III]. The white frame indicates the 8"x8" FOV of PMAS, which is sampled at a resolution of 0.5" per spatial element. North is up, East to the left.
The small panels to the right show three characteristic monochromatic
images at the wavelengths of Halpha 6563, [O I] 6301, and [S II] 6731
(from left to right).
While the recombination line image reflects the fact that this is
a high surface brightness, relatively young object where the central star
can hardly be discerned against the nebula (the brightest region to be
compared with the yellowish region in the HST image), the low ionization
frames clearly reveal the star (roughly 1" SW of the center), and more
structure to be compared with the reddish areas of the HST frame.
Note in particular the feature in the lower left corner which is
identified as a FLIER.
The partly patchy appearance of these monochromatic images is due to
the fact that the fiber bundle used for this test run was a lab prototype
which is far from ideal and exhibits several defects. The systematic
effect in the rightmost column is caused by the hot pixel CCD defect
which eventually lead to spilling charge into one third of the detector
area, making it impossible to take long exposures. Despite these shortcomings,
these and other test observations have demonstrated that the
instrument will be ready for science observations as soon as the provisional
subsystems have been replaced by the final ones (expected
for September 2001).
6 nights have been granted for the next PMAS run, starting October 23, 2001. These nights will be used for testing the science configuration with a science grade CCD and an optimized fiber bundle, and for our science verification programme. Later during the year, a mosaic consisting of 2x 2Kx4K CCDs is planned to replace the single chip, which will then double the wavelength coverage of the spectrograph. We will continue to produce fibers and optimize the fiber module, with an option to either increase the lensarray size, or to accomodate a twin configuration of two lensarrays, which would be advantageous for sky background subtraction in a beam switching mode. With the installation of a new lensarray, and replacing the 100um fibers with fibers of 50um core diameter, we a planning to reach the full design capacity of 1024 fibers of the fiber spectrograph (32x32 spatial elements).
The AIP intends to make PMAS available to the scientific community and offers support for interested observers. The P3d data reduction software package has been developed by Thomas Becker (AIP) prior to the installation of PMAS and was tested with real data from existing instruments like MPFS (Selentchuk), INTEGRAL (WHT), and SPIRAL (AAT). This software will be made available for future PMAS observers with the goal to reduce the data nearly on-line as part of the observing run on Calar Alto.
For more details concerning the design of PMAS, see list of
publications .
The PMAS Team is member of the OPTICON 3D Spectroscopy Working
Group which is intended to
bring together 3D expertise in Europe and to develop data format
standards and common software tools. This group is including 11 experienced
teams in Cambridge, Durham for the U.K., Leiden in the Netherlands,
Lyon, Marseille, and Paris in France, ESO, MPE-Garching and AIP in Germany,
Milan in Italy, and the IAC for Spain.
The OPTICON 3D Spectroscopy Working
Group Homepage is maintained at AIP.