Project STELES – SOAR Telescope Echelle Spectrograph
For technical details see:
INDEX
Abstract………………………………………………………………………………………2 1- Scientific Drivers 1.1 Introduction……………………………………………………………………………..3 1.2 Research fields that will benefit from STELES data…………………………….4 1.3 Access to the Instrument…………………………………………………………….7 1.4 Expected results……………………………………………………………………….7 2- Results transference methodology.……………..…………………………………..7 3- Project Methodology 3.1 Technical viability……………………………………………………………….…..…8 3.2 Instrument operation……………………………………………………….………….8 3.3 Maintenance plan………………………………………………………………………8 4- Management Approach 4.1 Construction……………………………………………………………………………9 4.2 Technical management……………………………………………………………….9 4.3 Follow up and documentation………………………………………………………9 4.4 Financial management………………………………………………………………10 4.5 Risk analysis………………………………………………………….……………….10 4.6 LNA infrastructure for production and assembling STELES………….……..11 5 Personnel 5.1 The instrument team and participants in the project……………………….…11 5.1 Potential users…………………………………………………………………….....12 6- Budget…………………………………………………………………………………..15 7- Simplified Schedule – indicating principal stages and estimate times……..16 8- Outreach………………………………………………………………………….…..…16 9- Bibliography 9.1 Technical publications……………………………………………………….……..16 9.2 Sample of science papers……………………………………………………....….17
Appendix A Equipment to be purchased in Brazil………………………………………...……….19 Equipment to be imported……………………………………………………………….19 Supplies and items to be purchased in Brazil…………………………….…..…….19 Supplies and items to be imported…………………………………………………....20 Third part labor in Brazil…………………………………………………………………22 Third part labor to be contracted outside Brazil………………………………....…22 Travel expenses…………………………………………………………….………..…...24 Daily expenses…………………………………………………………….…….…….….24
AppendixB Letter from SOAR Director approving STELES as a second generation instrument………………………..………………………….…………………….…....…24
ABSTRACT
High resolution spectroscopy has had an enormous scientific value for Brazilian astronomers, however, up to now we have not had access to an efficient instrument in a telescope with enough number of nights such as the 4.1 meter SOAR. For that purpose we intend to build the optical and mechanical modules of the STELES spectrograph, which the Brazilian community is committed to offer to the SOAR consortium. We developed a design with higher, wider spectral response and at lower cost throughput than instruments of the same class currently in operation. The project is already approved by the SOAR Board of Directors, 1/3 of the parts have been already purchased, and we plan to commission the instrument before 2010. The complex and diverse technologies of the instrument, such as fine mechanics, optics, electronics, cryogenics and software are interesting for the development of the national industry. In addition to burst the productivity of the Brazilian Astronomy, the instrument will have significant international visibility, since it will also be used by our partners from the USA and Chile.
Layout of STELES spectrograph, showing the red and the blue arms
1- Scientific Drivers
1.1 Introduction
High-resolution spectroscopy (HIRES) is one of the most powerful astronomical techniques. It enables determination of physical parameters such as temperature, gravity, rotation, radial motion and chemical composition. Abundances can be derived for stars in the Milky Way, in nearby galaxies and also in distant protogalactic clouds projected in the line-of-sight of quasars, probing the local and high-redshift Universe. Comparison between near and far universe gives information about its evolutionary phases, such as the first stellar generation, the epoch of re-ionization, and the formation of the galaxies.
The Brazilian astronomical community has intensively been using high-resolution spectroscopy. Before the 1990’s, the access to this technique was provided mainly through foreign facilities: ESO (European Southern Observatory) and CTIO (Cerro Tololo Interamerican Observatory), but after that, only projects in collaboration with researchers from the host institutions were possible, narrowing the access to the Brazilian community. In the period 1998-2002, under an agreement between ESO and ON (Observatório Nacional), Brazil rented the 1.5-m telescope at La Silla, enabling many projects to be developed. Currently, the access to HIRES instruments are through bHROS (optical range) and PHOENIX (near infrared) on Gemini South, but it is limited to a few nights per year for the whole Brazilian community. Therefore, we need urgently a HIRES facility on SOAR, where we have ~100 nights/year. The STELES spectrograph ( will reach objects as faint as V=18 (S/N=10 in one hour) with a spectral resolution R=50000, enabling science observations of objects hundreds of times fainter than what can currently be done in the main Brazilian facility, the Observatório Pico dos Dias (OPD). A set of projects that will take advantage of the proposed instrument is presented below.
The construction of STELES (SOAR Telescope Échelle Spectrograph) is already approved by the SOAR project as a second generation instrument, such that our community is committed to deliver it on time. STELES will be part of the set of instruments of the SOAR telescope, built by an international consortium between Brazil(CNPq), USA(NOAO, MichiganStateUniversity, University of North Carolina) and Chile(Conicyt). The expected high quality data are prone to produce important scientific results for us and the whole SOAR partnership, resulting in a good visibility to our technical capability. The development of a world-class instrument is also important to acquire a forefront instrumental knowledge.
The layout of STELES was developed in the period 2001-2006, as part of the Instituto do Milênio para Evolução de Estrelas e Galáxias na Era dos Grandes Telescópios: implementação de Instrumentação para os Telescópios SOAR e Gemini leaded by Beatriz Barbuy (IAG-USP). The layout is already reviewed and approved by the SOAR consortium. After that, the Millennium Institute bought the CCD arrays for the blue and the red arm, the controller and the cryogenic dewars, amounting to 1/3 of the total cost of the project.
Building instrumentation for a 4-m class telescope requires complex technologies involving engineering skills on fine mechanics, electronics, software, cryogenics and optics. All the different parts are developed simultaneously, around the core that is the optical design. The present project is concerned about the optics and mechanics of the instrument.
The Optical concept of STELES was designed by Bernard Delabre from ESO, one of the most prestigious optical engineers specialized in astronomical instrumentation. The project was conceived in an intense collaboration with Bruno Castilho and Clemens Gneiding from LNA/MCT. New achievements resulted in an instrument that is more efficient, cheaper and with higher throughput than similar instruments at work, such as FEROS, for example.
The optics of this kind of instrument cannot be ordered from commercial factories, but instead it has to be designed individually to achieve optimal performance. It is divided in 4 basic modular parts: a) the blue camera, b) the red camera, c) the fore-optics and d) the intermediate lenses and collimators. The mechanics is split into the fore-optics and the optical bench where the optical elements are mounted. In this project submitted to FAPESP we plan to buy the optical and mechanical parts and assemble the instrument. As in the previous phases of this instrument, the tight collaboration between IAG/USP and LNA/MCT is crucial for its successful completion. LNA infrastructure in optical metrology and characterization will be used to test and evaluate the optical components and verify if they are within the project specifications. LNA will be in charge of the optical assembly and alignment procedures, and the tools for these tasks are already purchased under FINEP and LNA projects.
1.2 Research fields that will benefit from STELES data
HIRES has a great impact in all fields of Astrophysics. STELES will combine high spectral resolution (6 Km/s/pixel), high stability, broad wavelength coverage (300-1000 nm), and good performance in the UV, and a fast, optimized data reduction software. Many existing projects in the Brazilian community will benefit from these capabilities and opportunities will be opened for new projects, which are currently impossible to be conducted. We list below a sample of such projects. Please notice that the references are intended to exemplify what the Brazilian community is doing, and so, is not exhaustive in any sense. Other papers from the same authors can be found in the web.
a- Abundance analyses in the near-UV: Key projects requiring good efficiency in the near-UV include spectroscopy of the strong electronic OH lines near 3100Å. These particular OH lines are detectable to very low metallicity and are useful in attempts to derive oxygen abundances in the oldest stars in the Galaxy: such O abundances probe the very earliest chemical evolution in the Milky Way. In addition to OH, the element beryllium is only detectable in the near-UV, via Be II lines near 3130Å. Beryllium is produced only through cosmic-ray spallation reactions and is a key probe in understanding cosmic-ray nucleosynthesis over the chemical evolutionary history of the Galaxy (Castilho et L. 1999, Cayrel et al. 2001).
b- Chemical Evolution of the Galaxy: The chemical evolution of the Galaxy follows the changes in elemental abundances from their initial values into the present compositions of the disk, bulge and halo. Some examples of specific topics related to this subject are the gradients of metallicity (Daflon and Cunha 2004) and radial and temporal variations of the star formation rate (Rocha-Pinto et al. 2000, Maciel et al. 2006).Also, detailed chemical composition analyses reveal measurable heterogeneities, possibly associated with age and position in the Galaxy, suggesting that local enrichment events may play a role in the overall galactic metal production (Castro et al. 1999). Concerning the major components of the Galaxy, the study of the bulge is essential to understand the mechanisms of formation of our Galaxy: the bulge may have been formed in the beginning of our Galaxy, although part of the bulge may have formed significantly later. The study of the oldest halo stars provides crucial lithium abundances that result from primordial Big Bang nucleosynthesis. Young stars in the Galactic disk trace the present distribution of chemical abundances and can be used to determine abundance gradients: these gradients provide constraints to Galactic models of star formation and chemical evolution (Barbuy et al. 2003).
c- Stars enriched by the r- and s-process: It has been suggested that one likely explanation for the highly r-process enhanced stars that have been identified recently is that they are members of a binary system with a massive companion that exploded as a type II supernova, which would now be a collapsed object such as a black hole. By monitoring radial velocities of the many examples of these stars we hope to find in the near future will be of extreme importance. The same applies to the metal-deficient stars that are moderately enhanced in their r-process elements (Barklem et al. 2005, Rossi et al. 2005). Additionally, s-process enriched stars such as barium stars have a complicated pattern of overabundance of s-process elements that has defied any straightforward interpretation. The well accepted model of wind accretion establishes that a trend of overabundance with the binary system period should be observed, which is not corroborated by observations. Even the nuclear reaction responsible for neutron production is controversial, since the preferred 13C (alpha, neutron) 16O reaction predicts a trend of both the overall overabundance and the neutron exposure parameter with metallicity, again directly contradicting the observations (Smiljanic, Porto de Mello, da Silva, 2006 A&A, accepted). Detailed analysis of large samples of these objects, involving many elements, is the only way to provide the necessary observational constraints to clarify these issues.
d- Light Element Abundances: The primordial abundance of light elements and their subsequent Galactic enrichment requires high-quality data to constrain and test their production and evolutionary models. One important problem in this subject is the intrinsic dispersion of Li abundances in dwarf stars of halo globular clusters. Given that globular clusters are among the oldest objects in the Galaxy, their initial Li abundance must be very close to the primordial value. Precise abundance determinations for Li (both 6Li and 7Li) and Be (with only one stable isotope, 9Be) provide essential information relevant to early Galactic cosmic-ray fusion and spallation nucleosynthesis, as well as primordial Big Bang Nucleosynthesis. Beryllium is an important addition to lithium, but Be is much more difficult to observe. The spectral regions containing the Be II (3130.41Å and 3131.06Å) and Be I lines (3312Å) are crowded and close to the atmospheric cutoff: a near-UV optimized spectrograph, such as STELES, would be an important addition to light element studies.
e- Cluster Analyses: The determination of accurate abundances in globular cluster stars over a range of magnitudes, covering effective temperatures from 4000 to 6000 K, can address the issue of possible variations in chemical composition existing among stars belonging to the same cluster. This topic is particularly relevant in investigating possible stellar processes, such as diffusion, dredge-up (Barbuy et al. 2006, Alves-Brito et al. 2005). The chemical distribution in sparse young clusters can be studied in significant samples of OB stars in OB associations (Daflon, Cunha and Buttler 2004). STELES will be able to investigate stars down to the main-sequence turn-off (V=17) in several clusters, combining high efficiency with a wide spectral range. Superb image quality, will also allow for spectroscopy in relatively crowded cluster fields.
f- Long-Term Velocity Monitoring of Carbon-Enriched Metal-Poor stars: A large fraction of the stars with [Fe/H] < -2.5 exhibit anomalously strong CH G-bands (and often C2 and CN features) indicative of very high carbon abundance, despite the stars overall low metallicity. It seems quite unlikely that all of the stars involved are members of close binaries that have undergone mass transfer of carbon-enriched material from their companions. Hence, sorting out which of the stars are radial velocity variables and which are not is an important program. Since the periods of known mass-transfer binaries can reach up to 7-8 years (or more, in some cases), gathering data for their study is a challenge with most telescopes. Correlations between measured abundances of (e.g., s-process) elements and orbital properties would provide valuable clues for understanding the range of phenomena and the nature of the progenitors in these systems (Lucatello et al. 2006, Rossi et al. 2005).
g- Asteroseismology of stars: Some structural properties of stellar interiors can be probed by the study of the weak non-radial pulsations that have been observed in objects all over the HR diagram, requiring simultaneous measurement of many spectral lines leads to the detection of multiperiodic oscillations. A large fraction of high mass stars with envelopes display non-radial pulsations, indicating that this is a relevant ingredient for mass ejection, in addition to the classical stellar winds. Asteroseismology is an important tool for stellar evolution studies (Floquet et al. 2002, Levenhagen et al. 2003).
h- Stellar winds: mass loss affects significantly the lifetime of high mass stars. In late evolutionary phases, the wind can be so dense as to preclude direct assessment of the underlying photosphere. Fortunately, presently there are computing codes able to derive parameters for the wind and the photosphere in a coherent way. They are especially useful to interpret spectra of O-type stars, Wolf-Rayets, Luminous Blue Variables and central stars of planetary Nebulae, taken at high resolution and encompassing broad wavelength range (Steiner and Damineli 2004). Long-term monitoring is crucial to unveil the nature of evolved massive stars (Damineli 1996) and their mass loss rates. Most of the known objects of this type can be reached from the Southern Hemisphere.
i- Stellar rotation: All stars rotate at some level, but this parameter was included in the models only very recently. Predictions of luminosity, evolutionary tracks and times, chemical yelds are different from the case of non-rotating stars. There are still few observational tests for such models, but recently it was confirmed that the angular momentum migrates to the surface as the star approach the exhaustion of H in the core (Groh et al. 2006). This result has a strong impact on massive stars, since it affects the understanding of mass loss rate, bipolar flows and hypernovae explosion. The discovery was done with spectrographs similar to STELES. Studies on other fast B-type rotators have been done by (Vinicius et al. 2006) and the induced mixing has been tested for a sample of stars in the Galaxy and in the LMC (Daflon et al. 2001, Korn et al. 2005).
j- Young stars: Many parameters from Pre-Main Sequence (PMS) stars can be derived from HIRES, such as chromospheric activity (Vieira et al. 1999), Lithium abundance, rotation, binarity (Pogoding et al. 2006), magnetic fields (Pogodin et al. 2005) and cluster motion. Nearby young associations can indicate how the local Interstellar Medium evolved along the last 100 million years. On the other hand, that is the age at which we expect planets being formed around those stars. Determining the rotational velocity of those stars can give information about the evolution of rotation along the latest PMS phases and planetary formation (Torres et al. 2006). Abundances and mettalicities have been derived, based on spectral synthesis (G. Rojas, J. Gregorio-Hetem 2005), of young stars that are candidates to present debris disks (Gregorio-Hetem & Hetem 2002). One of our main interest is to characterize the stars which will be prime targets of NICI imaging as contribution from our team to the NICI planet campaign. Our projects are also related with the study of chemical abundance of the Galaxy and the analysis of stellar clusters.
k- Doppler tomography of accretion disks: Monitoring the variability of emission line profiles can be used to map discs and motion of circumstellar gas flows in binary systems with mass transfer. Although those systems, as seen from Earth, span only a few micro-arcsecond, their spatial structures can be resolved in detail with this technique (Diaz and Ribeiro 2003). The stochastic nature of the emission in lines can be analyzed by using large number of spectra with fine spectral and temporal sampling (Diaz 2001). This technique can be applied to many types of objects, like Cataclismic Variables with discs, Algol binaries, Classical T Tauri and Herbig Ae/Be.
l- Absolute dimension determination and spectral disentangling: the absolute dimensions of stars, mainly of masses and radii, but also effective temperatures, are the most important parameters in controlling the models for stellar formation, structure and evolution. The tendency of hierarchical clustering is revealing that more and more eclipsing binaries are in fact triples or quadruples (Casey et al. 1998). Many systems, especially the ones with young components, can only be analyzed through spectral disentangling techniques. As a by-product, the disentangled spectra are fully adequate for abundance studies. High-precision (S/N) and high-resolution spectroscopy are the desired qualities of the spectroscopic data for these tasks (Torres and Vaz 2006).