Need help understanding stellar spectroscopy data from ESO

Need help understanding stellar spectroscopy data from ESO

We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

The European Southern Observatory webpage has a web page that has tabular spectrogram data from A. J. Pickles, University of Hawaii. There are over 130 .dat files there. Each one represents a spectrogram from a star.

My question is at the bottom.

First though, I will excerpt some lines from a few of the .dat files. You can read the tabular data by unzipping the ".dat.gz" files and opening the resulting ".dat" in a plain text editor.


Description of File Name:

  • UK = They all start with UK
  • W = Weak metalicity (can be: W for weak, S for strong, or nothing for normal)
  • G = Spectral class G (can be: O, B, A, F, G, K, M)
  • 0 = Spectral subtype 0 (can be: 0-9)
  • V = Roman numeral 5 Yerkes luminosity class (can be: Ia, Ib, II, III, IV, V, VI, VII)

Excerpt of File Contents:

#iRMS=9.222171502e-05 0 # lk ukf_wg0v uks_wg0v fh fse # 1150.0 0.004218 0.000000 0.004218 0.000000 1155.0 0.002700 0.000000 0.002700 0.000000 1160.0 0.001559 0.000000 0.001559 0.000000… [4,765 lines truncated]… 24990.0 0.016113 0.000000 0.000000 0.016113 24995.0 0.016087 0.000000 0.000000 0.016087 25000.0 0.000000 0.000000 0.000000 0.000000


#iRMS=1.052067091e-05 0 # lk ukf_g0i uks_g0i fh fse fk # 1150.0 0.000000 0.000000 0.000000 0.000000 0.000000 1155.0 0.000000 0.000000 0.000000 0.000000 0.000000 1160.0 0.000000 0.000000 0.000000 0.000000 0.000000… [4,765 lines truncated]… 24990.0 0.000000 0.000000 0.000000 0.000000 0.000000 24995.0 0.000000 0.000000 0.000000 0.000000 0.000000 25000.0 0.000000 0.000000 0.000000 0.000000 0.000000


#lRMS=0.06043180451 iRMS=0.04044797271 0 # lk ukf_m3iii uks_m3iii fh fl fd fm fsv # 1150.0 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1155.0 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 1160.0 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000… [4,765 lines truncated]… 24990.0 0.302861 0.000000 0.000000 0.000000 0.000000 0.323665 0.000000 24995.0 0.301125 0.000000 0.000000 0.000000 0.000000 0.321810 0.000000 25000.0 0.300519 0.000000 0.000000 0.000000 0.000000 0.321163 0.000000


#lRMS=0.0004083506647 iRMS=0.0004083506938 0 # lk ukf_o5v uks_o5v fh fse fl # 1150.0 12.354247 10.980931 12.354247 0.000000 0.000000 1155.0 9.928268 8.291242 9.928268 0.000000 0.000000 1160.0 8.410753 7.608442 8.410753 0.000000 0.000000… [4,765 lines truncated]… 24990.0 0.000000 0.000620 0.000000 0.000000 0.000000 24995.0 0.000000 0.001047 0.000000 0.000000 0.000000 25000.0 0.000000 0.001561 0.000000 0.000000 0.000000

In the accompanying paper, it helps to decode the columns:

  • 1st column: The wavelength in angstroms. The rows start at 1150 Å and increment every 5 Å up to 25000 Å. This can be seen across the X axis of the plots. (I don't know why whoever made these plots decided to start them at 8000 Å, but the actual data begins at 1150 Å.)

  • 2nd column: The corresponding flux at that wavelength. Fluxes are in Fλ (F_lambda) units. This is the Y axis of the plots.

  • 3rd column: "The standard deviation of the optical combination… The normalized source contributions are averages where multiple components from a source exist. The first line summarizes RMS errors of these individual averages. The standard deviation column is generally zero in wavelength regions where only one source is present, but is sometimes replaced by the standard deviation of the average forming that single component."

  • Subsequent columns: The source contributions. Sometimes there are multiple spectra surveys that took measurements at a particular wavelength. The average (arithmetic mean) of these sources appear in column 2. The standard deviation of these sources appear in column 3.


I thought that the Y axis of a spectrogram was supposed to represent intensity of radiation, but in this data they refer to it as flux and are measuring it in terms of "Fλ units". I am very dumb with regard to spectroscopy. Could somebody please explain to me how flux compares to intensity, and what exactly Fλ units are?

Spectroscopy: The Key to the Stars

More can be learned about physical processes going on in stars and nebulae by understanding and analyzing their spectra than by any other means.

Many amateur astronomers who use CCD cameras are taking up spectroscopy as part of their observational program, but until now the physics that underlies astronomical spectroscopy has been confined to advanced academic books.

In Spectroscopy – the Key to the Stars, Keith Robinson describes the physics and physical processes that cause the stellar spectra to be as they are… spectra that amateur astronomers can image with today’s commercially-made equipment. Written specifically for amateur astronomers, this book assumes only a basic knowledge of physics but looks in detail at many topics, including energy levels in atoms, the molecular spectra of red stars, emission lines in nebulae, and much, much more.

Here is everything you need to know about how the atomic processes in stars and nebulae produce the spectra that amateur astronomers can image, and why spectroscopy is such a powerful tool for astronomers.

Keith Robinson obtained a degree in physics from the University of Lancaster, and is a Fellow of the Royal Astronomical Society.

"If you ever wondered what the big deal is about spectroscopy or wished you understood it a little better, this book’s for you. Robinson takes a step-by-step approach to spectroscopy, each chapter building on the ones before it. … The book is a worthy addition to any advanced amateur astronomer’s library." (Michael Bakich, Astronomy, February, 2007)

"In this informative monograph, Robinson (Royal Astronomical Society) explains the basic concepts in terms that a general reader can master. Topics such as the characteristic radiation expected to be emitted by atoms, by ionized gas, and by molecules are addressed using illustrations and word descriptions of the physical processes. … the interested reader will find this book a stimulating introduction. Summing Up: Recommended. General readers lower-division undergraduates." (D. E. Hogg, CHOICE, Vol. 44 (11), July, 2007)

"In Spectroscopy: The Key to the Stars, Keith Robinson makes spectroscopy approachable for those who are interested in expanding their observational repertoire. … Not only is this a good read for any observer thinking of taking up spectroscopy, but it’s also suitable for high school or first-year college students in astronomy and physics." (Carolyn Collins Petersen, Sky & Telescope, Vol. 115 (1), January, 2008)

"This is a small book (160 pages) written for amateur astronomers who use CCD cameras and include spectroscopy as part of their observational program. The main purpose of the book is to describe the physics and the physical processes behind the stellar spectra. … the topics considered are clearly and concisely described. The amateur astronomers, who are not familiar with physics or who have forgotten the essentials of this science, will read it … with interest and pleasure." (Emile Biemont, Physicalia Magazine, Vol. 29 (4), 2007)

Need help understanding stellar spectroscopy data from ESO - Astronomy

An Introduction to Stellar Astrophysics
author: Francis LeBlanc
2010, Wiley, paperback edition

Homework Sets: 60% of grade
Hour tests (3 of them): 40%
Final Exam: 0%
(last Hour Test will be during the final exam period)
Students with special needs may request appropriate
accommodation call UT's office of Services for
Students with Disabilities, 471-6259.

Not required, but a good idea for practical training

Subject Matter, Goals, and Miscellaneous Comments

What is it? To whom am I speaking? Astronomy 352K is a junior/senior-level introduction to stellar astronomy and astrophysics, with emphasis on observational and empirical methods for studying stars via the light they emit. It is designed with upperdivision astronomy majors in mind, but it is also suitable for students majoring in closely related fields such as physics, mathematics, or engineering. See an additional remark in the textbook paragraph below.

Prerequisites? I expect you to have taken Physics 316 (Electricity & Magnetism) and its associated lab course Physics 116L, which have as their prerequisites Physics 301 (Mechanics) and 101L, and relevant math courses. It is acceptable to have taken instead the Engineering Physics courses 303K and 303L, with their lab courses. Astronomy draws on such a wide variety of areas in physics that we cannot expect you to have prior preparation in all of them, and so we will introduce physical ideas and laws as needed. (Examples include the theory of radiation, atomic structure, and statistical mechanics.) We will usually be interested mainly in applying physical principles, rather than in deep and lengthy derivations. In general the mathematical manipulations expected of you (e.g., on homework sets) will be pretty straightforward.

Background? We do not assume that you have strong (indeed, any!) previous background in astronomy, although many of the students already will have taken other upperdivision astronomy courses or at least had an introductory astronomy course such as AST 307 or 301. If you find that there are gaps in your basic astronomical knowledge, please ask me or the TA to explain or elaborate (either in class or during office hours). You might also find it helpful to consult one of the many fine introductory textbooks that are widely available (I can lend you one of them). It should take you only a few evenings to master all of the relevant material that is contained in these books.

Overlap with other courses? There is a small amount of overlap between AST 352K and AST 358 (Galaxies and the Universe), AST 353 (Astrophysics), and AST 352L (Positional, Kinematical, and Dynamical Astronomy). We will try to avoid excessive redundancy, but that is inevitable in some subject areas, since not all members of the present class will have taken these other courses. Note that the department will be offerring AST 353 for Spring 2012. I intend at least part of AST 352K to lead pretty directly into AST 353.

Textbook? We will be using Francis LeBlanc's An Introduction to Stellar Astrophysics. This book is fairly new The book covers material appropriate to both this course and AST 353 (the parts that include discussion of stellar interiors and nuclear energy generation). It is likely the AST 353 instructor will adopt the same text. This is a nice single source for basic information about stars, and I like its presumption that many readers will have had no prior introduction to astronomy in any detail.

Class notes? I also am posting copies of my notes on the class web site. I have alternated teaching this course with Prof. Harriet Dinerstein, and between us we have pretty much settled on the topics and presentation order that we like. Harriet made a major upgrade in the class notes a few years ago, and the current version has a lot of her ideas in it. You will easily notice that my presentations will alternate between working through these notes and trying to resonate with the textbook. There are sections of the course that will be "all book", or "all notes", or a mixture.

My bias in this course? I regard AST 352K as a vital link between the basic, often elegant physics and mathematics that you have ingested at UT for the past two-three years, and the real, often messy world of astronomical research. Astrophysics combines elements from all areas of physics to offer coherent theoretical models for how the solar system, galaxy, and universe are constructed and how they have and will evolve. If you are looking for that in this course, forget it. Theoretical astrophysics cannot really derive rational models for an object without appealing to observational astronomy. I am not an astrophysicist as the term is sometimes meant (that is, I am not a theorist). I am an observational astronomer, and proud of it. And observational astronomy is what you will find covered in this course. Not how is the universe constructed, but how does one practically assemble the basic data about particular astronomical objects (stars) that can be gainfully used in constructing the story of the universe?

So forget cookbook problems? Not entirely, but we will deal as much as possible with real data from the literature that have been obtained at various astronomical facilities over the past decades, and which now are readily available for study. Many of the homework problems will encourage you to seek data from basic astronomical catalogs. These sources can be found on appropriate web sites.

Homework, and your approach to it? The homework sets are the keys to what I want to get done in this course. I intend to be deliberately vague in some of the assignments. I admit that this is a sometimes maddening ploy. Real astronomical research usually does not admit cookbook solutions to interesting problems, and I want you to get used to that. Such an approach is also a signal on my part that I encourage interactions with you outside class. Feel free to discuss with me the course material, problem sets, or any other astronomical topic that come to mind. On the first page I give the formal office hours, but you of course may set up appointments with me at other rational times of day. Notice also that I give my work number, email address, and my home phone number. I greatly prefer contact in person or a phone call emails to me can get deeply buried. I would not give out my home phone number if I did not expect calls in the evening whenever you need to. (It is true that we are the only Snedens in the Austin telephone directory.) I want to help you do well in this course, but I need you to make contact! Don't be shy: remember that the truly stupid question is the unasked one.

Caveat Emptor! Two warnings must be given, one applicable to all faculty members here and one specific to me. First, in this department you deal with professional astronomers. The good part is that you get very close to current research, and that can be very exciting. The bad part is that we tend to travel a lot (most obviously to observatories in remote and exotic locales), and I will need to excuse myself from class a couple of times during the semester. At present I have trips scheduled for September 15-18 (no classes missed), and October 7-17 (missing class on October 11 & 13). A substitute lecturer will pinch-hit for me in class on those occasions. All class meetings will occur as scheduled. Second, I am currently Letters Editor of The Astrophysical Journal. This means that inevitably I am pulled in many different teaching/service/research directions simultaneously, and frankly I am very busy. However, this should not become your problem! I expect you to work in this class, and you should expect no less of me. Do not feel the slightest hesitation in pushing me to make time for you outside of class politely in the beginning, but more firmly if I do not respond well. Your interaction in this course can only aid your understanding.

Finally, a Carrot Instead of a Stick: I formally teach only one semester each year (because of my editing duties). In spring 2012 (an "off" semester) I probably will be offering a guided research position to an interested student who has successfully (A-B grade) completed AST 352K this semester. Usually the student will sign up for some course number that signifies independent study. Real investigation into the chemical composition of stars will occur! Some background buildup will be necessary, but then collaborative research can commence. A grade for that course will be the least interesting outcome, as successful work (often going beyond the spring semester) should lead to a published paper and/or attendance at a professional meeting.

Need help understanding stellar spectroscopy data from ESO - Astronomy

We share the Universe with hundreds of billions of other galaxies, the very first of which formed when the Universe was still very young.

The ELT will explore these early galaxies in exquisite detail and follow their evolution through cosmic time. This will help us understand how galaxies form and evolve, providing a key contribution to our knowledge of the Universe.

The ELT will explore these early galaxies in exquisite detail and follow their evolution through cosmic time. This will help us understand how galaxies form and evolve, providing a key contribution to our knowledge of the Universe.

We share the Universe with hundreds of billions of other galaxies, the very first of which formed when the Universe was still very young.

The ELT will explore these early galaxies in exquisite detail and follow their evolution through cosmic time. This will help us understand how galaxies form and evolve, providing a key contribution to our knowledge of the Universe.

Very distant galaxies are so far away their light takes billions of years to reach us. With modern telescopes, astronomers can travel back in time to reveal these extremely distant, and therefore very young, galaxies as they were in the early Universe. There are limits to how far back we can look with current telescopes. After the Big Bang, the Universe grew cool and dark. Only once dark matter and gas had clumped together did stars and galaxies begin to shine. Energetic radiation from those galaxies then ionised the remaining neutral gas, but exactly when and how this all occurred remains one of the biggest mysteries in cosmology. As an enormous telescope equipped with the most advanced instruments, the ELT will be able to observe these first galaxies and revolutionise our perception of the Universe.

When astronomers use current telescopes to study light from far-away galaxies, they observe the combined light emitted by all the individual stars. However, to make significant progress in our understanding of the evolution of the Universe, we need to look at these stars individually. For galaxies in our Local Group and beyond the ELT will be able to resolve individual stars, allowing us to perform a sort of galactic archaeology, using the stellar fossil records to decipher how galaxies formed and evolved.

The ELT will also contribute to a better understanding of the complex interplay between galaxies and the surrounding intergalactic medium (IGM). The IGM provides the reservoir of gas for the ongoing infall of fresh material into galaxies. At the same time, it acts as a repository for the gas driven out of galaxies by energetic processes such as active galactic nuclei and supernova explosions.

The ELT will observe distant galaxies to gather clues on how these buildings blocks of our Universe formed and evolved. It will also search the early Universe to discover the very first galaxies.

It’s not just empty space in between galaxies: the gas in the intergalactic medium contains most of the visible matter in the Universe and is key to the process of galaxy formation and evolution. The ELT will help us study its properties.

How did galaxies form and evolve?

Galaxy formation and evolution is driven by a number of mechanisms and a complex interplay between them. These include the hierarchical merging of dark matter halos (galaxies with smaller dark matter halos are believed to have formed first and merged to form galaxies within larger halos), the accretion and cooling of gas, and the gravitational fragmentation and the formation of molecular clouds. Star formation, nucleosynthesis, metal-enriched outflows that are driven by stellar winds, supernova explosions and energetic output from accretion onto supermassive black holes also influence galaxy formation and evolution.

Thanks to recent deep surveys, it has become easier to observe properties of galaxies such as their luminosity, size, star formation rate, mass — and the mass of the stars within them, as well as how they evolve with redshift. But we now need more detailed spatially and spectrally resolved studies of individual galaxies to improve our understanding of their underlying physical processes. A comprehensive picture of galaxy formation should also describe a galaxy’s internal structure, for example, how stellar populations and dust are distributed or what properties its central region has. Current facilities, however, are not able to describe these internal properties in detail in far-away galaxies as they cannot reach the required small spatial scales. We need to get closer to the scale on which star formation occurs, which is that of giant molecular clouds the ELT will enable us to do just this.

The Hubble Space Telescope has extended our view of evolving galaxies from the optical to the near-infrared, probing the emission from many stars in galaxies at high redshift and resolving the shapes and colours of galaxies. At the same time, integral field spectrometers, such as those on ESO’s Very Large Telescope, have routinely mapped the motion of ionised gas and the physical conditions of galaxies. The ELT will allow astronomers to observe fainter and more distant galaxies, up to when they were first building up their stellar populations and gas reservoirs, so studying them in more detail would much improve our current picture of galaxy evolution.

Characterising galaxies with a wide range of masses, star formation activities, and surroundings requires a giant leap in spatial resolution. This is uniquely provided by the ELT, with its 39-metre main mirror. and diffraction limit of 13 mas (

100pc at z > 1) in the K band (λ

2μm). Both imaging and integral field spectroscopy (with the MICADO and HARMONI instruments) will enable deep views of galaxies across cosmic times that are

15 times sharper than is currently possible with Hubble, and

6 times sharper than possible with the James Webb Space Telescope. HARMONI will overcome present limitations by tracing structures just a few tens of parsecs wide, and providing detailed studies by exploiting spectroscopic information with high angular and spectral resolution and sensitivity.

Gain in resolution with MICADO: real data from HST/NIC2 compared to three different simulations from ISAAC, NIC2 and MICADO. Credit: MICADO Consortium

Some 380,000 years after the Big Bang, the temperature of the Universe was low enough that the hydrogen-dominated IGM, which pervades space, became neutral. The IGM today, however, is fully ionised, heated by the ultraviolet emission from galaxies and active galactic nuclei. However, how and when the IGM turned from neutral to fully ionised remains a matter of great debate.

The prime sources of reionisation have remained elusive so far and the quest for such sources, which produced the ultraviolet radiation that reionised the IGM, is closely related to the search for the first, most distant galaxies. Ultra-deep observations of faint sources with the ELT’s HARMONI and MOSAIC instruments will allow astronomers to determine the ionisation state of the IGM at redshifts from 5 to 13. In other words, the ELT will push back the limits of the observable Universe to the end of the Dark Ages, when the first light-emitting objects, which ionised much of the content of the Universe, switched on.

Scientists will be able to use MOSAIC to study Lyman α galaxies at high redshift. Credit: MOSAIC Consortium

The gas in the intergalactic medium is revealed by the numerous hydrogen absorption lines that are seen in the spectra of quasars. Astronomers have seen that at high redshift the IGM contains most of the baryons in the Universe and is therefore the baryonic reservoir for galaxy formation. In turn, galaxies emit ionising photons and expel metals and energy through powerful winds that determine the physical state of the gas in the IGM.

This interplay of galaxies and gas in the IGM is central to galaxy formation, but it is complex, and happens on scales of the order of just one megaparsec. The ELT’s MOSAIC spectrograph will be able to explore the distant Universe (at z

2.5) in three dimensions to study the IGM structure and its chemical properties, and to look for correlations between the position of galaxies and gas density peaks in the IGM.

A spectroscopic survey with MOSAIC will enable the mapping of the IGM in the distant, young Universe. Credit: C. Laigle, J. Japelj

The absorption lines in the spectra of quasars arise from moderate density fluctuations in a warm, photo-ionised IGM. The spatial distribution of the IGM is related to the distribution of dark matter. The full density field can be reconstructed using a grid of lines-of-sight, but this requires at least about 900 randomly distributed targets per square degree. At a redshift of 2.5 there are

900 Lyman Break Galaxies per square degree and around 50 (brighter) quasars. With MOSAIC, the ELT will be able to use these beacons to map the structure of the IGM at redshifts between 2.2 and 2.4 with moderate spectral resolution.

The ELT’s HIRES instrument will also study directly the IGM. One exciting prospect is detecting elements synthesised by the first stars in the Universe — the massive, metal-free Population III stars. The ability of HIRES to obtain high-signal-to-noise, high resolution spectroscopy for bright (AB mag ≤ 21) quasars will allow astronomers to study absorption lines too weak for current telescopes, resulting in an order-of-magnitude increase in the sample of useful background sources at high redshifts.

Surveys, Catalogues, Databases, and Archives of Astronomical Data

5.2.3 Spectral Photographic Surveys

In the early 20th century the photographic surveys were carried out not only for catalogues of star positions but for spectral determination also. The first star photographic spectra applicable for scientific purposes were obtained by Henry Draper in the late 19th century with an 11 inch Draper instrument of Harvard College Observatory (HCO) equipped with the objective prism cell which contained two quartz prisms. Later the director of the HCO E. C. Pickering made efforts to equip several instruments with objective prisms and plate holders in order to make photographs of star fields with spectra (see Fig. 5.6 ). Thanks to him, the extensive survey of stellar spectra over both hemispheres to 9th visual magnitude began ( MacConnell, 1995 ).

Fig. 5.6 . A Bache telescope objective prism photograph used for the Henry Draper Catalogue, showing spectra of stars in Carina, recorded May 1893. The exposure time was 140 minutes ( Hearnshaw, 2009 ). Reproduced with permission of the Licensor through PLSclear.

All the plates were processed manually. All the objects registered in the frames were classified and occupied their places in the star hierarchy according to specific features of their spectra. That tremendous job resulted in the Henry Draper Catalogue (HD) of stellar spectra and the Harvard system of spectral classification of stars depending on their stellar temperatures . The HD was published in 1918–1924. It contains 225,000 stars extending down to 9 m . Altogether more than 390,000 stars were classified at Harvard.

Spectral photographic observations with objective prism remain actual until now but their goals transformed from the comprehensive classification of objects in star fields into the surveys of specific groups of astronomical objects. The number of objects fixed on photographic plates increased to tens of millions. The manual processing of observations is history. The extraction of spectra is performed on digitized images in automatic or semiautomatic modes.

In the second half of the 20th century, the First Byurakan Survey (FBS) or Markarian Survey, 1965–1980, covering 17,000 deg 2 ( Markarian et al., 1989 ) and its continuation, the Second Byurakan Survey (SBS), 1978–1991, 965 deg 2 , covering some 3,000,000 low-dispersion spectra ( Markaryan et al., 1983 Stepanian, 2005 ), are the largest photographic spectral surveys in the Northern sky initially targeted on a search of galaxies with ultraviolet excess. Both have been carried out with the Byurakan Observatory 102/132/213 cm Schmidt telescope using an objective prism. The spectra on the plates are about 1 mm long. Objects with 13 m < B < 16.5 m have useful spectra. Each FBS plate contains low-dispersion spectra of some 15,000–20,000 objects, so there are about 20 million objects in the whole survey ( Nesci et al., 2009 ). The plate collection of FBS was digitized using the commercial scanner EPSON 1680 Pro ( Fig. 5.7 ). Yet one survey related to the objective prism spectroscopy is Case Low-Dispersion Northern Sky Survey ( Pesch et al., 1995 ) see also the updated versions of Markarian galaxies (Mrk), catalogues by Mazzarella and Balzano (1986) , Petrosian et al. (2007) .

Fig. 5.7 . A piece of the digitized FBS image showing its low-dispersion spectra ( Mickaelian et al., 2009 ). The figure is published under the License CC BY 4.0 .

Another two wide-angle Hamburg surveys for the Northern and Southern sky were targeted onto quasar detection and investigation. The Hamburg Quasar Survey (HQS) covers 14,000 deg 2 in the Northern sky (1985–1997, Engels et al., 1988 Hagen et al., 1999 ). The next one is its enhancement to the Southern sky. The Hamburg-ESO Survey (HES) is based on digitized objective-prism photographs taken with the ESO Schmidt telescope, covering essentially the entire Southern extragalactic sky ( Wisotzki et al., 2000 ) (1990–1996). Targets of this survey are bright QSOs and Seyfert galaxies. It covers 9000 deg 2 in the Southern sky with the magnitude range 13 < B < 18 . All spectroscopic plates of both collections were digitized using the Hamburg PDS 1010G microdensitometer with further extraction of spectra by application of the specially designed numerical algorithm.

Among the extragalactic catalogues created at the same time we mention the three Reference Catalogues of Bright Galaxies (RC1, RC2, RC3) by de Vaucouleurs et al. (1964, 1976, 1991) Southern Galaxy Catalogue (SGC) ( Corwin et al., 1985 ) A Revised Shapley–Ames Catalogue of Bright Galaxies (Rev SA) ( Sandage and Tammann, 1987 ), and Catalogue of Principal Galaxies (PCG) by Paturel (1989) . So, a revised version of the Shapley–Ames catalogue contained data on galaxies brighter than 13.5 m , including its redshifts. This catalogue and the compilation of all available data on bright galaxies by de Vaucouleurs catalogues became the sources of distances which made it possible to obtain the first three-dimensional distributions of galaxies in the universe. Much more detailed information on the spatial distribution of galaxies was obtained on the basis of redshifts, measured at the Harvard Center for Astrophysics (CfA catalogue) for all Zwicky galaxies brighter than m p h = 14.5 . Later this survey was extended to galaxies brighter than m p h = 15.5 (the Second CfA catalogue) and to galaxies of the Southern sky (Southern Sky Redshift Survey) see A survey of galaxy redshifts (z-catalogue) by Huchra et al. (1983) . At this time the first catalogues of superclusters of galaxies were prepared: Superclusters by Oort (1983) , A supercluster catalogue by Bahcall (1983) , and All-sky catalogs of superclusters of Abell-Aco clusters by Zucca et al. (1993) .

But the modern era of galaxy redshift catalogues started with the Las Campanas Redshift Survey (LCRS). Here, for the first time, multiobject spectrographs were used to measure simultaneously redshifts of 50–120 galaxies ( Shectman et al., 1996 ). The LCRS covers six slices of size 1.5 ∘ × 80 ∘ , the total number of galaxies with redshifts is ∼26,000, and the limiting magnitude is 18.8 m . “Generally the construction of a redshift survey involves two phases: first the selected area of the sky is imaged with a wide-field telescope, then galaxies brighter than a defined limit are selected from the resulting images as nonpointlike objects optionally, colour selection may also be used to assist discrimination between stars and galaxies. Secondly, the selected galaxies are observed by spectroscopy, most commonly at visible wavelengths, to measure the wavelengths of prominent spectral lines comparing observed and laboratory wavelengths then gives the redshift for each galaxy” (cited by “SkyServer – Algorithms”).

These first redshift databases and catalogues of galaxy clusters allowed for the discovery of the three-dimensional cell structure of the universe (filamentary distribution of galaxies and galaxy clusters form superclusters, while there are huge voids without galaxies). These results demonstrated that “the pancake scenario of structure formation by Zeldovich et al. (1982) fits observations better than the hierarchical clustering scenario by Peebles (1980) . More detailed studies of the structure formation by numerical simulations showed that the original pancake scenario by Zeldovich also has weak points – there is no fine structure in large voids between superclusters observed in the real universe ( Zeldovich et al., 1982 ) and the structure forms too late, thus a new scenario of structure formation was suggested based on the dominating role of the cold dark matter in structure evolution. In a sense the new scenario is a hybrid between the original Peebles and Zeldovich scenarios: structure forms by hierarchical clustering of small structures within large filamentary structures – superclusters” ( Einasto, 2001 ).

18.1 A Stellar Census

By the end of this section, you will be able to:

  • Explain why the stars visible to the unaided eye are not typical
  • Describe the distribution of stellar masses found close to the Sun

Before we can make our own survey, we need to agree on a unit of distance appropriate to the objects we are studying. The stars are all so far away that kilometers (and even astronomical units) would be very cumbersome to use so—as discussed in Science and the Universe: A Brief Tour—astronomers use a much larger “measuring stick” called the light-year. A light-year is the distance that light (the fastest signal we know) travels in 1 year. Since light covers an astounding 300,000 kilometers per second, and since there are a lot of seconds in 1 year, a light-year is a very large quantity: 9.5 trillion (9.5 × 10 12 ) kilometers to be exact. (Bear in mind that the light-year is a unit of distance even though the term year appears in it.) If you drove at the legal US speed limit without stopping for food or rest, you would not arrive at the end of a light-year in space until roughly 12 million years had passed. And the closest star is more than 4 light-years away.

Notice that we have not yet said much about how such enormous distances can be measured. That is a complicated question, to which we will return in Celestial Distances. For now, let us assume that distances have been measured for stars in our cosmic vicinity so that we can proceed with our census.

Small Is Beautiful—Or at Least More Common

When we do a census of people in the United States, we count the inhabitants by neighborhood. We can try the same approach for our stellar census and begin with our own immediate neighborhood. As we shall see, we run into two problems—just as we do with a census of human beings. First, it is hard to be sure we have counted all the inhabitants second, our local neighborhood may not contain all possible types of people.

Table shows an estimate of the number of stars of each spectral type 1 in our own local neighborhood—within 21 light-years of the Sun. (The Milky Way Galaxy, in which we live, is about 100,000 light-years in diameter, so this figure really applies to a very local neighborhood, one that contains a tiny fraction of all the billions of stars in the Milky Way.) You can see that there are many more low-luminosity (and hence low mass) stars than high-luminosity ones. Only three of the stars in our local neighborhood (one F type and two A types) are significantly more luminous and more massive than the Sun. This is truly a case where small triumphs over large—at least in terms of numbers. The Sun is more massive than the vast majority of stars in our vicinity.

Stars within 21 Light-Years of the Sun
Spectral Type Number of Stars
A 2
F 1
G 7
K 17
M 94
White dwarfs 8
Brown dwarfs 33

This table is based on data published through 2015, and it is likely that more faint objects remain to be discovered (see Figure 1 ). Along with the L and T brown dwarfs already observed in our neighborhood, astronomers expect to find perhaps hundreds of additional T dwarfs. Many of these are likely to be even cooler than the coolest currently known T dwarf. The reason the lowest-mass dwarfs are so hard to find is that they put out very little light—ten thousand to a million times less light than the Sun. Only recently has our technology progressed to the point that we can detect these dim, cool objects.

Dwarf Simulation.

Figure 1. This computer simulation shows the stars in our neighborhood as they would be seen from a distance of 30 light-years away. The Sun is in the center. All the brown dwarfs are circled those found earlier are circled in blue, the ones found recently with the WISE infrared telescope in space (whose scientists put this diagram together) are circled in red. The common M stars, which are red and faint, are made to look brighter than they really would be so that you can see them in the simulation. Note that luminous hot stars like our Sun are very rare. (credit: modification of work by NASA/ JPL-Caltech)

To put all this in perspective, we note that even though the stars counted in the table are our closest neighbors, you can’t just look up at the night sky and see them without a telescope stars fainter than the Sun cannot be seen with the unaided eye unless they are very nearby. For example, stars with luminosities ranging from 1/100 to 1/10,000 the luminosity of the Sun (LSun) are very common, but a star with a luminosity of 1/100 LSun would have to be within 5 light-years to be visible to the naked eye—and only three stars (all in one system) are this close to us. The nearest of these three stars, Proxima Centauri , still cannot be seen without a telescope because it has such a low luminosity.

Astronomers are working hard these days to complete the census of our local neighborhood by finding our faintest neighbors. Recent discoveries of nearby stars have relied heavily upon infrared telescopes that are able to find these many cool, low-mass stars. You should expect the number of known stars within 21 light-years of the Sun to keep increasing as more and better surveys are undertaken.

Bright Does Not Necessarily Mean Close

If we confine our census to the local neighborhood, we will miss many of the most interesting kinds of stars. After all, the neighborhood in which you live does not contain all the types of people—distinguished according to age, education, income, race, and so on—that live in the entire country. For example, a few people do live to be over 100 years old, but there may be no such individual within several miles of where you live. In order to sample the full range of the human population, you would have to extend your census to a much larger area. Similarly, some types of stars simply are not found nearby.

A clue that we are missing something in our stellar census comes from the fact that only six of the 20 stars that appear brightest in our sky— Sirius , Vega , Altair, Alpha Centauri, Fomalhaut, and Procyon—are found within 26 light-years of the Sun ( Figure 2 ). Why are we missing most of the brightest stars when we take our census of the local neighborhood?

The Closest Stars.

Figure 2. (a) This image, taken with a wide-angle telescope at the European Southern Observatory in Chile, shows the system of three stars that is our nearest neighbor. (b) Two bright stars that are close to each other ( Alpha Centauri A and B) blend their light together. (c) Indicated with an arrow (since you’d hardly notice it otherwise) is the much fainter Proxima Centauri star, which is spectral type M. (credit: modification of work by ESO)

The answer, interestingly enough, is that the stars that appear brightest are not the ones closest to us. The brightest stars look the way they do because they emit a very large amount of energy—so much, in fact, that they do not have to be nearby to look brilliant. You can confirm this by looking at Appendix J, which gives distances for the 20 stars that appear brightest from Earth. The most distant of these stars is more than 1000 light-years from us. In fact, it turns out that most of the stars visible without a telescope are hundreds of light-years away and many times more luminous than the Sun. Among the 6000 stars visible to the unaided eye, only about 50 are intrinsically fainter than the Sun. Note also that several of the stars in Appendix J are spectral type B, a type that is completely missing from Table.

The most luminous of the bright stars listed in Appendix J emit more than 50,000 times more energy than does the Sun. These highly luminous stars are missing from the solar neighborhood because they are very rare. None of them happens to be in the tiny volume of space immediately surrounding the Sun, and only this small volume was surveyed to get the data shown in Table.

For example, let’s consider the most luminous stars—those 100 or more times as luminous as the Sun. Although such stars are rare, they are visible to the unaided eye, even when hundreds to thousands of light-years away. A star with a luminosity 10,000 times greater than that of the Sun can be seen without a telescope out to a distance of 5000 light-years. The volume of space included within a distance of 5000 light-years, however, is enormous so even though highly luminous stars are intrinsically rare, many of them are readily visible to our unaided eye.

The contrast between these two samples of stars, those that are close to us and those that can be seen with the unaided eye, is an example of a selection effect. When a population of objects (stars in this example) includes a great variety of different types, we must be careful what conclusions we draw from an examination of any particular subgroup. Certainly we would be fooling ourselves if we assumed that the stars visible to the unaided eye are characteristic of the general stellar population this subgroup is heavily weighted to the most luminous stars. It requires much more effort to assemble a complete data set for the nearest stars, since most are so faint that they can be observed only with a telescope. However, it is only by doing so that astronomers are able to know about the properties of the vast majority of the stars, which are actually much smaller and fainter than our own Sun. In the next section, we will look at how we measure some of these properties.

Key Concepts and Summary

To understand the properties of stars, we must make wide-ranging surveys. We find the stars that appear brightest to our eyes are bright primarily because they are intrinsically very luminous, not because they are the closest to us. Most of the nearest stars are intrinsically so faint that they can be seen only with the aid of a telescope. Stars with low mass and low luminosity are much more common than stars with high mass and high luminosity. Most of the brown dwarfs in the local neighborhood have not yet been discovered.

New Atomic Data for Hot Star Studies

The spectrum and term analysis of Co III measured using Fourier transform and grating spectroscopy

D. G. Smillie, J. C. Pickering, G. Nave, and P. L. Smith
Astrophysical Journal Supplement 223, 12 (11pp), March 2016
Download PDF Copyright (2016) The American Astronomical Society.

Together with collaborators at NIST we have published new atomic data for doubly ionised cobalt,Co III including accurate wavelengths (130-256 nm) and atomic energy levels. These new data are important in ongoing studies of Hot Star using the Hubble Space Telescope, particularly for the Advanced Spectral Library(ASTRAL) Treasury project, which is the stellar equivalent of the Hubble Deep Field. These hot stars shine brightly in the ultraviolet. By understanding the composition of hot stars we can understand stellar and galaxy evolution. Our new data, measured using Fourier Transform spectroscopy at Imperial College London and NIST gives at least order of magnitude improvement in accuracy for Co III .

Spectral 'ruler' is first standardized way to measure stars

The first standardized way to measure stars has been developed for Gaia mission. Credit: Amanda Smith/Cambridge Institute of Astronomy

Previously, as with the longitude problem 300 years earlier for fixing locations on Earth, there was no unified system of reference for calibrating the heavens.

But now, when investigating the atmospheric structure and chemical make-up of stars, astronomers can use a new stellar scale as a 'ruler' – making it much easier for them to classify and compare data on star discoveries.

In fact, the work is a critical first step in the Gaia satellite's mission to map the Milky Way, as the unprecedented levels of stellar data that will result need "consistent stellar parameters", the same way we need values to measure everything from temperature to time, say astronomers.

The guidelines are free to download and are already being used by the world's largest astronomy projects. The work has recently been published in the journal Astronomy & Astrophysics.

The team, including Dr Paula Jofre from the University of Cambridge's Institute of Astronomy, selected 34 initial 'benchmark' stars to represent the different kinds of stellar populations in our galaxy, such as hot stars, cold stars, red giants and dwarfs, as well as stars that cover the different chemical patterns – or 'metallicity' – in their spectrum, as this is the "cosmic clock" which allows astronomers to read a star's age.

This detailed range of information on the 34 stars form the first value set for measuring the millions of stars Gaia aims to catalogue. Many of the benchmark stars can be seen with the human eye, and have been studied for most of human history – dating back to the very first astronomical records from ancient Babylon.

"We took stars which had been measured a lot so the parameters are very well-known, but needed to be brought to the same scale for the new benchmark - essentially, using the stars we know most about to help measure the stars we know nothing about," said Jofre.

"In previous galactic studies, the Sun is used as the standard to show a method is working, along with a few other well-known stars. But I choose this one because it works for my method, you choose a different one for different reasons data may not match.

"This is the first attempt to cover a wide range of stellar classifications, and do everything from the beginning – methodically and homogenously."

Launched at the end of last year, Gaia will gather data on over a billion stars in the Milky Way, allowing astronomers to study for the first time in close detail its myriad stars and planetary systems. Petabytes of data will be sent back to Earth – roughly the equivalent of all the information held in all the libraries of the world today.

The new value system was needed to ensure that analysis of this extraordinary amount of data is done in the most effective and efficient way, a template to measure the vast stream of data against.

Jofre focused on spectroscopic data to work out metallicity: the chemical combination that makes up a star. Just as a raindrop can split sunlight into the colours of the rainbow, spectroscopes split the light from a star into its chemical elements – and the results can be read like a musical score, with high notes or low notes in the scale giving clues as to the star's age. On average, the higher a star's metal content the younger it is.

Jofre created a 'spectral' library, combining the best data on the atmospheric structure of benchmark stars to determine a uniform scale for metallicity. Together with definitive scales for the stars' temperatures and surface gravities, produced with colleagues at the University of Uppsala and the University of Bordeaux, her work completes the measuring system that will be used to gauge data from Gaia.

"Now this set of data scales for the benchmark stars can be used as a way of making definitive measurements of others stars – invaluable to astronomers working on a wide range of projects," Jofre said.

The benchmark stars are already being used as a standardising model by Gaia's sister project, the Gaia-ESO survey, which is observing stellar spectra at a high resolution from the Very Large Telescope in Chile. They will also provide the basis for the thousands of reference stars needed to set the parameters for the hugely ambitious Gaia satellite once it starts mapping the entire galaxy – the "pillars for the enormous calibrators".

The fact that the ideal benchmark stars needed to be ones we already have a lot of data on means that many are bright and relatively near to the Earth – and have been the subject of wonder across civilisations.

Aldebaran, Arcturus, Pollux, Procyon and Alpha Centauri have played a part in the culture and mythology of mankind since they were first identified thousands of years ago. Babylonian astronomers used them as a reference point to describe the positions of the moon and planets as they moved through the night sky, appearing in the Babylonian Astronomical Diaries dating back to almost 1000 years BC.

"Many people interested in astronomy know these stars, their position in constellations, and the best time of year to see them. It is amazing that there is still so much to learn about the physics of these most well-known stars," said Dr. Ulrike Heiter from the Uppsala University.

"While stars do move over millennia, for humans they are fixed points – used to navigate the Earth for centuries. We are still using them as fixed points, but this time for navigating the galaxy," Jofre said.

UK Gaia lead Professor Gerry Gilmore added: "Advances in understanding the history and structure of our Galaxy with ambitious projects are possible only because, like Newton, we see farther by standing on the shoulders of giants. For reliably determining what chemical elements the stars are made of, those giants are the benchmark stars. All our vastly expanding knowledge depends on really understanding the few."

Anyone Can Do Astronomy with Python and Open Data

Ole Moeller-Nilsson, CTO at Pivigo, was kind enough to share his insights on how a beginner can easily get started exploring astronomy using Python. This blog post grew out of a presentation he gave at PyData London meetup on March 7th.

Python is a great language for science, and specifically for astronomy. The various packages such as NumPy, SciPy, Scikit-Image and AstroPy (to name but a few) are all a great testament to the suitability of Python for astronomy, and there are plenty of use cases. [NumPy and AstroPy are NumFOCUS fiscally sponsored projects SciPy and Scikit-Image are affiliated projects.] Since leaving the field of astronomical research behind more than 10 years ago to start a second career as software developer, I have always been interested in the evolution of these packages. Many of my former colleagues in astronomy used most if not all of these packages for their research work. I have since worked on implementing professional astronomy software packages for instruments for the Very Large Telescope (VLT) in Chile, for example.

It struck me recently that the Python packages have evolved to such an extent that it is now fairly easy for anyone to build data reduction scripts that can provide high quality data products. Astronomical data is ubiquitous, and what is more, it is almost all publicly available—you just need to look for it.

For example, ESO, which runs the VLT, offers the data for download on their site. Head over to and create a user name for their portal. If you look for data from the instrument SPHERE you can download a full dataset for any of the nearby stars that have exoplanet or proto-stellar discs. It is a fantastic and exciting project for any Pythonista to reduce that data and make the planets or discs that are deeply hidden in the noise visible.

I encourage you to download the ESO or any other astronomy imaging dataset and go on that adventure. Here are a few tips:

  1. Start off with a good dataset. Have a look at papers about nearby stars with discs or exoplanets and then search, for example: . Notice that some data on this site is marked as red and some as green. The red data is not publicly available yet — it will say under “release date” when it will be available.
  2. Read something about the instrument you are using the data from. Try and get a basic understanding of how the data is obtained and what the standard data reduction should look like. All telescopes and instruments have publicly available documents about this.
  3. You will need to consider the standard problems with astronomical data and correct for them:
    1. Data comes in FITS files. You will need pyfits or astropy (which contains pyfits) to read them into NumPy arrays. In some cases the data comes in a cube and you should to use numpy.median along the z-axis to turn them into 2-D arrays. For some SPHERE data you get two copies of the same piece of sky on the same image (each has a different filter) which you will need to extract using indexing and slicing.
    2. The master dark and bad pixel map. All instruments will have specific images taken as “dark frames” that contain images with the shutter closed (no light at all). Use these to extract a mask of bad pixels using NumPy masked arrays for this. This mask of bad pixels will be very important — you need to keep track of it as you process the data to get a clean combined image in the end. In some cases it also helps to subtract this master dark from all scientific raw images.
    3. Instruments will typically also have a master flat frame. This is an image or series of images taken with a flat uniform light source. You will need to divide all scientific raw images by this (again, using numpy masked array makes this an easy division operation).
    4. For planet imaging, the fundamental technique to make planets visible against a bright star rely on using a coronagraph and a technique known as angular differential imaging. To that end, you need to identify the optical centre on the images. This is one of the most tricky steps and requires finding some artificial helper images embedded in the images using skimage.feature.blob_dog .

    Using the tools offered by NumPy, SciPy, AstroPy, scikit-image and more in combination, with some patience and persistence, it is possible to analyse the vast amount of available astronomical data to produce some stunning results. And who knows, maybe you will be the first one to find a planet that was previously overlooked! Good luck!

    2. Potentially Volatile Phosphorus-Bearing Molecules

    In this section, we tackle the challenging problem of enumerating and prioritizing the phosphorus-bearing molecules (P-molecules) that could be spectroscopically observed in planetary atmospheres. There are two main approaches to this problem:

    • a targeted approach, developed in sub-section 2.1, that iteratively builds up a list of molecules based on known or proposed chemistry in planetary atmospheres including Jupiter, Earth, and Venus

    • a reaction-agnostic approach, developed in sub-section 2.2, that simply enumerates all molecules that fulfill certain criteria.

    2.1. Targeted Approach

    Our goal in this section is to identify target P-molecules that may be detectable in planetary atmospheres, including species that are predicted to be important for understanding the phosphorus chemistry on Venus. Table 1 details the small number of atmospheric P-molecules that have been explicitly considered.