The Role of Images in Astronomical Discovery
Page 34
list. “Corrected Hubble types, complete coverage of the redshifts (adding to the earlier
data by Humason and Mayall), and modern, mostly photoelectric magnitudes for every
Shapley–Ames galaxy were listed.”35 In the later version of 1987, Sandage and Tammann
provided new morphological types for about 200 southern galaxies and updated distances
for many local galaxies. As a working object, the RSA advocated the Hubble–Sandage
33 S. van den Bergh, A Preliminary Luminosity Classification of Late-Type Galaxies, and A Preliminary Luminosity Classification for Galaxies of Type Sb, The Astrophysical Journal, 1960, Vol. 131, pp. 215–223 and pp. 558–573.
34 D. M. Elmegreen and B. G. Elmegreen, Arm Classifications for Spiral Galaxies, The Astrophysical Journal, 1987, Vol. 313, pp. 3–9.
35 A. R. Sandage, Centennial History of the Carnegie Institution of Washington, Volume 1: The Mount Wilson Observatory, Cambridge: Cambridge University Press, 2004, p. 492.
14:11:09, subject to the Cambridge Core
.012
.012
14:11:09
, subject to the Cambridge Core
Fig. 10.8 Page of the Takase, Kodaira and Okamura photometric atlas showing its display of galaxy image, isophotal map and radial light profile. With permission of University of Tokyo Press.
10. Atlases of Galaxies, Picturing “Island-Universes”
221
classification scheme and provided the visual tools to apply it to a whole new set of galax-
ies. This catalogue and mini-atlas became an obligatory desktop reference and handbook for
observational work, from planning to data reduction, analysis and interpretation. A limited
number of archetypal objects and a concise format helped in memory recall, i.e. observers
didn’t need to have larger-sized atlases to hand. Nowadays the RSA, as a hard-copy book
or e-book, continues to be a standard reference, just like a favorite recipe book.
1984. An Atlas of Selected Galaxies: With Illustrations of Photometric Analyses , The
Trust for Quantitative Imaging
The first historical stage in atlas making, initiated by Hubble, was based on recognizing
patterns from visual inspections of photographic images; it was essentially qualitative and
subjective. Hubble’s and Arp’s pure morphological and qualitative approaches were consid-
ered insufficient by many. Wishing for a more quantitative approach, Japanese astronomers
Bunshiro Takase, Keiichi Kodaira and Sadanori Okamura sought to follow in the footsteps
of José Luis Sérsic.
Using the Japan-based Okayama 188-cm reflector and the Kiso Observatory 105-cm
Schmidt telescope, Takase and his colleagues took photographs of galaxies through a set of
colour filters, which were reduced to produce digitized photometric data with the help of
high-performance measuring instruments and computerized image processing techniques.
This photometric atlas in fact represented the first effort to include clusters of galaxies in a
systematic photometric survey (Fig. 10.8).
They used the results of new digital photometry to classify galaxies. In their 1984 An
Atlas of Selected Galaxies: With Illustrations of Photometric Analyses, the authors declared
that galaxy classification required this new and more quantitative phase.36 The atlas illus-
trated the computerized analysis of galaxy images using various image-processing tech-
niques. The purpose was to demonstrate how photographs of galaxies could be reduced
into quantitative data in the forms of calibrated isophotal maps, colour maps and radial
luminosity profiles. The purpose of the small atlas was to demonstrate the technique of
photographic-plate digitalization and computer analysis; it was not a systematic presenta-
tion of a large number of objects.
The intention of this quantitative analysis was to reduce the subjectivity and thus estab-
lish an “objective classification”: the trio of Japanese astronomers claimed rightly that quan-
titative data were needed to provide reliable physical information on the luminosity and
colour distributions for the different structural elements of galaxies. Although photomet-
ric data had been used in the earlier atlas of Sérsic, Takase et al.’s work marked the entry
of galaxy atlases into the digital age, with a more advanced mathematical treatment that
revealed hidden structural features of galaxy morphology. A highly innovative feature of
the atlas was the application of two-dimensional Fourier analysis to quantitatively identify
36 B. Takase, K. Kodaira and S. Okamura, An Atlas of Selected Galaxies: With Illustrations of Photometric Analyses, Tokyo: University of Tokyo Press and Utrecht: VNU Science Press, 1984.
14:11:09, subject to the Cambridge Core
.012
222
Part III – Organizing the World of Galaxies
and discriminate the modes of oscillations of galactic disks, which give rise to the various
spiral patterns.
1988. Atlas of Galaxies Useful for Measuring the Cosmological Distance Scale ,
Superlative Imagery
The Hubble Space Telescope (HST) launched in 1990 was equipped (and still is) with cam-
eras that provided a field of view of only a few arcminutes, a fraction of the angular size
of the Moon, which is significantly smaller than that provided by most ground-based tele-
scopes. Furthermore, this moderate-sized telescope operating in space held the extraordi-
nary promise of an unprecedented angular resolution, which would reveal spatial details
ten times finer than the best ground-based telescopes. This was a very significant jump in
angular resolution, to the point that much more accurate visual references were required to
see where the telescope was pointing and looking at. Among the several pre-launch prepa-
rations for the HST, an unusual venture was set up to help observers to find their way among
the promised exquisite images: it was the production of a giant and finely illustrated atlas
of galaxies. Designed and put together by Allan Sandage and John Bedke, the Atlas of
Galaxies Useful for Measuring the Cosmological Distance Scale appeared in 1988.37 The
giant atlas, with 40 × 51-cm-sized pages, gets close to some of the largest-ever-published
scientific atlases (Figs. 0.6 and 10.9).38
The completion of the Mount Wilson/Palomar/Las Campanas survey of galaxies by
Sandage and collaborators in 1985 had provided most of the large-scale photographic mate-
rial needed for this special atlas; some of it had already been used in the RSA. The sample of
galaxies, to be potential HST targets, included 322 objects selected again from the Shapley–
Ames set of bright galaxies. Sandage and Bedke stated that the purpose of this unusual atlas
was to provide a selection of excellent galaxy photographs, with images of adequate scale
for the resolution of the extragalactic stellar content with the powerful HST cameras.
Because the Virgo cluster of galaxies was to be a prime target of the HST programs, many
objects from the central region of the Virgo cluster of galaxies were included in the new
atlas. The last eight plates of the atlas showed several Virgo cluster galaxies, in anticipation
of their use for determining the Hubble constant, i.e. the expansion rate of the universe,
with greater accuracy; this was one of the reasons for building the HST in the first place.
The ultimate theoretical goal was to obtain a more accurate determination of the spacetime
these, the precise age of the universe could be inferred.39
To achieve this goal, a reliable and precise extragalactic distance scale, i.e. based on
absolute distances, was required. The determination of distances to galaxies is based on the
measurement of the properties of primary objects (luminous stars such as Cepheids, HII
37 A. R. Sandage and J. Bedke, Atlas of Galaxies Useful for Measuring the Cosmological Distance Scale. Washington: Scientific and Technical Information Division, NASA, 1988.
38 L. Daston and P. Galison ( Objectivity, New York: Zone Books, 2007, pp. 23–25) describe James Bateman’s double-elephant 17-kg and 68 × 98 cm folio atlas The Orchidaceae of Mexico and Guatemala (London: Ridgway, 1837–1843). This rare book can be found on sale currenly for close to $120,000 US dollars.
39 Sandage and Bedke also use the expression “the Creation” or “since Creation” instead of the Big Bang.
14:11:09, subject to the Cambridge Core
.012
10. Atlases of Galaxies, Picturing “Island-Universes”
223
Fig. 10.9 Atlas of Galaxies Useful for Measuring the Cosmological Distance Scale, by Sandage and
Bedke (1988). Courtesy of NASA Scientific and Technical Information Division.
regions, supernova remnants) and comparing them with nearby Milky Way analogues for
which distances are relatively well established. A high spatial resolution is required, and
few galaxies are suitable. This is a result of a combination of the background light produced
by the galaxy’s billions of unresolved stars, the crowding together of the brighter stars, dust
obscuration and a high inclination to the line of sight of the galaxy plane or just distance, all
of which greatly reduce the ability to resolve individual stars. The new atlas would alleviate
14:11:09, subject to the Cambridge Core
.012
224
Part III – Organizing the World of Galaxies
all these problems. It provided a set of galaxy photographs with images of an adequate scale
for estimating how finely the stellar content could be resolved with the more powerful HST.
The images of the atlas were assumed to be of good quality and were displayed at a scale
suitable to help positioning HST observing fields for optimal control of stellar crowding.
The technical objective was to resolve and recognize the largest possible number of stars,
then to make use of the primary (Cepheid variables) and secondary targets (e.g. red and
blue supergiant stars, novae, supernovae), the so-called distance indicators. The HST was
expected to provide a spatial resolution gain of a factor of 10, and Sandage and Bedke
illustrated this by putting images of the nearby galaxy Messier 33, and of Messier 101,
located 10 times further away, side by side. As a working object, the atlas aimed to educate
the potential users of HST ahead of time. The enormous size of the atlas still makes it a
most conspicuous object in libraries and researchers struggle to find adequate space for it
in their offices.
1988. The Color Atlas of Galaxies , Searching for the Meaning of Colours in Galaxies
Until the late 1980s, all galaxy atlases had used black-and-white images, reproduced from
photographic plates. Today, we see so many spectacular colour images of galaxies that
it is hard to imagine that, 30 years ago, producing colour photographs of galaxies was a
real challenge. The American astronomer James D. Wray published the first Color Atlas
of Galaxies in 1988.40 It was an effort to produce true-colour pictures of galaxies with
the highest fidelity. Wray used a meticulously chosen set of photographic emulsions and
filters to ensure “true colours,” as the author’s “overriding consideration was the need for
the colours to be reliable and interpretable.” This meant reproducing a colour palette that
would correspond to what a very sensitive human eye would perceive when looking at these
galaxies.
The atlas comprised colour images of more than 600 galaxies of all forms, including
“peculiar” and interacting ones. Wray assembled his core sample of galaxies from de Vau-
couleurs et al.’s Reference Catalog of Bright Galaxies (RC1),41 and added galaxies known
or suspected to have nuclear activity or extended dust content. The imaging campaigns
for the image database were conducted at McDonald Observatory in Texas, Las Campanas
Observatory and Cerro Tololo Inter-American Observatory (CTIO) in Chile.
As the author surmised, “color images offer a source of aesthetic appreciation and scien-
tific investigation.” For example, colour allows one to distinguish a very young stellar popu-
lation (bluish knotty regions) from old populations (extended yellowish light component).42
To ensure colour authenticity, a colour reference was established, based on photometric data
from individual bright stars in the field. The use of the Eastman three-colour dye transfer
process enabled a correct colour balance at all brightness levels and a consistent relative
surface brightness from galaxy to galaxy throughout the atlas. The manuscript submitted
40 J. Wray, Color Atlas of Galaxies, Cambridge: Cambridge University Press, 1988.
41 G. de Vaucouleurs, et al., Third Reference Catalogue of Bright Galaxies, New York: Springer, 1991.
42 W. Baade had introduced the concept of stellar populations in his book Evolution of Stars and Galaxies, C. Payne-Gaposchkin (editor), Cambridge, MA: Harvard University Press, 1963.
14:11:09, subject to the Cambridge Core
.012
10. Atlases of Galaxies, Picturing “Island-Universes”
225
by Wray was excellent. Despite all these efforts, however, the resulting product was disap-
pointing because of the very poor quality of printing.
Nevertheless, the atlas benefited from the enthusiastic support of none other than Allan
Sandage who emphasized that “the colors are true.” Writing in the preface, Sandage envis-
aged that the work would become an exceptional pedagogical tool to teach three central
themes: (1) the Hubble classification system, (2) Baade’s stellar population concept, and
(3) the life cycle of stars determining the stellar content of galaxies. In glowing words,
Sandage continued: “From this book, galaxies will be chosen for work with the HST on
the resolved stars for distance determination, applicable to the yet unsolved problem of the
value of the Hubble constant; for work on the ground-based telescopes on star formation
rates in galaxies of different Hubble types, studies of dust distributions, . . . ” Sandage ended
with unrefrained praise: “This atlas is important and . . . will be viewed some years hence
as one of the major reference works produced in astronomy in the last half of the 20th cen-
tury.” Sandage had obviously reviewed the original manuscript with splendid photographic
copies. Commenting on these originals, the American astronomer Harold G. Corwin Jr.
writes, “The colors were not deeply saturated as most color astronomical photos seem to be,
but were as subtle as I imagined the galaxies might be as viewed with a huge telescope.”43
1990. Photometric Atlas of Northern Bright Galaxies , Mapping Light across Galaxies
In 1990, the Japanese astronomers Keiichi Kodaira, Sadanori Okamura and Shin-Ichi
Ichikawa issued a new photometric galaxy atlas, the Photom
etric Atlas of Northern Bright
Galaxies, based on photographic plates, obtained with the Kiso 105-cm Schmidt telescope,
of 791 galaxies, chosen from a subset of the RSA listing.44 Their work came in the wake
of Takase et al. (1984), summarized above. They were continuing their colleagues’ quan-
titative photometric approach. The new atlas went beyond just the demonstrative stage. It
was part of a long-term project to produce a large database of the two-dimensional surface-
brightness distributions of bright galaxies. It was a colossal enterprise, with images of all
the galaxies recorded in the photographic V band (500–650 nm). The original photographs
were all digitized using a Perkin-Elmer PDS photodensitometer, a complex and relatively
expensive machine at the time, which converted the darkening level of the photographic
emulsion into a quantitative measure of brightness.
Their effort was driven by the growing need to have increasingly accurate photometric
data. Although not pursuing a complete survey, the authors’ aim was to obtain homogeneous
optical surface photometric data for as many nearby galaxies as possible, with isophotal
maps and photometric parameters processed in a consistent way. From a scientific perspec-
tive, this atlas was to provide a quantitative galaxy database to be used as a reference for
observations of galaxies conducted at other wavelengths, as new observing capabilities in
the radio and X-ray domains began to provide maps and images of these objects (Chapter 7).
43 H. G. Corwin Jr., e-mail communication with the author (December 2016).
44 K. Kodaira, S. Okamura and S.-I. Ichikawa, Photometric Atlas of Northern Bright Galaxies, Tokyo: University of Tokyo Press, 1990.
14:11:09, subject to the Cambridge Core
.012
.012
14:11:09
Fig. 10.10 The Carnegie Atlas of Galaxies, interior pages. With permission of Carnegie Science.
, subject to the Cambridge Core
10. Atlases of Galaxies, Picturing “Island-Universes”
227
Kodaira, Okamura and Ichikawa produced a homogeneous and well-calibrated database
of the largest quantitative sample of northern bright galaxies of its time. Each galaxy was
presented in three panels: a photograph of each galaxy from the computer display screen; the