by Rene Roy
the advancement of our knowledge in key areas of astrophysics. He contributed funda-
mentally to our understanding of the universe by decrypting stellar evolution. Sandage
made fundamental contributions to constructing and interpreting Hertzsprung–Russell
(H–R) diagrams of star clusters that display stars with their luminosity as a function of
effective temperature. As non-homomorphic representations, such H–R diagrams have been
crucial in revealing that stars evolved not by sliding down the “main sequence,” but by
drifting on well-defined tracks across the diagram, changing in luminosity and effective
temperature throughout their lifecycle, at rates depending on the initial mass of the star.
Sandage delved into the nature of mysterious radio sources, discovering the first quasar,
and helped to refine the extragalactic distance scale. He measured the properties of expand-
ing spacetime.2 Sandage’s dream was nothing less than “decoding cosmic evolution.”
Throughout his whole career Sandage accomplished most of his work with indomitable
energy and passion by taking, analyzing and interpreting images. He pushed for building
one of the best imaging telescopes, the 2.5-m Irénée du Pont telescope at Las Campanas,
Chile, with its large field of view. For him, images were full of secrets just waiting to be
deciphered by the persistent and scrutinizing mind, which could reveal what he expected to
be the deeper nature of our universe. A true artist of the scientific image, he shared his work
and strong vision in several magnificent ways. As main author of four of the great galaxy
1 A. Sandage, Centennial History of the Carnegie Institution of Washington, Volume 1: The Mount Wilson Observatory, Cambridge: Cambridge University Press, 2004, p. 178.
2 D. Lynden-Bell and F. Schweizer, Allan Rex Sandage, 18 June 1926–13 November 2010, Biographical Memoirs of Fellows of the Royal Society, 2012 (arXiv:1111.5646).
247
14:11:34, subject to the Cambridge Core
.014
248
Conclusion
atlases, Sandage passed on an inspiring and lasting legacy. Few astronomers have used and
demonstrated in more vivid ways the power of images for scientific discovery.
Science seeks order in the natural world of “things.” As I have shown in this book,
images have played an essential role in discovering order in the sidereal world, in unveiling
galaxies as “island-universes” and in establishing the astounding structure and dynamics of
the expanding universe. Allan Sandage was certainly a master and a leading explorer in this
complex process. Along with Gérard de Vaucouleurs and the iconoclastic Halton “Chip”
Arp, he drew for us some of the best roadmaps to the universe.
Images as Roadmaps to the Universe
But what about all these images? What has been their role on this astounding path of dis-
covery? I surmise that images of galaxies crystallize our cosmic knowledge along four cog-
nitive dimensions. “Without a visual first-hand impression we are literally fumbling in the
dark, but once we have an image, we rarely look back at it, as it becomes a ‘simplicity’.”3
Images of galaxies help us (i) to grasp the vast physical scale of the universe, (ii) to climb
the ladder of cosmic complexity, (iii) to sharpen our aesthetic and epistemological sense in
positioning ourselves in nature at large and, finally, (iv) to drive the design and building of
increasingly powerful instruments, while initiating new research programs.
First, images are most powerful in representing at a human scale objects of fantastically
different spatial scales. For example, objects as different as superclusters of galaxies that
embrace 10 million light-years of length in space, animals of human size, plant cells of a
few micrometers or silicon atoms on the surface of a crystal at the nanometer scale can
all be displayed in a page of a book or a scientific atlas or on a computer screen. At the
upper rungs of the cosmic ladder, images help us visualize the immense magnitude of the
universe: from successive images going from the Earth–Moon tandem, to our solar system
planetary world, to nearby star clusters like the Pleiades, to the larger globular clusters, to
the Milky Way and its Magellanic Cloud satellites, to the Local Group of galaxies, to the
Virgo cluster of galaxies, and further still. Hence, we move from scales of light-seconds
to millions and billions of light-years. Images help us to assemble a cognitive chart of the
universe and of its components and to grasp their wildly diverging physical scales.
Secondly, images of astronomical objects and their history reveal the amazing diversity
and range of structures of celestial objects. “The study of size, shape, brightness, central
concentration, degrees of mottling, rotation, change, and movement, as well as questions
of morphology, identity, classification, and evolution, are current to contemporary nebu-
lar and extragalactic astronomy, just as they were in the nineteenth century.”4 Seeing and
viewing while scripting notes was an initial step in exploring celestial features. However,
even the best and most detailed written notes did not suffice. It would have been challenging
for anyone to take the observing notes of a skillful nineteenth-century observer and try to
3 Lars Lindberg Christensen, e-mail to the author of November 2016.
4 O. W. Nasim, Observing by Hand: Sketching the Nebulae in the Nineteenth Century, Chicago: University of Chicago Press, 2013, p. 233.
14:11:34, subject to the Cambridge Core
.014
Conclusion
249
reconstruct even an approximate picture just from his/her comments and descriptors.
Sketching of nebulae, in support of the act of viewing and observing, illustrated their
complexities. Then twentieth-century technologies totally transformed these processes.
Astrophotography, aperture synthesis using radio waves and X-ray imaging revolutionized
astronomical imaging; they provided us with tools to examine cosmic objects in the finest
details over the complete electromagnetic spectrum in previously undreamt-of ways. These
latter steps were essential to understand the nature of cosmic objects and to reveal the com-
plex underlying physical processes.
Using a battery of telescopes and sensors, we became able to monitor objects over time,
either by following single objects or by comparing vast ensembles of cosmic objects of the
same class. We established that astronomical timescales are to be measured in thousands,
millions or billions of years. Variations of brightness of many astronomical objects over
time have turned out to be incredibly rich, even over relatively short timescales. Powerful
imaging techniques have revealed significant variations in some objects over time on the
hourly and even daily scale. Some phenomena, such as pulsars, fluctuate on millisecond
timescales, which is amazingly short for astronomical objects. The firmament of fixed stars
has been blown apart.
Thirdly, images of galaxies may not be as spectacular as those of solar system objects
taken by interplanetary space probes. They may not offer the diversity of forms or colours
of butterflies, birds, or palm trees, and far less than many of the natural objects illustrated
so spectacularly by Ernst Haeckel in Kunstformen der Natur, a book of lithographic and<
br />
halftone prints. We may also not see the finest morphological or kinematic details such
as those of the atmosphere of Jupiter, the rings of Saturn or the icy pingos of Pluto. Nor
do we see in the average galaxy the explosive range of colours of galactic nebulae (unless
false colours are used to highlight features). Despite these relative pictorial shortcomings,
galaxies at first sight do display order and stability on the grandest scale of nature. And
at the other extreme, they unveil the ultimate cosmic cataclysm, the grandiose chaos of
colliding and merging galaxies. Reconstructed by computer simulations, real images of
merging galaxies make us realize our smallness: how fleeting and contingent our existence
is, how small our world, itself sailing around an inconspicuous dwarf star carried around
with billions of others in the giant Milky Way merry-go-round.
Fourthly, many impressive telescopes have been built and continue to be built, each gen-
eration more powerful and an improvement on the previous one. The first great reflectors
of the twentieth century were designed in great part to observe and explore the nature of
“nebulae,” when most of them became “galaxies.” Creative opticians and builders drove on
relentlessly for better and deeper imaging capabilities: in succession, we had William Her-
schel’s 40-ft reflector, William Parsons’ 6-ft Leviathan, the Crossley 36-inch, the Ritchey–
Hale–Pease’s Mount Wilson 60- and 100-inch, the Palomar 5-m, Irénée du Pont 2.5-m, and
finally the Hubble Space Telescope, the latter being one of the most expensive and pro-
ductive scientific machines ever built. The 6.5-m James Webb Space Telescope, due to be
launched in 2018, has as one of its core scientific missions the imaging of the first galaxies
that formed more than 12 billion years ago. On the ground, following the dozen or so 8–
10-meter giants (Subaru Telescope, twin Gemini telescopes, the Very Large Telescope, the
14:11:34, subject to the Cambridge Core
.014
250
Conclusion
Keck twin telescopes and a few others), three new mammoth telescopes of 25–40 m in size
are under construction. The Atacama Large Millimeter Array (ALMA), the great submilli-
meter wavelength interferometer on the altiplano of northern Chile, is providing unique
views of young planetary systems surrounding nearby stars and of molecular clouds in dis-
tant forming galaxies. Their imaging capabilities would have been inconceivable only 50
years ago.
The Digital Universe
We are at the dawn of a new age in astronomical imaging. Soon, telescopes will generate
terabytes of imaging data, day and night. What will happen with the widely used digital
imaging and computer processing of astronomical images? What is the impact of the com-
puter age? There is no doubt that the digital age is changing the role and use of atlases of
galaxies, and of scientific imaging. This is evident from just flipping the pages of the weekly
issues of the journals Science and Nature, or browsing the daily posting of new images on
the NASA website, Astronomical Picture of the Day.5 Modern computer-based catalogues
and sophisticated query tools have become part of everyday research life. A powerful syn-
ergy is growing between the new tools and the large synoptic surveys of the sky from the
ground and space. Multiwavelength views have superseded photographic plates. By clev-
erly combining images of a given object obtained at different wavelengths and stacking
them like a sophisticated pack of virtual cards, researchers now work with multiwavelength
data cubes and multidimensional data entities.
With all these electronic advances, atlases are also entering a new age.6 With the power-
ful panoramic detectors, we have moved from producing analogue images to digital ones.
This emphasizes the key difference between analogue and digital. Photographic plates, the
best analogue photographic detectors, were non-linear in their response to light: that is, an
increase of light by a factor of four did not translate into a factor of four of darkening of
the emulsion; the chemical reaction was instead some complex response function that was
tricky to calibrate and required time and training on the part of the observer. With digital
electronic detectors like CCDs, the response is linear; ten times as many photons result in
ten times more electrons. Furthermore, digital detectors or imagers generate a quantifiable
number ready for the mathematical operations of computer image processing. The han-
dling of images has become a sort of mind aerobatics, and many astronomers have become
extremely adept at this. Several user-friendly tools and applications have been created and
shared, enabling complex image manipulations to be carried out by amateur astronomers
and students, not only professionals.
What will the future galaxy atlases be? It is now possible to do things that were impos-
sible with printed atlases, while still serving the same purposes and in a far more versatile
way. We have the ability to make images with extremely large numbers of pixels covering
5 http://apod.nasa.gov
6 I am most grateful to Lars Linberg Christensen (European Southern Observatory) and Anton Koekemoer (Space Telescope Science Institute) for sharing their perspectives on the role of atlases in the digital age.
14:11:34, subject to the Cambridge Core
.014
Conclusion
251
a much higher dynamic range of brightness. This provides a far greater cognitive power
in viewing the overall scene or for zooming in on details. With a few keyboard clicks, we
are able to vary image parameters such as the filter or the wavelength range we wish to
use to examine a field, bringing out the underlying physics. Computer-aided, one can also
skim through thousands of images in a matter of minutes, arrange appropriate sequences
and tune into the desired wavelength of interest. These manipulations can reveal subtle
commonalities and differences between images.
Astronomers have assembled a colossal treasure: millions of images of the sky at multi-
ple wavelengths and multiple tools to analyze them. For example, the Virtual Observatory
is a collection of astronomical data centers. Among its many tasks and projects, it offers
software systems and powerful image processing capabilities to researchers. By appropri-
ately mastering and enabling tools of artificial intelligence, astronomers have new ways of
browsing through archives and are able to combine images obtained, for the same object, in
the X-ray, the visible and the radio domains. It is important to note some of the challenges
of multiwavelength imaging. When images come from different telescopes and instruments
spanning the electromagnetic spectrum, how do we ensure proper alignment of the images?
If we correct the images for geometric distortions, are we compromising the measurements?
The alignment of features in different spectral regions does not necessarily mean that they
are the same physical object.7 The tools offered by the Virtual Observatory help to address
these issues and make automated multispectral imaging analysis, interpretation and classi-
fication rewarding. Perhaps, one day, the galaxy researcher will be able to construct his or
her atlas on demand.
Galaxy Images for Everyone
The best example of the new age is the Hubble Legacy Archive.8 This is one of mankind’s
finest modern atlases and it is growing. The people at the Space Telescope Science Institute
(STScI/NASA), the Space Telescope European Coordinating Facility (ST-ECF/ESA) and
the Canadian Astronomy Data Centre (CADC/NRC/CSA) have also made considerable
efforts to visualize the Hubble Space Telescope data as well and as efficiently as possible.
The data in the thumbnails and “interactive displays” are uniformly treated and provide an
excellent way for scientists to get a first-hand impression of the data.
Innumerable things remain to be discovered from archival material, as old photographic
plate collections are digitized and loaded into modern archives. It is crucial that this old
observatory material is protected and not lost. The colossal effort of cleaning and digitizing
the 500,000 photographic plates of the Harvard College Observatory plate stacks deserves
particular praise.
7 Glenn Mackie, The Multiwavelength Atlas of Galaxies, Cambridge: Cambridge University Press, 2011.
8 https://hla.stsci.edu/hlaview.html
14:11:34, subject to the Cambridge Core
.014
252
Conclusion
Several observatory outreach departments are deploying significant efforts to produce
ethically correct, impressive colour imagery from the raw telescopic data.9 Started many
years ago as Pictures of the Week projects for both the European Southern Observatory and
Hubble Space Telescope, this developing archive assembles images from existing databases
or published material. Images are tagged with various metadata (coordinates, orientation,
wavebands, etc.) providing context that may be useful for both laypeople and experts. Often
representing the best images of the objects in question, this project constitutes a slowly
accumulating archive, close to an electronic atlas. AstroPix is another project combining
all available images (across all wavelengths) into one searchable interface.10 Finally, images
are also uploaded to Wikipedia that may arguably be the biggest atlas in the making.
What a long and fascinating road we have traveled since Al-Sufi’s description of the