The Role of Images in Astronomical Discovery
Page 26
states of parallel versus opposite alignment. The transition is called “forbidden,” meaning
that the probability of the flip is extremely low, occurring only once every few million years.
Most often the flip is triggered by collisions between hydrogen atoms which don’t result
in a radio photon, as the two colliders get away with the energy. However, because there
is a lot of hydrogen in many low-density regions in galaxies, the 21-cm line is produced
continuously throughout the universe, and can be observed at great distances.
The Race to the Spin
Dutch astronomer Konrad C. van de Hulst (1918–2000), then a Ph.D. student working
with Jan Oort (1900–1992) at Leiden Observatory, predicted the spin transition of neu-
tral hydrogen in a 1945 paper.23 In 1949, Iosif Shklovsky (who had imagined a mechanism
for synchrotron radiation) confirmed that the hydrogen radio line should exist and could be
detected. The Dutch group searched for several years. It was Harold I. Ewen (1926–2013),
then a brilliant young American astronomer and graduate student at Harvard University,
working with Edward Purcell (1912–1997), who found it first.24 The duo had developed a
new method of radio observing called “switched frequency” mode. Using a quickly assem-
bled horn antenna, made of plywood covered with copper foil mounted outside a laboratory
window, they detected the hydrogen spectral line. They soon shared their technique with
Dutch and Australian colleagues who almost immediately also found the line.25 Neutral
hydrogen, or HI emission, was ubiquitous.
The discovery was the start of another immensely fruitful episode in radio astronomy.
The HI radio line became a superb tool to map the gas in the interstellar medium of our
Milky Way and in nearby galaxies. The spiral structure and the rotation of the Milky
Way was confirmed for the first time by HI observations conducted by Oort and his team
(Fig. 7.6).26,27 Following other successful observations by the pioneering teams, the radio
telescopes being constructed and the existing ones were modified to better detect and mea-
sure the 21-cm line, and record radio images of galaxies and other objects.
A strong astrophysical driver for these facilities was to employ the 21-cm line to map
the kinematics and dynamics of the cold hydrogen gas in galaxies and then use Newtonian
physics to infer the total mass of galaxies. The HI line became most advantageous because
it is intrinsically very narrow, due to the very long lifetime of the spin quantum energy
states. Moreover, the interstellar medium is mostly transparent to this sharp radio line.
23 H. C. van de Hulst, Radio Waves from Space: Origin of Radio Waves, Nederlands tijdschrift voor natuurkunde, 1945, Vol. 11, pp. 210–221.
24 H. I. Ewen and E. M. Purcell, Radiation from Galactic Hydrogen at 1,421 MHz, Nature, 1951, Vol. 168, p. 356.
25 C. Alex Muller and Jan H. Oort, The Interstellar Hydrogen Line at 1,420 MHz and an Estimate of Galactic Rotation, Nature, 1951, Vol. 168, pp. 356–358.
26 J. H. Oort, Observational Evidence confirming Lindblad’s Hypothesis of a Rotation of the Galactic System, Bulletin of the Astronomical Institutes of the Netherlands, 1927, Vol. 3, pp. 275–282.
27 Jan H. Oort, Frank J. Kerr and Gart Westerhout, The Galactic System as a Spiral Nebula, Monthly Notices of the Royal Astronomical Society, 1958, Vol. 118, pp. 379–389.
14:06:02, subject to the Cambridge Core
.009
7. A Symphony of Waves
167
Fig. 7.6 Transformational Image: The Milky Way Map of Neutral Hydrogen. Distribution of
neutral hydrogen in the plane of the Milky Way observed at 21 cm.28 Maximum densities in the
z-direction are projected onto the plane and contours of equal densities are drawn (darker is denser).
The Sun is at the juncture of the axis, and galactic longitudes are indicated. The mapping of the struc-
ture was derived from the observed intensities and radial velocities of the 21-cm line along different
lines of sight. The authors assumed circular symmetrical symmetry and used the model of the rota-
tion curve derived by Maarten Schmidt.29 Elongated features are mimicking spiral arms. This is a fine
example of non-homomorphic representation (Chapter 4). From Oort, Kerr and Westerhout (1958),
Monthly Notices of the Royal Astronomical Society.
28 J. H. Oort, F. J. Kerr, and G. Westerhout, The Galactic System as a Spiral Nebula, Monthly Notices of the Royal Astronomical Society, 1958, Vol. 118, pp. 379–389.
29 M. Schmidt, A Model of the Distribution of Mass in the Galactic System, Bulletin of the Astronomical Institutes of the Netherlands, 1956, Vol. 13, pp. 211–222.
14:06:02, subject to the Cambridge Core
.009
168
Part II – Images as Galaxy Discovery Engines
With the advent of aperture synthesis in the 1960s, radio interferometers were built and
used in Westerbock (Holland), Mullard (UK), Penticton (Canada) and Greenbank (USA).
They revealed many outstanding features of the neutral hydrogen in galaxies. At the rela-
tively small scale of a few light-years, the gas is distributed in filaments. On a larger scale,
the spiral arms are seen in neutral hydrogen; they are regions corresponding to an increase of
interstellar gas density. Within a few years, the research groups at Leiden and then Gronin-
gen, in Holland, had mapped the spiral structure of the Milky Way and were able to derive
the rotation curve of the Milky Way inside the Sun’s galactic orbit (Fig. 7.6).
Mapping neutral hydrogen became a “sport.” It was quickly found that neutral hydrogen
extended much further than the disks of galaxies seen in the visible light, sometimes twice,
even three times, further (see Plate 7.5). Maps of the radial velocity of the neutral hydrogen
were made of many galaxies (see Plate 4.2). But a mystery appeared. Instead of a falling off
in a Keplerian way as a function of distance, which would indicate a tapering galaxy mass
distribution with radial distance, the rotation velocity was found to remain constant to the
edge of the neutral hydrogen disk (Chapter 8). In addition to rotation, peculiar structures like
warps and non-symmetrical velocities were found, revealing morphological and kinematic
anomalies. Most were assigned to interactions with companion galaxies, to mergers or to
other non-axisymmetric behavior that is indicative of past cannibalism, where a galaxy has
evolved by swallowing a companion or smaller satellite galaxies.
Images in the Submillimeter Range
The successful technique of aperture synthesis was also being applied to shorter radio wave-
lengths, for example in the millimeter wavelength range, where molecules, in particular CO,
can be studied and molecular spectroscopy can be carried out. The Atacama Large Milli-
meter Array (ALMA) in northern Chile, with its 64 antennae, is the most recent and pow-
erful of these new millimeter/submillimeter arrays, and is capable of imaging galaxies at
very large distances (Plate 6.1).
But what about more direct radio imaging, as we do with our digital cameras? Surmount-
ing severe technical challenges, radio astronomy did not lag behind other astronomy fields
for long in developing panoramic electronic detectors. Over the last 20 years, we have wit-
nessed the arrival of multi-pixel radio-wavelength detectors equivalent to the CCDs used
for visible light. These arrays of detec
tors are capable of making direct radio images of the
sky in the millimeter and submillimeter wavebands. The current technique uses arrays of
miniature bolometers, which are lined up in rows and columns like a chessboard to register
radio signals from a portion of the sky that is focused by a parabolic antenna of high surface
precision.
A bolometer is a kind of thermometer. The radio photons are absorbed in each individual
bolometer by a thin metal film cooled to a very low temperature, about 0.280 K, just above
absolute zero. Heated by the sky’s radio signal, the metal changes its temperature as it
absorbs the radiation. A heat-sensitive semiconductor measures the change as a voltage blip,
the voltage being proportional to the intensity of the incoming radiation. In current devices,
14:06:02, subject to the Cambridge Core
.009
7. A Symphony of Waves
169
the thermal, electrical and mechanical structure of the bolometer array is set on a single
silicon wafer. Freestanding silicon nitride membranes, of less than one micron in thickness,
are laid out with thin layers of titanium, which absorb the radiation. The temperature change
in the absorbers is measured through semiconducting germanium chips, doped by a neutron
transmutation, with electrical resistance that is acutely dependent on temperature. Making
images with impinging radio waves is another amazing achievement of human ingenuity.
To complete this chapter on imaging outside the optical window, let us swing to the other
end of the electromagnetic spectrum, the X-ray domain.
From Conic Sections to X-Ray Telescopes
Many regions in the interstellar medium are filled with hot gases of a few thousand degrees;
these are the diffuse galactic nebulae; the Orion Nebula is the most spectacular such object
nearby. The gas is ionized; it is called a plasma: atoms, mostly hydrogen and other heavier
atoms such as oxygen, carbon and nitrogen, have lost some or most of their electrons due
to ionization by ultraviolet radiation or collision with surrounding fast electrons. When the
free electrons pass nearby protons or ions, the attractive electrical force decelerates them.
While braking, the electrons lose some of their kinetic energy, which is converted into
photons; this overall process is termed bremsstrahlung. This results in a broad continuum
of wavelengths. When the gas is extremely hot, i.e. millions of kelvins, the braking emission
peaks in the X-ray region. While synchrotron radiation from relativistic charged particles
that are spiraling in magnetic fields is responsible for most of the radio spectrum at longer
wavelengths, braking radiation, or bremsstrahlung, is the main source of emissions at the
shorter radio wavelengths.
With the advent of the space age, astronomers saw a new opportunity opening: observing
from outer space. They swiftly designed telescopes to be flown in Earth’s orbit above the
blocking atmosphere, gaining access to the shorter-wavelength domain, ultraviolet light,
X-rays and gamma rays. For visible or infrared light, astronomers use lenses made of glass
with a refractive index greater than one, or slightly curved glass mirrors. Light hits them at
near normal angles and is redirected. To capture X-ray images of cosmic sources and image
these, experimenters faced a challenge. X-ray photons have sufficiently high energies that
they penetrate the material and are absorbed instead of being reflected by it, as light is by
a metallic surface. To deflect X-rays, a completely new design of telescopic optics was
required. The successful approach was to have X-rays hit a metallic surface at a shallow
angle (from ten arcmin to a few degrees) and to deviate the rays only slightly, like a pebble
being skimmed at the surface of a pond and bouncing up. This led to grazing-incidence
metallic mirrors (Fig. 7.7).
An imaging X-ray telescope became a nested set of elongated metal cones follow-
ing the principles outlined by German physicist Hans Woltjer (1911–1978) in 1952.30 In
30 H. Woltjer, Spiegelsystems streifenden Einfalls als abbildende Optiken für Röntgenstrahlen, Annalen der Physik, 1952, Vol. 445, pp. 94–114.
14:06:02, subject to the Cambridge Core
.009
170
Part II – Images as Galaxy Discovery Engines
Fig. 7.7 Mirror elements of the Chandra X-ray Observatory showing the nested conic mirrors
focusing X-rays on the detector. Credit: NASA, CXC, D. Berry.
his autobiography, American astronomer Riccardo Giacconi described how the twentieth-
century collectors of X-rays had actually been designed in ancient times. While reading Lo
specchio ustorio, ovvero trattato delle sezioni coniche (The burning mirror, or treatise
on conic sections) published by the Italian mathematician Francesco Bonaventura (1598–
1647) in 1632, Giacconi was struck by the drawing of Figure XIII in the ancient work.
Stunned, he realized that he had replicated the principle of the collector that Bonaventura
had conceived some 330 years earlier.31 More startling, Bonaventura even referred in his
own work to a book with the same title written by the ancient Greek mathematician Apol-
lonius of Perga (c. 262–190 BC). Apollonius is noted for his work on conic sections, which
influenced many scholars of the Renaissance and later periods. He had shown that the main
geometrical curves (circles, ellipses, parabolae and hyperbolae, all names he created) were
obtained by mentally pushing a plane through a cone and contemplating the figure of their
intersection. So the twentieth-century technology for X-ray imaging is rooted in Antiquity.
X-ray astronomy has been an area of frontier work. The recent Chandra X-ray Observa-
tory has been used to observe many clusters of galaxies and galaxies. The observatory was
named for the Indian-born American astrophysicist Subrahmanyan Chandrasekhar (1910–
1995) who obtained the Nobel Prize in Physics in 1983 for his mathematical theory of
black holes. Then Giacconi was the recipient of the same award in 2002 for his work
31 R. Giacconi, Secrets of the Hoary Deep, A Personal History of Modern Astronomy, Baltimore: Johns Hopkins University Press, 1998, pp. 137–138.
14:06:02, subject to the Cambridge Core
.009
7. A Symphony of Waves
171
Fig. 7.8 Transformational Image: X-ray Contour Map in the 0.7–3.0 keV energy band of the
Giant Elliptical Galaxy Messier 87. The intensity and radial distribution of the very hot gas indicates
that the galaxy possesses a massive dark halo tens of trillions times the mass of the Sun. Credit: D.
Fabricant, M. Lecar, and P. Gorenstein (1980), The Astrophysical Journal. C
AAS. Reproduced with
permission.
on cosmic X-ray sources, in particular clusters of galaxies, and for his designs for X-ray
telescopes.
Measurements of the radio, optical and X-ray bremsstrahlung, of its shape and intensity,
provide key diagnostics of the temperature and density of the interstellar gas. Observed
from above the Earth’s atmosphere, many stars, nebulae and galaxies have been found to
be sources of X-rays (see Plate 7.6). Unexpectedly, the most spectacular sources of extended
X-ray emission have turned out to be large elliptical galaxies and clusters of galaxies
(Fig. 7.8). Clusters of galaxies are
huge conglomerates, containing from hundreds to thou-
sands of galaxies of all sorts, packed in a volume of a few million light-years in diameter.
For example, the Coma cluster, well known to us from the time of William Herschel, hosts
at least 1,000 bright galaxies (Fig. 1.4).
Except for a few cases, X-rays from the galaxy clusters do not come from the individual
galaxies. The whole cluster volume is shining in X-rays, as if galaxies in these clusters
are orbiting through a giant bath of extremely hot gas. The X-ray emission is produced
14:06:02, subject to the Cambridge Core
.009
172
Part II – Images as Galaxy Discovery Engines
by a colossal mass of ionized hydrogen and helium, with traces of heavy elements such
as iron, at temperatures of tens of millions of kelvins (Plate 8.1). By mapping the neutral
hydrogen of spiral galaxies in the inner parts of large clusters of galaxies, radio astronomers
have shown them to be severely depleted of their neutral hydrogen (see Plate 7.7). As they
revolve through the hot intracluster medium, cluster galaxies are shred, like trees losing
leaves in a windstorm.
Strangely, the hot gas is gravitationally bound by the strong gravitational potential and
is not boiling off the galaxy clusters. It has turned out that the intracluster gas is another
indicator of some invisible mass or dark matter. Above all, clusters of galaxies imaged by
X-rays have revealed more acutely than other astronomical objects the presence of dark
matter on a large scale. The total masses of large galaxy clusters as derived from a range of
indicators – stellar light, intracluster gas X-ray emission and galaxy kinematics – are huge.
They are of the order of 1013 to 1014 solar masses. Let us now plunge into the topic of dark
matter that has been mentioned so many times. Imaging with several techniques has been
a key tool in revealing the presence of this mysterious entity.
14:06:02, subject to the Cambridge Core
.009
8
Imaging the Invisible
Systems constructed by science resemble imposing edifices which, on
closer inspections, reveal a crack or two.
Stanley L. Jaki 1
In this darkness there is a small admixture, a few percent of the whole,