Astronomy › Section T · Galaxies and quasars › Topic T.6
Quasars
In 1963 a faint blue "star" turned out to be one of the most powerful objects in the universe, billions of light-years away. Quasars outshine whole galaxies from a region not much bigger than the Solar System.
Radio stars
After the Second World War, astronomers turned surplus radar equipment and new radio dishes to the sky and found it full of strong radio sources. The third Cambridge catalogue, the 3C catalogue of 1959, listed hundreds. Most sources were matched with galaxies or gas clouds, but a few coincided with faint, blue, star-like points of light. In 1962 Cyril Hazard and colleagues at Parkes in Australia timed the Moon passing in front of the source 3C 273, which fixed its position precisely, and it matched a 13th-magnitude blue star. No ordinary star gives off such strong radio waves. These became known as quasi-stellar radio sources, or quasars.
Schmidt's insight
The spectra of these radio stars made matters worse: they showed bright emission lines that matched no known element at their observed wavelengths. In 1963 Maarten Schmidt, at Caltech, studied the spectrum of 3C 273 and saw that several of its lines had the same spacing pattern as the Balmer lines of hydrogen, only shifted to wavelengths about 16% longer. Nobody had considered such a huge shift, because no star moves fast enough to produce one. The lines in other quasars were soon identified the same way, with even larger redshifts. Later, many quasars were found with no radio emission at all, so today the general name is quasi-stellar object, or QSO, still usually called a quasar.
Worked example: the redshift of 3C 273
In 3C 273's spectrum, hydrogen-beta, whose rest wavelength is 486.1 nm, appears at 562.9 nm. Find the redshift and estimate the recession speed. 1. Shift: \(\Delta\lambda = 562.9 - 486.1 = 76.8\) nm. 2. Redshift: \(z = \frac{\Delta\lambda}{\lambda} = \frac{76.8}{486.1} = 0.158\). 3. Simple speed: \(v \approx cz = 0.158 \times 3.00 \times 10^5 = 47{,}400\) km/s. 4. Check: every line in the spectrum must give the same \(z\), which is how Schmidt knew he had found hydrogen. 5. Interpret: with \(H_0\) about 70 km/s per Mpc, Hubble's law suggests roughly 680 Mpc, about 2 billion light-years. At this redshift \(v = cz\) already overestimates the speed slightly.
Very large redshifts
Many quasars have redshifts above 1, and the formula \(v = cz\) would give speeds faster than light, which is wrong. One common way to express a large redshift as an equivalent speed is the special-relativity Doppler formula:
which always gives speeds below \(c\). Strictly, though, a distant quasar's redshift comes from the stretching of space while its light travelled, as The expanding universe explained. So astronomers quote \(z\) itself and use a model of the expansion to turn it into a distance or a time. More than a million quasars are now known. As of late 2026, the most distant quasar on record, reported in July 2026 from the Euclid space telescope, is EUCL J172902.75+641018.1 at \(z = 7.77\), seen about 660 million years after the Big Bang. It overtook J0313-1806, at \(z = 7.64\), which had held the record since 2021.
Worked example: a quasar at z = 3
Lyman-alpha, an ultraviolet hydrogen line with rest wavelength 121.6 nm, appears in a quasar at 486.4 nm. 1. Redshift: \(z = \frac{486.4 - 121.6}{121.6} = 3.00\). 2. Equivalent speed: \((1+z)^2 = 16\), so \(\frac{v}{c} = \frac{16 - 1}{16 + 1} = 0.882\). 3. Check: the simple formula would give \(3c\), impossible, while 0.882 is below 1. 4. Interpret with the standard model of the expanding universe: since \(1 + z = 4\), distances have grown fourfold since the light set out, when the universe was only about 2.1 billion years old. 5. Note how far the ultraviolet line has moved: into the blue part of visible light.
Are they really so far away?
If quasars were nearby objects with some other cause for their redshifts, they would need far less power, and in the 1960s and 1970s a few astronomers argued exactly that. The evidence went against them. Quasars sit in clusters and groups of galaxies with the same redshifts. Gas clouds in intervening galaxies print absorption lines on quasar spectra, always at lower redshifts than the quasar. Most convincingly, images taken from the 1990s with the Hubble Space Telescope, above the blurring atmosphere, revealed faint host galaxies around quasars, both spirals and ellipticals, with the same redshifts. Quasars are the brilliant centres of distant galaxies.
Staggering power
3C 273 shines at about magnitude 13, bright enough for a good backyard telescope, yet its light has travelled for about 2 billion years. To appear that bright from so far, it must emit roughly 4 trillion times the Sun's luminosity in visible light, more than a hundred times the light of the whole Milky Way. Quasars also pour out ultraviolet and X-rays, and some radio waves. The most luminous known, J0529-4351, reported in 2024, is estimated at about 500 trillion Suns. Many quasar host galaxies look disturbed or are colliding with neighbours, a clue to what fuels them.
Small size from flickering
Many quasars vary in brightness by tens of percent over months, weeks or even days. That limits their size. If a region one light-month across brightened everywhere at once, light from its far side would reach us a month after light from its near side, smearing any change over at least a month. So an object that changes in time \(\Delta t\) can be no larger than about \(c\,\Delta t\). A quasar that varies over 10 days is at most 10 light-days, \(2.6 \times 10^{11}\) km or about 1,700 AU, across: about 30 times the width of Neptune's orbit, tiny for the power of a hundred galaxies.
A family of active galaxies
Quasars have quieter relatives. In 1918 Heber Curtis photographed a straight jet coming from the centre of the giant elliptical galaxy M87. In 1943 Carl Seyfert described spiral galaxies with unusually bright, compact nuclei whose emission lines are broadened by gas moving at thousands of km/s. Radio astronomers found radio galaxies with huge lobes on either side. All these are now called active galactic nuclei, or AGN: compact centres producing enormous energy in a tiny volume. Quasars are the most luminous AGN, and the next topic, Supermassive black holes, explains the single engine behind them all.
Common mistakes to avoid
First, using \(v = cz\) for large redshifts, which can give speeds above light. Second, thinking quasars are stars in our galaxy; they are the centres of distant galaxies. Third, assuming a quasar's small size means low power; flickering shows the opposite, huge power from a tiny region. Fourth, getting the size limit backwards: a change over 10 days means at most 10 light-days across, not at least. Fifth, thinking all quasars are radio sources; most are radio-quiet. Sixth, mixing up a line's rest and observed wavelengths when computing \(z\).
Key takeaways
Quasars were found as star-like radio sources; in 1963 Schmidt saw that 3C 273's lines were hydrogen lines redshifted by z = 0.158. Redshift is z = Δλ/λ; v ≈ cz works only for small z, while large redshifts are quoted as z and interpreted with a model of expansion. Host galaxies with matching redshifts prove quasars are distant. They emit trillions of times the Sun's luminosity, yet brightness changes over days to months show that the source is only light-days to light-months across. Quasars are the most luminous active galactic nuclei, alongside Seyfert and radio galaxies.
Flashcards
Flashcard
Extremely luminous, star-like centre of a distant galaxy
Test your understanding
Question 1 of 5
Lyman-alpha, rest wavelength 121.6 nm, appears in a quasar's spectrum at 608.0 nm. What is the quasar's redshift?
Question 2 of 5
For a quasar with redshift z = 3, what equivalent speed does the relativistic Doppler formula give?
Question 3 of 5
A quasar doubles in brightness over 4 days. What does this imply about the region producing the light?
Question 4 of 5
What evidence best showed that quasars lie at the great distances their redshifts imply?
Question 5 of 5
What did Maarten Schmidt recognise in 1963 about the spectrum of 3C 273?
