Radio Astronomy: Observing the Invisible Universe PDF Lecture 10 - Pulsars

LECTURE 10 Pulsars: Clocks in Space 104  Radio Astronomy: Observing the Invisible Universe T his lecture is about what you find when you look for a radio source that is twinkling. For stars, twinkling means that the light is flickering a bit and m...

LECTURE 10 Pulsars: Clocks in Space 104  Radio Astronomy: Observing the Invisible Universe T his lecture is about what you find when you look for a radio source that is twinkling. For stars, twinkling means that the light is flickering a bit and moving from side to side, and often, if you look at a star close to the horizon on a clear night, you can even see it changing colors as it dances around. Scintillation „„ Blobs in the Earth’s turbulent atmosphere act like little lenses or prisms, and the wind blows them past the star, making the star dance and twinkle. The technical term is scintillation. „„ The Moon, Sun, and planets don’t scintillate, because they have too large an angular size. The light from one point of the planet scintillates, but the light from another part of the planet scintillates in a different way, and all the different scintillations add up to a fairly smooth image, though a bit fuzzy. „„ It takes 2 things to make a radio source scintillate. First, the radio source has to have a small angular size. It doesn’t have to be small itself, but if it is big it has to be very far away to appear small in the sky. Objects with a negligible angular size are referred to as point sources. They appear in the sky as a point without any structure. A type of radio galaxy called a quasar is very bright and very distant and can look like a point source to us. „„ The second thing you need to make a point source scintillate is lumps of ionized gas to refract the radio waves and bend them. „„ The signal from a quasar is turned into a lumpy image, and as the ionized gas clouds move around, we alternately see bright and then faint radio emission. That’s the signal of scintillation—a change in brightness with time as the radio source is focused and defocused on our telescope. 105  Lecture 10 Pulsars: Clocks in Space „„ At radio wavelengths, the strength of scintillation increases with the wavelength squared. Longer wavelengths (lower frequencies) scintillate much more than higher frequencies. „„ There are many regions of ionized gas between us and a distant quasar, but close to home there’s a pretty powerful lumpy ionized medium that radio waves have to traverse: the solar wind. „„ The Sun is putting out ionized gas every moment, and it flows past Earth. This is the solar wind. It arises from magnetic storms on the surface of the Sun. It has lumps, and while most of the time it doesn’t affect our radio measurements, at low frequencies it can make point radio sources scintillate. 106  Radio Astronomy: Observing the Invisible Universe Hewish’s Telescope „„ Antony Hewish built a telescope to discover quasars by their scintillation from blobs in the solar wind. At the time, in the mid- 1960s, quasars were mysterious beasts, and they still are in many ways. „„ If you’re trying to understand the properties of something, you need large samples of objects or phenomena to separate the general from the particular. Hewish wanted to discover lots and lots of quasars, and in the process, he and his team discovered something else quite unexpected. „„ Because the scintillation is strongest at low frequencies, Hewish designed his telescope to work around 80 megahertz, at a wavelength of 3.7 meters, just below the FM band in the United States. „„ Hewish ended up covering about 4.5 acres—nearly 60 tennis courts—with his antennas. The antennas were just copper wire, strung between wooden poles. „„ Hewish’s graduate student, Jocelyn Bell Burnell, was involved in the construction and operation of the telescope. „„ When the Hewish telescope was finished, the observations began. The data were recorded on a chart with pen and ink. Because the scintillations are rapid, the paper moved fast and the pen could record fluctuations as short as 1/10 of a second. The telescope produced 4 beams on the sky in different directions, and the data were filtered in various ways, so there were about 100 feet of chart recording that came out every day—and had to be examined every day, by hand. 107  Lecture 10 Pulsars: Clocks in Space The Discovery of Pulsars „„ Radio telescope sidelobes can be reduced but not eliminated entirely, even from big modern dishes. The result is that radio telescopes have some sensitivity to signals coming from all directions, not just where the main beam is pointed. „„ Bell Burnell and Hewish’s telescope ran day and night, 7 days a week, and Bell Burnell had to sort through all that data. She discovered scintillating sources, which meant new quasars, and also lots of examples of terrestrial interference. There was also a strange faint signal—some “scruff” on the chart recordings—that didn’t look like scintillation and didn’t look like interference. „„ After a while, Bell Burnell recognized that one particular patch of scruff had reappeared several times from about the same fixed direction in space. That was peculiar. Something was producing regular radio pulses. It didn’t seem fixed to the Earth but appeared earlier each day, like something fixed to the stars. „„ Could this be some signal from another civilization? They looked at it with another radio telescope at Cambridge, and after a few fumbles detected it. So, it was not being generated in their own equipment. „„ Study of thousands of feet of charts showed a few other directions with pulsing radio sources. They had discovered something new: pulsars. „„ They reported their results in a paper in February 1968. By late spring, more pulsars had been discovered. Pulsars were not only real, but there seemed to be quite a lot of them. But what were they? „„ We now understand that a pulsar is the remnant of a massive star that exploded as a supernova at the end of its life. In order for a star to become a supernova, it has to have 8 times the mass of the Sun. These massive stars are short-lived, and upon exploding, their interior collapses and forces electrons onto protons to form neutrons, 108  Radio Astronomy: Observing the Invisible Universe thus producing a neutron star. A cubic inch of a neutron star—about This ball of matter is incredibly the size of a sugar cube—contains dense—as dense as matter can the same mass as Mount Everest. get. „„ A neutron particle left by itself in free space decays in just a few minutes into a proton and electron, but bound up in a bundle with others, it’s quite stable. „„ The neutron star has a diameter of just a few kilometers. It has a mass between 1 and 2 times the mass of the Sun, so that’s a lot of matter compressed into a tiny volume. „„ A star has a magnetic field, just as the Earth does. That magnetic field is threaded through the interior of a star, and when the star collapses, it drags the field down with it, amplifying it enormously. We’re left with a collapsed stellar core that’s rotating rapidly because of conservation of angular momentum. „„ Earth’s north magnetic pole is not at the North Pole. It’s the same for a neutron star; a neutron star’s magnetic pole can be offset from its rotational pole. The magnetic field produces strong radio emission, and if the magnetic pole sweeps past us, we see it as a radio pulse. And we call it a pulsar. „„ The pulses from a pulsar are extremely regular. The first pulsars that were detected had periods of about 1 second, meaning that the neutron star was spinning around about once a second. But now we have pulsars whose periods span from a few seconds to a few thousandths of a second. 109  Lecture 10 Pulsars: Clocks in Space „„ The pulsars are losing energy as they rotate, so they are slowing slightly, but very predictably. Pulsars slow down and eventually die. „„ The discovery of pulsars initiated a flurry of observations around the world. Although pulsars were discovered in observations at 80 megahertz, in modern times they’re most commonly observed between 300 megahertz and 3 megahertz using large single dishes. „„ In a typical pulsar pulse, there’s a main pulse and sometimes a secondary pulse, or interpulse. If we think about the pulsar as having 2 magnetic poles, then it makes sense that sometimes we see emission from both. The average pulse shape is very stable, but any one individual pulse may show large variations in intensity and shape from the average. „„ How do pulsars produce radio radiation in narrow beams? Nearly 50 years after their discovery, there’s still lots of mystery about the basic pulsar emission mechanism. The emission is certainly nonthermal. It’s broadband, and its intensity increases rapidly to lower frequencies, opposite of the Planck curve—a clear sign of nonthermal emission. But what exactly makes it? „„ We know that energetic particles in magnetic fields produce nonthermal emission. Pulsars have very strong magnetic fields, far in excess of anything that we could produce on Earth. All we need is particles. It’s likely that charged particles from the neutron star’s surface are accelerated by intense electric fields. „„ Spiraling in the magnetic fields, these particles emit high-energy photons, which in turn are converted into electrons and positrons by the intense magnetic field. It is these electrons and positrons that produce the radio emission we observe. This sounds plausible, but many details still don’t fit this picture. „„ The first observers of pulsars discovered that the arrival time of pulses depended on the frequency that was observed. A given pulse 110  Radio Astronomy: Observing the Invisible Universe arrived first at the higher frequencies and then progressively later at lower and lower frequencies. But all frequencies left the pulsar at the same time. The delay comes from interstellar electrons. „„ Ionized gas between us and the pulsar delays the pulse arrival. The size of the delay increases as frequency decreases. This process is called dispersion. It depends on the amount of ionized gas between us and the pulsar. „„ This is important for 2 reasons. First, manmade signals from Earth don’t travel through the interstellar medium to reach us, so their pulses are not dispersed unless they’re transmitted that way. Looking for the signature of dispersion gives us one way to discriminate between pulses from space and terrestrial interference. „„ Second, the amount of dispersion gives us an estimate of the distance to the pulsar. The farther the pulsar, the more it should be dispersed. There are many uncertainties, but it’s so difficult to get distances in radio astronomy that anything that gives us something even approximating a distance is grasped like a lifesaver. Suggested Reading Graham-Smith, Unseen Cosmos, chap. 6. Hewish, “Pulsars and High Density Physics.” Kellermann and Sheets, eds., Serendipitous Discoveries in Radio Astronomy. Verschuur, The Invisible Universe, chap. 8. 111  Lecture 10 Pulsars: Clocks in Space Questions to Consider 1. A pulsar has what is called a duty cycle, which is the fraction of the time, in every pulsar period, when the pulse is actually on. A typical duty cycle is less than 10%, meaning that if a pulsar has a period of 1 second, the pulse is seen for less than 0.1 seconds. What does this tell us about the size of the area on the pulsar that emits the pulse? 2. Hewish’s array could produce 4 beams in different directions because the individual dipoles have a very broad beam pattern. Could you produce 4 beams in very different directions if the array were made up of dishes? 3. Besides the slowdown of a pulsar’s spin because of energy loss, other factors could make a pulsar appear to slow down or even spin up. What could those be? (Answer provided in the next lecture.) 112  LECTURE 11 Pulsars and Gravity 113  Lecture 11 Pulsars and Gravity I n this lecture, you will learn how to use pulsars—spinning tops with the mass of the Sun that give off a radio pulse every second or so— to confirm Einstein’s prediction of gravitational radiation. You will also learn how to confirm Einstein’s theory by measuring a pulsar’s Doppler shift. The Doppler Shift of Pulsars „„ The Doppler shift describes how a wave is stretched or compressed as the source of the wave moves toward us or away from us. A pulsar can have a Doppler shift. „„ We can think of the pulses as peaks of a wave; we just Measuring Doppler Shift don’t have the rest of the wave. And those peaks will You can measure be shifted by any motion. the Doppler shift of the ticks from „„ We can measure a Doppler any clock. To do shift from a pulsar—sort of. this, find a small The rub is that with atoms portable clock and molecules, we have a that gives off clearly defined rest frequency loud ticks. Tie it set by quantum physics. We to a meter-long can measure velocity directly string and whirl it around your because we know the rest head. frequency. „„ But for a pulsar, its actual You will hear a steady beat because rotational period can be the clock is not changing its distance anything from seconds to from you, but someone standing milliseconds. All we can nearby will hear the increase and measure is the observed decrease of the clock tick rate as the period, which is the clock on the string approaches them combination of its rest period and then recedes from them. and any Doppler shift. 114  Radio Astronomy: Observing the Invisible Universe „„ The terms “period” and “frequency” describe the same thing— namely, some sort of repetitive phenomenon—but they are the inverse of each other. Something that happened 10 times a second has a frequency of 10 hertz and a period of 0.1 seconds. Something that happens every 2 seconds has a period of 2 seconds and a frequency of 1/2 hertz. „„ The Doppler shift is in play for every pulsar in the sky, only we can’t use it because we don’t know the actual period of the pulsar. We can’t use the Doppler shift unless we see the period change. „„ In fact, the period we observe of every pulsar changes all the time because we’re observing pulsars from a moving platform— the Earth—which spins on its axis toward the east and away from the west. So, as we observe a pulsar rising, we see that its period increases, and as it sets, its period decreases. „„ The Earth also goes around the Sun at 30 kilometers per second or so. If we measure a pulsar in the spring of the year and then later in the fall, we will have seen a change in its period because of the Earth’s revolution around the Sun. „„ But we can correct for all of these motions and measure the Doppler shift of pulsars so precisely that we even need to make corrections to pulse arrival times because Earth’s motion is influenced by the gravity of the planet Jupiter. „„ This Doppler effect is stronger in some parts of the sky than others. Because we know the Earth’s velocity very accurately, this Doppler shift can be used to locate the position of pulsars quite precisely. How Do Pulsars Get Their Spins? „„ All stars rotate. A star that goes supernova blows off its outer layers, but the inner parts collapse, preserving their angular momentum and spinning up as they fall. 115  Lecture 11 Pulsars and Gravity „„ The result is a neutron star spinning quite fast. The neutron star in the Crab Nebula currently has a period of 33 milliseconds; it rotates more than 30 times each second. At its birth, it certainly had an even shorter period, around a few milliseconds. „„ As pulsars age, they slow down and their magnetic field weakens and period increases. Eventually, they cease being observable as pulsars. Normal Pulsars Young pulsars spin rapidly and possess a Old pulsars have slower periods and weaker Magnetic Field Strength (Gauss) strong magnetic field. 1013 magnetic fields. Pulsar’s life journey 1012 e h lin at De The pulsar’s emission becomes too weak for 11 10 Earth-based detection Extinct pulsars the pulsar is extinct. live here! 0.1 1 Spin Period „„ Millisecond pulsars are old. They have run out of steam and dropped out of sight. But then, for some of them, an amazing thing happens: They get rejuvenated and spin up—a lot. Their magnetic field is still weak, so their radio emission is weak, but they are spinning fast. „„ These pulsars were formed in a binary system, which is where a pair of stars is in mutual orbit around each other. The most massive star went supernova and became a neutron star. That neutron star did the pulsar thing, and over time, its magnetic field weakened and its rotation slowed. 116  Radio Astronomy: Observing the Invisible Universe „„ But then, as the other star evolved, it started losing matter and dumping some of the matter on the neutron star. That spun the neutron star back up, to even faster periods than it had before. It made a millisecond pulsar. „„ So, a pulsar can get its spin from either its creation in a supernova explosion or by mass transfer from a companion. „„ All pulsars are neutron stars. But not all neutron stars are pulsars. To be a pulsar, a neutron star has to have a magnetic field strong enough to produce radio emission, and that radio beam has to sweep across the Earth. Because the beams are relatively narrow, we probably see only 10% of the active pulsars in the Milky Way. How Do We Measure Pulsars So Precisely? „„ There is radio noise in all of our measurements, and it comes from our electronic equipment, the galactic nonthermal background, the atmosphere, and other sources. The challenge of radio astronomy is how to dig the signal out of the noise. „„ But we have one important thing going for us: Noise is truly random. It fluctuates from instant to instant. If you average 2 noisy signals, the noise in one tends to cancel the noise in the other. „„ The amplitude of the noise—that is, the height of the noisy signal— decreases as the square root of the number of samples that you have. If you average 4 noisy signals, the noise in the average has decreased by a factor of 2. „„ We are able to detect pulsars, and almost every other signal in radio astronomy, because as we add data and average, the noise level decreases while the signal stays constant. 117  Lecture 11 Pulsars and Gravity „„ We can look at the pulsar and measure the pulse and average down the noise in time, but because pulsar radio emission is broadband, we can also measure it at a number of frequencies and average over frequency as well as time. „„ Because of interstellar dispersion, the pulses arrive at slightly different times at different frequencies. We have to shift the samples before adding them lest we blur out the signal. But we can measure the dispersion quite accurately, and the shift is easy. „„ Using signal-averaging techniques, we can also measure the period derivative, which is the technical name for the spin-down rate, and also the dispersion measure. This allows us to predict the arrival time of pulses many months into the future. „„ But how do we determine the period in the first place? One way is to average the data using a guess for the period. If we are off, the pulse is blurred because data taken at different times do not add in phase. If we are spot-on, the pulse appears at its sharpest. In detecting new pulsars and measuring their rotational periods, quite a lot of this searching goes on, and it’s computationally intensive. Confirming Einstein’s Theory of Relativity „„ In 1972, Joseph Taylor, a young faculty member at the University of Massachusetts in Amherst, wrote a proposal to the National Science Foundation seeking funds to search for new pulsars using Puerto Rico’s Arecibo radio telescope, which is 1000 feet across and has no rival in its sensitivity to pulsars. 118  Radio Astronomy: Observing the Invisible Universe Arecibo telescope „„ At the time of Taylor’s proposal, there were about 100 known pulsars, but no pulsar had been found in a binary system. Taylor got the grant, began to buy the hardware necessary to build a digital pulsar detection system, and recruited a graduate student, Russell Hulse, as a collaborator. „„ They did their search at frequencies around 400 megahertz. At every point in the sky where they took data, their programs searched more than 500,000 combinations of period, dispersion, and pulse shape seeking a detection. „„ If you take enough samples of random noise, every once in a while the noise combines to look like something real. But it’s not. Given that each spot on the sky would be analyzed in half a million different ways, Hulse set his threshold for detection at 7 standard deviations. This means that a signal would have to be 7 times the expected noise level before the computer would flag it as possibly being real. 119  Lecture 11 Pulsars and Gravity „„ Taylor and Hulse took their system to the Arecibo telescope and began searching the sky for new pulsars. The observations were performed on and off for more than a year. In the end, they found 40 new pulsars. One in particular caught their eye. It had a period of around 60 milliseconds, which made it the second-fastest pulsar then known, second only to the Crab pulsar. „„ When Hulse reobserved this particular pulsar several times, the pulsar’s period couldn’t be pinned down. He set up a new data analysis scheme with faster time sampling of the incoming signal. The new data showed that this pulsar’s period seemed to be changing in a regular way. „„ Then, Hulse realized that the changes could arise from a changing Doppler shift of the pulsar, which meant that it must be in orbit around another star—a pulsar in a binary system. „„ In short order, Joe Taylor arrived at Arecibo carrying new equipment that made study of this pulsar much easier. Here’s what we now know about this amazing system. 120  Radio Astronomy: Observing the Invisible Universe „„ There are 2 neutron stars in orbit around each other. Both have about 1.4 times the mass of the Sun, and one of them is a pulsar with its radio beam intersected by Earth. They are quite close together; at their closest, they are only a few times farther apart than the Earth and Moon. They orbit around each other every 7.7 hours. That rapid orbit produces a large Doppler shift in pulse arrival times. „„ In classical physics, there is nothing to keep 2 objects from orbiting each other forever. But Albert Einstein changed all that with his theory of relativity. Under this theory, space is curved around massive objects, and as they interact gravitationally, they radiate gravitational waves. Eventually, the 2 neutron stars will merge as their orbit decays through radiation of gravitational waves. „„ The pulsar found by Hulse and Taylor confirmed this theory and provided direct experimental proof that changes in gravity travel at the speed of light. For this, they were both awarded the Nobel prize in Physics in 1993. „„ Pulsars can be used to probe the curvature of space predicted by Einstein’s theory of relativity. Every object distorts the space around 121  Lecture 11 Pulsars and Gravity it at least a little. When the pulsar is nearly exactly behind the star, its pulses have to travel an extra distance because of the distortion of space. When the pulsar is behind its companion, the pulses are delayed by the extra path they need to travel. This is a direct measurement of the curvature of space. Suggested Reading Graham-Smith, Unseen Cosmos, chap. 7. Hulse, “The Discovery of the Binary Pulsar.” Taylor, “Binary Pulsars and Relativistic Gravity.” Questions to Consider 1. Although all frequencies of a pulsar are emitted at the same time, the arrival time of a pulse on Earth depends on frequency because of delay by interstellar electrons (the technical term is dispersion). And the delay is proportional to frequency to the −2 power; a decrease in frequency by a factor of 2 increases delay by a factor of 4. Because the interstellar medium between a pulsar and us changes as ionized clouds drift past, the dispersion changes and pulse arrival times change. How might this be detected and corrected? 2. Searches for new pulsars are typically done at frequencies below 1 gigahertz, where pulsars are brightest, but accurate timing of pulsars is done at frequencies above 1 gigahertz, even though the pulsars are weaker there. Can you understand why, in view of the previous question? 3. Because we can measure the period of a pulsar extremely accurately and we have to compensate for the Doppler shift of the period arising from the Earth’s motion around the Sun, does this give us information on possible extra planets in the solar system, or even a dark companion star to the Sun? 122  Lecture 13 The Big Bang: The Oldest Radio Waves T his lecture is about the radio signals from the big bang. These radio waves are very old. They’re the oldest electromagnetic radiation in the universe. These old radio waves tell a story from a time when the universe was very young and very simple. Radio Noise „„ Objects in our world constantly emit radio waves—it’s thermal emission of radio waves. This means that our radio receivers are always putting out some sort of signal, even if nothing is coming in from the antenna. „„ Consider a camera. If you block the lens with something at room temperature—such as your hand—there will be no picture, because nothing in the camera gives off light. „„ On the other hand, if you block the input to a radio telescope receiver, the receiver would see blackbody emission from whatever was blocking it. If the object was at room temperature, around 300 kelvins, the radio emission would be quite bright. Objects at 300 kelvins don’t give off visible light, but they do emit radio waves. „„ Sources of radio emission—radio noise—that we have to contend with in our radio telescopes include noise from the receiver, antenna and horn, dish, ground, atmosphere, and unrelated celestial emission. „„ The signals that we’re trying to detect usually sit on top of a sea of unrelated emission. For this reason, radio-astronomical measurements are almost always differential, which means that we’re almost always looking at the difference of signals rather than the absolute signal. „„ Observations of pulsars give a great example. The pulse sits atop a large amount of radio emission—the so-called system noise, which 136  Radio Astronomy: Observing the Invisible Universe comes from the receiver, atmosphere, ground, and so on. Usually, our signals of interest are much less than 1% of the total system noise. „„ Everything except the pulsar is fairly constant with time. Therefore, we can just take the average intensity over the entire pulse period and subtract it from all the data to eliminate the radio signals that aren’t of interest. This is called time switching, because we are making a comparison between signals that arrive at a different time—pulse on compared to pulse off. „„ This technique can’t be used for most celestial radio sources because they’re pretty constant in time, but we can always do what is called position switching, where we compare the radio emission in one direction with what we get when we point the telescope in another, nearby direction. If we switch quickly enough, the receiver noise, atmosphere noise, and so on will remain constant and our difference will reflect the different radio emission that comes in from the 2 parts of the sky. 137  Lecture 13 The Big Bang: The Oldest Radio Waves „„ Another technique is frequency switching. If you are looking at a spectral line that covers a small range of frequencies, you can compare the signal at the frequency of the line with what the telescope receives at adjacent frequencies. The difference between them, in theory, should leave you only with the line emission. 138  Radio Astronomy: Observing the Invisible Universe „„ If you are switching in time, position, or frequency, you don’t have to know what the different parts of the system are doing. You don’t have to know the atmosphere, receiver, ground, etc., very well because their emission all cancels out. „„ There’s another comparison that can be done: load switching, where we are switching the receiver between a dish looking at the sky and some source of radio noise located just in front of the receiver. The load—that is, the noise source—should be a blackbody whose temperature we can measure precisely. If we can look at a few different temperature loads, we can determine the noise that the receiver adds to the system. „„ With the switch pointed at the load, we have the radio emission from the receiver plus the blackbody emission of the load. With the switch thrown toward the antenna, we have the same emission from the receiver, but instead of the load, we have something from the antenna, sky—everything else. „„ We can compare what’s coming in through the telescope to what comes from the load and say that the telescope gives us a certain antenna temperature. In other words, the radio emission from everything in front of the receiver can be related to one number: the temperature of a blackbody that would have given the same amount of signal. 139  Lecture 13 The Big Bang: The Oldest Radio Waves „„ This is one reason why it’s very convenient for radio astronomers to pretend that the signals we observe come from thermal objects— blackbodies—even if they don’t. „„ The intensity of radio spectral lines is usually given in units of temperature—antenna temperature or a related temperature. It is the temperature that a blackbody would have if it were producing the identical amount of emission as the spectral line does at that frequency. „„ In the same way, we can describe the receiver noise as having a certain antenna temperature. This is the temperature that a blackbody would have to supply the same amount of radio emission, or radio noise, as the receiver does. „„ In many instances, the antenna temperature is very close to the physical temperature of an object, or it’s easily related to it, so it’s not just an arbitrary intensity scale. „„ The system temperature is the amount of noise in our radio telescope system added by everything but what we want to look at in the sky. It is an amount of noise expressed as the temperature that a blackbody would have to produce that amount of noise. The Cosmic Microwave Background „„ In 1964 at Bell Telephone Laboratories in Holmdel, New Jersey, there was a very peculiar looking antenna, called a horn reflector, that no one was using because the project that it was built for—Project Echo, a very early attempt to communicate over long distances by way of a satellite in orbit—had come to an end. 140  Radio Astronomy: Observing the Invisible Universe „„ Horns funnel the radio waves down to the point where an antenna probe can pick them up. Horns have simple beam patterns. They don’t have many sidelobes because there is no blockage. And they don’t have legs holding up the receiver in front of them, so the telescope beam is quite clean. „„ Most radio telescope dishes are based on the parabola, which is symmetric around a central line and brings incoming plane waves to a single focus. Receivers often have a small horn to collect the signal and couple it to the antenna. The horn reflector design uses just a portion of a parabolic surface, and instead of a small horn in front of the receiver, it extends the horn and makes it part of the structure. 141  Lecture 13 The Big Bang: The Oldest Radio Waves „„ At the Bell Labs facility in 1964, Arno Penzias and Robert Wilson were young scientists interested in making very precise radio- astronomical measurements. Wilson, in particular, wanted to see if there was a faint radio halo around the Milky Way that might have been unintentionally removed in position-switched observations. „„ They had a new load that they could use—a cold load. Penzias designed it. It’s called a cold load because it uses liquid helium to produce very cold blackbody emission. Liquid helium boils at a temperature of about 4 kelvins, and they calculated that because of heat leakage into the load, it would appear to their receiver as a blackbody at about 5 kelvins. So, the load would give 5 kelvins. „„ What did they expect when they switched from the load to the antenna? The atmosphere would give about 2.5 kelvins, and there might be 1 kelvin from losses in the antenna itself. At their frequency of 4 gigahertz, any galactic synchrotron emission should have faded to below 1/10 th of a kelvin, and from the ground or other sources, they expected less than 1/10 th of a kelvin of radio noise. The antenna plus atmosphere, and so on, should be around 4 kelvins—20% cooler than the 5-kelvin load. 142  Radio Astronomy: Observing the Invisible Universe „„ They flung the switch to the cold load and measured the output of the receiver; then, they flung the switch to the antenna. The antenna was not cooler than the load. It was hotter than the load. „„ Any kind of error with the load would have increased the noise coming out of it and made the antenna seem cooler than the load. But they got the opposite result. The antenna, and whatever was coming into the antenna, was hotter than they expected. It was not a huge difference, but it was a difference, and it was different in a way that was not easy to explain. „„ Penzias and Wilson had detected a completely new source of radio emission that came from all directions in the sky. They estimated a temperature of 3.5 ± 1 kelvins. It’s the remnant of the big bang that marked the origin of the universe. „„ They published their findings in a short paper in The Astrophysical Journal. This radiation is now generally referred to as the cosmic microwave background. It’s also often called relic radiation because it is the relic of the big bang. „„ In the first few sentences of their paper, they made the key points: It’s isotropic, meaning that it comes from all parts of the sky. So, it’s not just part of the Milky Way. It’s unpolarized. Emission from a blackbody has no preferential polarization, unlike synchrotron emission or manmade transmissions. It’s free from seasonal variations. This makes it unlikely that it came from any particular object, which would rise and set at different times throughout the year. „„ Meanwhile, a group at Princeton University was working on the problem of the origin of the universe. They were led by Robert Dicke, 143  Lecture 13 The Big Bang: The Oldest Radio Waves who came up with the idea of switching a receiver between the antenna and a load. The technique is often called Dicke switching. „„ While Penzias and Wilson were still scratching their heads over their result, Dicke had a group at Princeton building a small horn antenna to work at a 3-centimeter wavelength to try and detect the signal. The groups learned of each other’s work, got together, and 2 short papers were published in The Astrophysical Journal: one from Penzias and Wilson reporting their detection and the other from Dicke and collaborators reporting the interpretation that it was blackbody radiation from the big bang. „„ A few Soviet scientists were also hot on the trail of the background radiation, and in retrospect, the radiation had shown up indirectly in the excitation of molecules observed at optical wavelengths. „„ When the Princeton group finished their antenna and measured the radiation, they got 3 ± 0.5 kelvins, in agreement with the Penzias and Wilson result of 3.5 ± 1 kelvins. And because the Princeton group worked around 10 gigahertz while the Bell Labs group worked at 4 gigahertz, the agreement meant that the spectrum was thermal blackbody emission. „„ For their discovery, Penzias and Wilson were awarded the Nobel prize in physics in 1978. The discovery of thermal radio emission from all parts of the sky pretty much clinched the big bang as our theory of the origin of the universe. The radiation has been measured quite a bit since then and has passed every test. „„ The COBE (Cosmic Background Explorer) satellite, launched in late 1989, measured the blackbody spectrum of the background radiation quite precisely and searched for fluctuations across the sky. Although the background radiation looked uniform across the sky when observed with the Bell Labs horn reflector antenna, it has to have structure at some level. 144  Radio Astronomy: Observing the Invisible Universe „„ This is because at the time that the radiation was created, the universe was already beginning to get lumpy. Gravity was drawing some regions together, and those slight density fluctuations should have left a mark on the radiation. Those density COBE satellite fluctuations later grew into galaxy clusters and into all the structures that we see around us. If we could measure their imprint on the cosmic background radiation, we’d be seeing the earliest-possible signs of structure in the universe. „„ COBE showed the blobby structure of the universe that evolved into all the amazing things that we see today. The structures have been analyzed extensively, and it turns out that they don’t make sense unless the universe is dominated by dark matter. 145  Lecture 13 The Big Bang: The Oldest Radio Waves „„ Work on the microwave background continues today. From the remnant radiation of the big bang, we know the age of the universe: 13.799 billion years, with an uncertainty of only 38 million years—an accuracy of 99.5%. Suggested Reading Kellermann and Sheets, eds., Serendipitous Discoveries in Radio Astronomy. Mather, “From the Big Bang to the Noble Prize and Beyond.” Mather and Boslough, The Very First Light. Smoot, “Cosmic Microwave Background Radiation Anisotropies.” Wilson, “The Cosmic Microwave Background Radiation.” Questions to Consider 1. Is it possible for something in interstellar space that is bathed by the radiation from the big bang, a 3-kelvin blackbody, to be cooler than 3 kelvins? 2. When we look back in space and time to a time when the universe was smaller than it currently is, do you think that the radiation from the big bang was hotter than 3 kelvins? 146