AstroTalk: Behind the news headlines of July 2013
Richard de Grijs (何锐思) (Kavli Institute for Astronomy and Astrophysics, Peking University)
The Universe in colour
Astronomical images, particularly those published in magazines like the Amateur Astronomer, usually exhibit beautiful colours that often really help you understand the intricate details of what you are looking at. In fact, this aspect of the “beauty” of the Universe was an important driver for me to pursue a career as a professional astrophysicist – and I am often still in awe of the outstanding beauty of many objects that are scattered all over the sky. Having been a successful professional astronomer for more than 20 years now has not diminished that basic human response. Of course, in the mean time, I have come to realize that the “raw” images professional astronomers obtain with the telescopes they use to study the Universe are anything but colourful – although they may still be beautiful!
When you use a professional telescope to take images of astronomical objects, most commonly these images are obtained through certain filters – usually coloured pieces of glass that are placed in the light path leading to your detector. These filters are designed such that only a small fraction of the incoming light is transmitted, while most of the light is actually absorbed. In essence, what this means is that your detector – usually a CCD camera similar to the photo camera in your mobile phone, although of much better quality1 – only records a limited range of wavelengths. Images coming straight from the telescope are, therefore, essentially monochromatic: they don’t contain much colour information, but that is exactly what we want as professional astronomers!
While you might be awestruck by the beautiful red or blue colours of an object in the sky, as scientists we want to know exactly how “blue” or how “red” that object is: we need a way to quantify the colour, so that we can begin to understand the physical processes that lead to the objects having certain colours. If you are in the fortunate position to be outside the city in a dark place and the sky is clear, have a careful look at some of the bright stars. You may be surprised to find that stars are not all “white”, but some of them are red (for instance, the star Betelgeuse in the constellation of Orion) while others are decidedly bluer. This offers us an immediate insight into the type of star and even the main physical processes that play a role. Young, massive stars are usually very hot and they tend to be white or blue. Just think of the colour of the hottest part of a flame: this also tends to be white or blue. Older stars, often those that are getting towards the end of their lifetimes, are yellow or red; you may have heard of “red giant” stars, which represent the end stages of stars like the Sun.
To quantify the colour of an object in the astronomical sense, we usually observe
1 See my article in these pages (December 2011, p. 20) about the new gigapixel camera on board the European Gaia satellite. it through a number of different filters, which each have their own transmission properties. For instance, we can employ filters that only allow a small fraction of the bluest light, the reddest light, or any other wavelength range to which our detector is sensitive to pass through. In optical astronomy, a commonly used filter that is designed to only allow blue light to pass through is called a B filter (“B” for blue), while red filters are usually referred to as I filters (“I” for infrared). Having obtained observations of the same astronomical objects through both of these filters, we can define an astronomical colour, in this case a “B–I” colour. Without going into the nitty gritty of how we calculate this colour, it suffices to know that we determine the intensity (“brightness”, usually referred to by the letter F, for “flux”) of the object in both the B and I filters, and we then take their ratio, FB / FI. This procedure forms the basis for calculating the B–I colour, which thus allows us to quantify how “blue” or how “red” our object is. In other words, we get a value for the colour, which can now be compared with the values we have obtained for other objects or with theoretical predictions based on our detailed physical understanding of the nature of the objects we are studying.
Now, you might wonder how this relates to the beautifully coloured images you see in this magazine and elsewhere in the popular astronomy press. These images are usually referred to as “true colour” images, but they are not really “true” in their colour palette: the colours have been significantly enhanced so that the human eye can actually appreciate the colour differences. Each image is captured using at least three different filters: at minimum a red, a green and a blue filter. We then combine those three (or more) images into one, using a Technicolor process pioneered in the 1930s. The same process occurs in digital photo cameras, except that in your camera, it’s done automatically. Such true- colour images can help us understand how different colours vary across an object (like a galaxy or gaseous nebula), and this process also allows us to visualize observations obtained at wavelengths that the human eye is not sensitive to, such as in the ultraviolet or infrared domains. Addition of those colours isn’t done at random, though. On the contrary, astronomers are very careful to stay as true to nature as they can. Thus, in full-spectrum images, the details that correspond to infrared light will have the reddest colour and the details corresponding to ultraviolet will have the bluest.
The first true colour measurement of an exoplanet
So, where does this leave us in our interpretation of astronomical data? Well, consider a news item I came across last month (July 2013). Astronomers using the Hubble Space Telescope have, for the first time, determined the true colour of a planet orbiting another star. If seen up close this planet, known as HD 189733b, would be a deep azure blue, reminiscent of Earth’s colour as seen from space.
But that’s where the similarities end. This “deep blue dot” is a huge gas giant planet orbiting very close to its host star. The planet’s atmosphere is scorching with a temperature of over 1000 °C, and it rains glass, sideways, in howling 7000 km/h winds. At a distance of 63 lightyears, this turbulent alien world is one of the nearest exoplanets to Earth that can be seen crossing the face of its star. It has been intensively studied by Hubble and other telescopes, and its atmosphere has been found to be dramatically changeable and exotic, with hazes and violent flares. Now, this planet is the subject of an important first: the first measurement of an exoplanet’s visible colour.
“This planet has been studied well in the past, both by ourselves and by other teams,” says Frédéric Pont of the University of Exeter (UK). “But measuring its colour is a real first – we can actually imagine what this planet would look like if we were able to look at it directly.” To measure what this planet would look like to our eyes, the astronomers measured how much light was reflected off the surface of HD 189733b.
HD 189733b is faint and located close to its star. To isolate the planet’s reflected light from this starlight, the team used Hubble’s Space Telescope Imaging Spectrograph to peer at the system before, during and after the planet passed behind its host star on its orbit. As it slipped behind its star, the light reflected from the planet was temporarily blocked from view, and the amount of light observed from the system dropped. But this technique also shows how the light changes in other ways. “We saw the brightness of the whole system drop in the blue part of the spectrum when the planet passed behind its star,” explains Tom Evans of the University of Oxford (UK). “From this, we can gather that the planet is blue, because the signal remained constant at the other colours we measured.”
The planet’s azure blue colour does not come from the reflection of a tropical ocean, but is due to a hazy, turbulent atmosphere thought to be full of silicate particles, which scatter blue light. Earlier observations using different methods reported evidence for scattering of blue light on the planet, but these most recent Hubble observations give robust confirming evidence. HD 189733b presented a favourable case for these kinds of measurements, because it belongs to a class of planets known as “hot Jupiters”. These massive planets are similar in size to the gas giants in our solar system, but instead lie very close to their parent star – this size and proximity to their star make them perfect subjects for exoplanet hunting. We know that hot Jupiters are numerous throughout the Universe. Since we do not have one in our own solar system, studies of planets like HD 189733b are important to help us understand these dramatic objects. “It’s difficult to know exactly what causes the colour of a planet’s atmosphere, even for planets in the solar system,” says Pont. “But these new observations add another piece to the puzzle about the nature and atmosphere of HD 189733b. We are slowly painting a more complete picture of this exotic planet.”
On larger scales: from the Milky Way to the Universe as a whole
Colours of objects in the Universe have long fascinated both professional astronomers and the general public. In January 2012, for instance, a team of astronomers from the University of Pittsburgh (USA) announced the most accurate determination yet of the colour of the Milky Way, our home galaxy: “a very pure white, almost mirroring a fresh spring snowfall,” according to professor Jeffrey Newman and his PhD student Timothy Licquia. It has been difficult to make the measurement for the Milky Way, since our solar system is located well within the Galaxy. Because of this, clouds of gas and dust obscure all but the closest regions of the Galaxy from view, preventing researchers from getting the “big picture”.
“The problem is similar to determining the overall colour of the Earth, when you’re only able to tell what [the US state of] Pennsylvania looks like," Newman noted. To circumvent this problem, Newman and Licquia set out to determine the Milky Way’s colour by relying on images from other, more distant galaxies that can be viewed more clearly. These galaxies were observed by the Sloan Digital Sky Survey (SDSS), a project that measured the detailed properties of nearly a million galaxies and has obtained colour images of roughly a quarter of the sky. Without the large set of galaxies studied by SDSS to compare to, an accurate colour determination was not possible. The new colour measurement is allowing researchers to better understand the development of the Milky Way and how it is related to other astronomical objects.
“The problem we faced was similar to determining the outside climate when you are in a room with no windows,” said Newman. “You can’t see what’s happening, but you can look online and find current weather conditions someplace where they should be about the same.” The team identified galaxies similar to the Milky Way in properties that were able to be determined – specifically, their total amount of stars and the rate at which they are creating new stars, which are both related to the brightness and colour of a galaxy. The Milky Way should then fall somewhere within the range of colours of these matching objects.
“Thanks to the SDSS, the large, uniform sample needed to select Milky Way analogues already existed,” said Newman. “Although it is a relatively small telescope, only 2.5 metres in diameter, SDSS has been one of the most scientifically productive in history, enabling thousands of new projects like this one.” Newman described the overall spectrum of light from the Milky Way as being very close to the light seen when looking at spring snow in the early morning, shortly after dawn, the whitest (natural) thing on Earth. Many cultures around the world have given the Milky Way names associated with milk. Human vision is not sensitive to colours seen in faint light, so the diffuse glow of the Galaxy at night appears white. That association has proven to be very appropriate, given the Milky Way’s true colour.
Astronomers divide most galaxies into two broad categories based on their colours: relatively red galaxies that rarely form new stars, and blue galaxies where stars are still being born. These latest measurements place the Milky Way near the division between the two classes. This adds to the evidence that although the Milky Way is still producing stars, it is “on it’s way out,” according to Newman. “A few billion years from now, our Galaxy will be a much more boring place, full of middle-aged stars slowly using up their fuel and dying off, but without any new ones to take their place. It will be less interesting for astronomers in other galaxies to look at, too: the Milky Way’s spiral arms will fade into obscurity when there are no more blue stars left.”
The Milky Way’s colour is very close to the “cosmic colour” measured by Ivan Baldry from Liverpool John Moores University (UK), Karl Glazebrook of Swinburne University of Technology (Australia) and their collaborators in 2002; these researchers measured the average colour of galaxies in the local Universe. They added up all the visible radiation emitted by a very large number of galaxies in a huge cosmic volume, and determined how all of that light might be perceived by the human eye. Their team used a survey of more than 200,000 galaxies (the 2dF Galaxy Redshift Survey) reaching to distances of a few billion lightyears to construct the distribution of the colours the eye would see.
Since our Universe is expanding, light from distant galaxies is stretched to longer (redder) wavelengths. The farther away the galaxy, the greater the amount of stretching that occurs. The researchers removed this effect before combining all the light to form a smoothed-out average colour. Then they took into account the mean response of the human eye to the various colours, eventually coming up with the final result: the colour of the Universe is a shade of beige! Several people who had read about this result proposed names for the colour and the one that stuck – suggested by Peter Drum – was “Cosmic Latte,” similar in colour to milky tea or coffee.
The cosmic star-formation rate – the birth rate of new stars – reached its peak some ten billion years ago before starting to decline. In fact, by about five billion years ago, most of the stars that would ever form in our Universe had already formed. Since young stars are hot and blue while old stars are cooler and redder, the colour of the Universe was certainly different in the past, when there were many more young stars than today. Based on the observed behaviour of the star- formation rate over time, and if we continue to associate the colour of the Universe with the names of drinks, you might say that the Universe evolved from a “Cosmic Blue Hawaii” (a cocktail drink) to a “Cosmic Latte”. I bet you never thought that the Universe could be that exciting!
Figure 1: The pink arrow shows the position of Betelgeuse in the constellation of Orion. (Credit: Wikimedia Commons.)
Figure 2: This illustration shows HD 189733b, a huge gas giant that orbits very close to its host star HD 189733. The planet’s atmosphere is scorching with a temperature of over 1000 °C, and it rains glass, sideways, in howling 7000 km/h winds. At a distance of 63 lightyears, this turbulent alien world is one of the nearest exoplanets to Earth that can be seen crossing the face of its star. By observing this planet before, during and after it disappeared behind its host star during its orbit, astronomers were able to deduce that HD 189733b is a deep, azure blue – reminiscent of Earth’s colour as seen from space. (Credit: NASA, ESA, M. Kornmesser.)
Figure 3: This plot compares the colours of solar system planets to the colour of the hot Jupiter HD 189733b. With the exception of Mars, the colours are primarily determined by the chemistry of the planets’ atmospheres. Earth’s blue atmosphere plus the blue tint of the oceans dominate our world’s hue. HD 189733b’s deep blue colour is produced by silicate droplets, which scatter blue light in the scorching atmosphere. (Credit: NASA, ESA, and A. Feild, STScI/AURA)
Figure 4: This illustration shows a “hot Jupiter” planet known as HD 189733b orbiting its star, HD 189733. The Hubble Space Telescope measured the actual visible light colour of the planet, which is deep blue. This colour is not due to the presence of oceans, but is caused by the effects of a scorching atmosphere where silicate particles melt to make “raindrops” of glass that scatter blue light more than red light. Because the planet is only 63 lightyears from Earth, a visitor would see many of the same stars we see in our nighttime sky, though the constellation patterns would be different. Our Sun and the nearest star to our Sun, α Centauri, are labelled here – two faint stars near the centre of the image. Also labelled is Sirius, the brightest star in our skies in the constellation of Canis Major. (Credit: NASA, ESA, and G. Bacon, AURA/STScI)
Figure 5: A star field image showing the star HD 189733 (centre). The Hubble Space Telescope has given astronomers a fascinating new insight into the atmosphere of HD 189733b. To the right of the star is the notable planetary nebula Messier 27. The field-of-view is approximately 0.9 × 0.6 degrees. (Credit: NASA, ESA, and the Digitized Sky Survey 2. Acknowledgment: Davide De Martin, ESA/Hubble)
Figure 6: Full-colour view of the Milky Way, clearly showing regions of obscuration in red. (Credit: Axel Mellinger)
Figure 7: The galaxy has a range of colours, but on average is a very specific white. (Credit: BBC)
Figure 8: The cosmic spectrum, as measured by Karl Glazebrook and Ivan Baldry.
Figure 9: The average colour of the Universe, as calculated by Karl Glazebrook and Ivan Baldry.
Figure 10: Evolution of the colour of the Universe from 13 billion years ago to 7 billion years from now, as calculated by Karl Glazebrook and Ivan Baldry.