Thursday, May 19, 2016

Coloring the Universe: How Beautiful Astronomical Images are Made

Astronomers have made discoveries that have completely changed our view of the Universe and our place in it. Their advanced telescopes have given us a kind of superhuman vision that greatly surpasses what our eyes are capable of in terms of sensitivity, resolving power and wavelength coverage. Their spectacular images rival the beauty of our finest works of art.

The technology and expertise required to obtain these images is just as impressive as their beauty. The book “Coloring the Universe: An Insider's Look at Making Spectacular Images of Space”, released in November last year, gives perhaps the best description available of how these beautiful images are obtained, ranging from a description of the instruments used, to the software techniques adopted to produce the best presentations. Travis Rector, Kim Arcand and Megan Watzke wrote the book.  Travis Rector, an astronomer and one of the world’s best at producing astronomical images, wrote a seminal paper giving a “practical guide” on “how to generate astronomical images from research data with powerful image-processing programs”. Kim Arcand and Megan Watzke are both award-winning science communicators and authors, with extensive experience in disseminating images to a global audience. (In full disclosure, Arcand and Watzke are Chandra X-ray Center colleagues and friends of mine, and I also reviewed a previous book by them, called "Your Ticket to the Universe". In between they’ve written a book called “Light” that is full of gorgeous images from many fields of science.)
Figure 1: The cover of Coloring the Universe, showing an optical image from the NSF’s Mayall 4-meter telescope at Kitt Peak National Observatory of IC 1396A, a dark nebula more commonly known as the Elephant Trunk Nebula. Credit: T.A. Rector (University of Alaska Anchorage) and H. Schweiker (WIYN and NOAO/AURA/NSF).

The text in Coloring the Universe is eloquent and accessible to a wide audience. The book has excellent organization, and each chapter is broken up into easily digestible subsections. Some of the topics covered include a comparison of human vision with telescopic vision and a discussion of what astrophysics can be learned from images. It also explains some details about observing at the world’s largest telescopes and discusses the different kinds of light that we observe.

As expected, the book is full of spectacular images, many produced by Travis Rector and his colleagues, with careful descriptions given in figure captions. It’s striking that many of the ground-based telescope images by Rector et al. are just as beautiful as those made by NASA’s Hubble Space Telescope (HST). For example, here is a HST image showing part of the Veil Nebula, the remains of a supernova in our galaxy:

Figure 2: An HST image showing part of the Cygnus Loop supernova remnant, the expanding remains of a massive star that exploded about 8,000 years ago. Credit: NASA, ESA, and the Hubble Heritage Team (STScI/AURA)  

For comparison, here is a ground-based image from a different part of the remnant:

Figure 3: An optical image from the Mayall 4-meter telescope, of the region known as Pickering's Triangle, also part of the Cygnus Loop supernova remnant. This image is rotated by 180 degrees from the one used in the book. Credit: T.A. Rector (University of Alaska Anchorage) and H. Schweiker (WIYN and NOAO/AURA/NSF).
In this case and many others, the much bigger field of view of ground-based telescopes compared to HST can compensate for their much lower spatial resolution, as I described in a blog post in 2014: “What Makes an Astronomical Image Beautiful?, based on a paper by Lars Lindberg Christensen and colleagues. One reason that HST images are often more familiar is that they have a more powerful publicity engine promoting them. Coloring the Universe allows part of this publicity imbalance to be rectified. 

Figure 4: An optical image using the NSF's 0.9-meter telescope at Kitt Peak National Observatory of the Rosette Nebula. Credit: T.A. Rector, B.A. Wolpa & M. Hanna (NOAO/AURA/NSF).

Criticism of astronomical images

Two motivations for writing Coloring the Universe are to demystify the production of astronomical images and indirectly respond to critics. For publicity, beautiful images give astronomy a clear advantage over many other fields of science, as we have found in publicizing Chandra results. There doesn't have to be exciting or important science for a particular result to receive widespread attention. However, not everyone can appreciate such beauty without deep skepticism, and over the years I’ve collected some critiques of astronomical images. (I forgive but I don’t want to forget.) For example, in a New York Times review of an exhibition of solar system images, the writer described the “sub-wooferish whooshes of sound” accompanying planetarium shows using HST images. He followed by writing “Well, the colors are as phony as the sound.” In another example, science writer Charlie Petit described HST images as being “simultaneously dreadfully misleading, worthwhile, and useful”, in one post at the Knight Science Journalism Tracker (now archived at the online magazine Undark). Referring to the famous HST image of the “Pillars of Creation” in a 2007 post, Petit said “The Tracker finds it worth posting in part just to put up the Hubble telescope’s unbelievable image from ten years ago (unbelievable is literally true. Its power came from extensive color tweaking that gives it far more drama than would greet the naked eye).”

It’s not just writers who have been critical. Washington Post writer Joel Achenbach wrote about the mostly negative reactions of astronomers to the Pillars of Creation image (scroll down to the text “From a 1997 story I did in the magazine”). The astronomers criticized the colors used and they also criticized the orientation of the image, as though it’s important for a public audience to maintain the arbitrary astronomer’s convention of North pointing up. These comments were collected almost 20 years and hopefully since then astronomers have gained a better appreciation for the optimal presentation of images for a public audience. I’m not sure this is the case, and for what it’s worth, a 2015 study by Kim Arcand and colleagues of what people think is “real” in astronomical images showed no significant difference between the opinions of self-rated experts and non-experts.

Critics like those described above would gain a better understanding of how images are made and the motivations behind these methods by reading Coloring the Universe. For example, the book contains a chapter called “Photoshopping the Universe: what do astronomers do? What do astronomers not do?” followed by a chapter called “The aesthetics of astrophysics: principles of composition applied to the Universe”. The latter includes a fascinating subsection on why the Pillars of Creation image looks so dramatic.

Responding to the criticism

As Kim Arcand and colleagues explain in discussing the aesthetics of images, the use of color leads to the most questions and comments. As noted earlier, this can result in claims that astronomical images are faked & that they're nothing like what the eye would see. My response to the former is “no!” and my response to the latter is: “why should they be, when telescopes can see so much better than our eyes?” It’s a common fallacy to assume that astronomical images are meant to show what our eyes can see, or might be able to if they were more sensitive. Astronomical images can convey an enormous amount of information, especially when they are not limited by the shortcomings of human vision. For example, images can show narrowband, optical images to pick out phenomena that our eyes are unable to discern. Even more significantly, they can show objects and phenomena in wavelengths that are well beyond the range of human vision, such as X-rays and radio waves.

Figure 5: A composite image of NGC 602, a cluster of bright young stars in the Small Magellanic Cloud, a nearby galaxy. Chandra data is shown in purple, optical data from HST is shown in red, green and blue and infrared data from the Spitzer Space Telescope is shown in red. Credit: X-ray: NASA/CXC/Univ.Potsdam/L.Oskinova et al; Optical: NASA/STScI; Infrared: NASA/JPL-Caltech

As various image experts have explained over the years, such as Megan Watzke, Kim Arcand and Robert Hurt, the colors used in astronomical images are often representative, which means that they are not intended to show how our eyes might see the object, but instead represent maps of the electromagnetic radiation produced at different wavelengths, and with a range of filters. Astronomers often use the term “false color”, but that terminology is misleading for non-expert audiences, as Robert Hurt and others have pointed out. In many cases images simulate what our eyes might see if they were sensitive to very different wavelengths, like Geordi La Forge's visor-enhanced vision in Star Trek the Next Generation.

A good example in Coloring the Universe is the Chandra image of Tycho’s supernova remnant, the remains of a supernova seen on Earth in 1572. Here, the shortest wavelengths are shown in blue, intermediate wavelengths are shown in green and the longest wavelengths are shown in red, in the same order and wavelength order as our vision at optical wavelengths. This use of color helps explain the astrophysics, as the outer blast wave has produced a rapidly moving shell of extremely high-energy electrons (blue), and the supernova debris has been heated to millions of degrees (red and green).

Figure 6: A Chandra image of Tycho’s supernova remnant. Credit: NASA/CXC/SAO

Coloring the Universe gives a much more detailed discussion of the meaning and value of multiwavelength images, and the use of color in making attractive images that give insight into their scientific content.

Without giving too much away, here are a few other highlights of the book:

  • It shows a sense of humor, e.g. in describing how colors should be used appropriately it notes that people in images wouldn’t be colored green “unless you’re in Roswell, New Mexico”. Also, since optical astronomers observe (hopefully) all night, the authors note, “like vampires we sleep during the day” and “Fortunately, we don’t sleep in coffins”. 
  • I liked the use of raw images to show what telescopes collect before processing has been done, and how calibration and multiple exposures correct for changes in charge coupled detector (CCD) sensitivity and gaps between CCDs.
  • It gives an excellent description of the history of astronomical images and their dissemination, including HST’s observations of Jupiter and Shoemaker-Levy in 1994 at a pivotal time for the production and dissemination of images. The World Wide Web started in 1991 but its use was limited until the Mosaic web browser was introduced in 1993. Another big step was the release of layering capabilities in Photoshop in 1994, allowing sophisticated color composite images to be created.
  •  An authoritative description is provided of the important role that images play in publicity, including the establishment of programs like Hubble Heritage, providing the opportunity to gather beautiful images that professional astronomers might have missed.

I’ll end with an excerpt from the summary of
Coloring the Universe:

“In this book we’ve talked about the complex process of converting what the telescope can see into something we humans can see. It’s a fundamental challenge because our telescopes observe objects that, with a few exceptions, are invisible to our eyes. That is of course the reason why we build telescopes. There would be no point in building machines like Gemini, HST and Chandra if they didn’t expand our vision. Astronomers use telescopes to study and understand the fundamental questions of how we came to be: from the formation and fate of the Universe; to the generation and function of galaxies; to the birth, life and death of stars inside galaxies; to the planets and moons around these stars; and to the origin of life here and possibly on worlds beyond our own.”

The text continues with an eloquent summary of the principles and motivation behind the production of astronomical images, which you’ll have to purchase the book to enjoy. I highly recommend that you do so, to savor the gorgeous images Coloring the Universe contains and to appreciate the ingenuity involved in producing these images.

End note: While waiting for Coloring the Universe to arrive in the post, you can watch this excellent talk by Jayanne English from the University of Manitoba, titled “Are images of space realistic?”

Friday, March 25, 2016

The Superb LIGO Press Conference Announcing the Direct Detection of Gravitational Waves

The recent announcement that gravitational waves had been directly detected was one of the most exciting and important events in the history of astrophysics. I watched the press conference with dozens of other people at the Harvard-Smithsonian Center for Astrophysics (CfA), and we enjoyed sharing the excitement of the Laser Interferometer Gravitational-wave Observatory (LIGO) team. I was tweeting so quickly that my hands started to get sore.

For a long time I’d been skeptical that they would ever make this detection, as the technical challenges are incredibly formidable. It wasn’t until team member Daniel Holz visited CfA for a talk and discussed LIGO’s progress that I started to think they could actually do it. I chatted with Holz after his talk about how the announcement of a detection might be made, and since then I’d wondered: would the quality of the result’s communication be as strong as the result itself? On February 11th, 2016 I found out that the answer was a resounding yes.

I think the press conference to announce the LIGO detection was superb, perhaps the best that I’ve ever seen. I can give my opinion with some authority as I've helped organize many press conferences for the Chandra X-ray Observatory. David Reitze, Gabriela (Gaby) Gonzalez, Rainer Weiss and Kip Thorne all did excellent jobs. In full disclosure, I have previously worked with the press officer Whitney Clavin who organized the press conference, but I had already made most of the notes for this blog post before I found this out.  Congratulations to Whitney and her communications team, including graphics experts Robert Hurt and Tim Pyle, for doing such a good job, as I describe in detail below. Of course, this communication effort would not have occurred without the astonishing and extremely important achievement by the entire LIGO team. Experimental mastery, theoretical expertise and communications prowess have combined here in a powerful way.

Below I describe some of the many highlights of the press conference and then discuss the analogy of sound being used to explain gravitational waves. I finish with criticism of the Nature Physics editorial that made derogatory comments about the LIGO press conference.

Figure 1: The LIGO Hanford Observatory. Credit: Caltech/MIT/LIGO Observatory.

The LIGO press conference highlights
  • The press conference kept viewers in suspense for a few minutes at the start of the press conference, with France Cordova’s introduction and a short video including hints that they’d been successful. Then the official announcement was made. With our Chandra press conferences we usually jump to the announcement quickly, to avoid losing viewers, however this result was so important that it transcended typical practice.
  • The actual announcement of the detection by David Reitze was clear, simple and deliberate: “Ladies and Gentleman, we have detected gravitational waves. We did it.”

Figure 2: David Reitze, Executive Director of the LIGO project, making the announcement that they had detected gravitational waves - “We did it” - in a screen capture from the press conference. Credit: Caltech/MIT/LIGO Observatory.
  • There was a clear feeling of celebration amongst the panelists, including Reitze briefly throwing his arms upwards in victory just after making the announcement of a detection, and a lot of hugging between panelists. All of this was perfectly appropriate and humanizing.
  • Reitze quickly emphasized that a lot of checking was done before they announced the result, and he also briefly described the great future ahead for gravitational wave astronomy 
  • Simple language was used, especially early on, as emphasized in a Forbes article: “How The Epic Discovery Of Gravitational Waves Was Brilliantly Communicated”, by Carmine Gallo.
  • The structure of the whole press conference was excellent, with the punch-line early, followed by a good executive summary, followed by filling in details about the instrument and the history later on. This structure resembled a good press release or news article. 
  • They did the publicity after a paper describing the main result was accepted for publication, unlike some recent press conferences, such as the one for BICEP2. 
  • They emphasized the superlatives and there were a lot of them, like “first” and “proof” and “mind-boggling” and “scientific moon-shot”. 
  • They used excellent graphics and props, including a good mixture of helpful, intuitive animations and straightforward graphics, to explain what gravitational waves are, how they affect objects and what was oberved. I like how Reitze showed a view of the black hole merger as a nearby viewer would see them, followed by a version showing the gravitational waves. Later on the animation shown by Rainer Weiss to explain how LIGO detects gravitational waves was particularly good.
  • Gaby Gonzalez did an excellent job explaining what I think was the key scientific figure in the press conference (see Figure 3 for more details).

Figure 3: The figure shown by Gaby Gonzalez demonstrating that they had really detected gravitational waves, because a very similar signal was seen at both observatories, separated in time by ~7 ms, roughly equal to the light travel time between the two observatories. As Gonzalez said about this figure, another screen capture from the press conference, “This is it. That’s how we know we have gravitational waves”. At CfA this figure drew gasps and cries of “whoa” from the audience. Credit: Caltech/MIT/LIGO Observatory.
  • They had a server that didn’t, as fair as I know, crash during the webcast. It was smart that they were prepared for a big following, despite the leaked rumors on Twitter by Lawrence Krauss (here and here) and in Science Magazine. They encouraged group watches of the press conference, presumably to help limit the risk of server crashes. It was also more fun to watch that way too. At CfA it felt a bit like watching an exciting sports event, with a bunch of parochial supporters. The atmosphere was electric and appropriately so.
  • They discussed some “gee whiz” aspects, including the huge amount of power produced when three solar masses is suddenly converted into gravitational waves: during the collision the total power output in the gravitational waves was 50 times greater, for that brief period, than all of the power produced by all of the stars in the universe put together. I also liked the analogy Reitze gave for explaining the sensitivity of their measurement:
“..if we were trying to measure the distance between the Sun and the nearest star, which is about 3 and a quarter light years away, LIGO is capable of measuring that, if it could do that, to the level of about the width of a human hair." 
That’s extraordinary, though it would have been even better for a general audience to give the distance in miles or kilometers – amounting to 30 trillion kilometers, or 30,000 billion kilometers – rather than light years.
  •  They didn’t use cliches like “Holy Grail” or “smoking gun”, even though it must have been tempting to do so, as such phrases were much more appropriate here than for any other science press conference I can recall.
  • There was a sense of humor at times, like Weiss’s remarks that Einstein could have built LIGO back in 1916 if the technology had been available: “He was smart enough and he knew enough physics. He wasn’t just a theorist”. This got a good laugh from the other panelists, then the audience. Weiss then immediately introduced Kip Thorne “who is a theorist, but really also an experimenter”. Again, laughter.
  • Gaby Gonzalez carefully and graciously acknowledged the large number of people who worked on LIGO.
  • There was a very clear explanation who the founders of LIGO are, Ronald Drever, Kip Thorne and Rainer Weiss, who will be hot favorites to receive the Nobel Prize in Physics where the limit, probably not coincidentally, is three Laureates in a given year. If I was on the Nobel Committee I would be pushing to award the prize to them later this year, or as soon as possible.
  •  The speakers provided clear comments on how their results would improve in the future, including the much better positions for events that will come from adding extra observatories similar to the two that already exist, but at different locations. This is crucial for follow-up searches for counterparts using observatories like NASA’s Chandra X-ray Observatory, ESA’s XMM-Newton Observatory or the Australia Telescope Compact Array, as suggested in searching for the possible counterpart to the newly discovered merger of two black holes.

The gravitational waves as sound analogy

Many of the words and graphics used by the panelists in this press conference can be adopted by speakers in scientific and especially in public talks about gravitational waves. One exception to this recommendation is that I think the panelists could have been clearer in explaining that it is an analogy to describe gravitational waves as sound. This analogy is vivid and has a couple of key strengths, as explained to me by LIGO team member and former CfA colleague Vicky Kalogera:
1. "The frequencies LIGO is sensitive to are in the same range human ears are sensitive to, so there is an “easy” conversion of LIGO signals into sounds one can make." 
2. "More importantly, “regular” (ie most electromagnetic) telescopes are pointing instruments, like our eye. In contrast the LIGO detectors work like human ears: LIGO can detect signals from above, below, back, and front." 
Despite these key strengths this is still only an analogy: gravitational waves are very different from sound waves, e.g. the former are transverse waves, or ripples in space-time and the latter are longitudinal waves traveling through a medium, such as air or water; the former can travel through the near-perfect vacuum of space and through dense objects and the latter cannot. Also, in a terrestrial setting the speed of propagation for these waves in very different, by a factor of almost a million.

To explain how the sound analogy was introduced in the press conference, here’s a quote from Reitze’s presentation:

“What LIGO does is that it actually takes these vibrations in space-time, these ripples in space-time and it records them on a photo-detector and you can actually hear them. So, what LIGO has done, it’s the first time that the universe has spoken to us through gravitational waves. And this is remarkable, up to now we’ve been deaf to gravitational waves, but today we are able to hear them. That’s just amazing to me. I think this is big, again because what’s coming now is we’re going to be able to hear more of these things. And no doubt we’ll hear things that we expected to hear, like binary black holes or binary neutron stars colliding, but we will also hear things we never expected.”

He correctly describes the vibrations as being in space-time, but then proceeds to describe gravitational waves as sound without being clear that this is an analogy. The people who may not know this is an analogy probably overlap a lot with those who don’t know what space-time or a photo-detector is, so his introduction might not have helped much. Later on Gaby Gonzalez played the “chirp” from the merger, without a clear explanation that the gravitational waves were converted into sound, which didn’t help.

Was this a problem for a general audience, who would generally hear about the result through the press, where further simplifications often occur? I didn’t make a systematic analysis of the press articles, but for an example, this is how Dennis Overbye started his page one New York Times article about the discovery:

“A team of scientists announced on Thursday that they had heard and recorded the sound of two black holes colliding a billion light-years away, a fleeting chirp that fulfilled the last prediction of Einstein’s general theory of relativity.”

This is misleading for non-experts. Also, the title for the video by Overbye is “LIGO Hears Gravitational Waves Einstein Predicted”. 

Figure 4: Page one of the New York Times on Friday, February 12th, 2016. Credit: New York Times and the Newseum.

For another example, the second paragraph of the page one Boston Globe story said:

“At 5:51 a.m. on Sept 14, a new, highly advanced observatory in Louisiana heard the explosion as it passed the Earth in ripples known as gravitational waves, a high-pitched chirp in the steady hum of the universe. Seven milliseconds later, the distinct sound was registered at a sister observatory in Washington.”

Only minor changes are needed to explain that gravitational waves are not the same as sound waves. For example watch Brian Greene’s appearance on “Late Night with Stephen Colbert”. First he does a great job, comparable in standard to the LIGO panelists, at explaining how gravitational waves distort objects as they pass through them. Then he introduces the sound analogy in the following way:

Brian Greene: “In fact you can actually in some sense hear the gravitational waves. They vibrate at a frequency that if you turn it into sound the human ear can hear.

Stephen Colbert: “So literally these waves can be turned into sound.”

Brian Greene: “They can”.

Of course, there’s no guarantee that writers will explain the analogy correctly because of this subtle change, but it gives them a better chance. Also, it’s not vital that non-experts understand the difference between gravitational waves and sound waves, however it certainly doesn’t hurt.

Another, related issue − that I never would have thought of myself − is that talking about “hearing” the universe and formerly being deaf to it, is ableist language. See this interesting discussion on Twitter led by Chanda Prescod-Weinstein.

In defense of the press conference panelists, I’ll note that the analogy has been used by scientists before. For example Janna Levin used it extensively in her TED talk “The sound the universe makes”. This talk included some brilliant explanations, such as a description of a journey into a black hole, but I didn't feel completely comfortable with heavy use of the sound analogy, placing it on my personal radar.

In science communication I think it’s useful to discuss the care that should be taken with using analogies, especially a cool example like this one, and how these can end up being distorted. A second, loosely-used analogy is saying that astronomers listen to cosmic phenomena using radio telescopes. Besides being potentially misleading, another problem for these two analogies with sound is that there are real examples of sound in space. To name three, sound exists in oscillating stars, in galaxy clusters and it existed in the early universe. I do have a bias here as I used to do research in oscillating stars and I helped organize a press conference publicizing the incredibly deep note in the Perseus galaxy cluster, however my point remains. Again, only small changes are needed.

A Nature Physics editorial criticizing LIGO’s communication

As noted above I was very impressed with the LIGO press conference. I was much less impressed with an editorial – here is a publicly available copy – that recently appeared in Nature Physics about LIGO’s communication efforts. (This contrasts strongly with the excellent coverage of the gravitational waves result itself by the Nature News team, including Davide Castelvecchi.) Here’s one excerpt from the Nature Physics editorial:

“Of course, the announcement, made on 11 February 2016, immediately hit the headlines. “We did it”, stated David H. Reitze, executive director of LIGO. The excitement was palpable. Some of us cried. But the public’s response was largely summed up by the satirical news source, The Daily Mash, with their headline: “Scientists completely fail to explain ‘gravitational waves’”.”

I disagree strongly with the final quote in this statement and it’s telling to note that a satirical news source was used. The editorial goes on to say “And it isn’t difficult to explain” by giving an explanation that is much less clear, for non-experts, than the one that the LIGO panelists gave. For example, nowhere does the Nature Physics article mention how the gravitational waves stretch and shrink objects as these ripples in space-time travel outwards from their source. The editorial also discusses fields and charges. Again, not very clear for a general audience.

The editorial then says:

“But what is truly mind-blowing is that not one of the telescopes operating at electromagnetic wavelengths has detected a counterpart event.”

This statement isn’t completely correct because a possible counterpart has been found with the Fermi Gamma-ray Burst Monitor, one that is plausible enough to motivate quick papers by smart theorists like Avi Loeb and Stan Woosley to explain this unexpected phenomenon, if true (the papers are here and here, respectively).

The editorial finishes by mentioning the prospects for detecting gravitational waves from just after the Big Bang, and concludes with:

“In the meantime, we should learn to explain the physics of these spectacular events to non-physicists.”

No, because we already have.