Quick: name the three largest known objects in the Kuiper belt. If you’ve been paying close attention you will instantly get Eris and Pluto, and, if pressed, you will admit that no one knows which one is bigger. And the third? An unscientific poll of people who should know the answer (my daughter, my wife, my nephew) reveals that not a single one does.
The answer, of course, is Makemake (you remember how to pronounce this, right? Mah-kay-mah-kay, Polynesian style). Makemake was discovered just months after the discoveries of Eris and of Haumea, and all were announced within days of each other. Eris and Haumea had important stories immediately attached to them (Eris was as big as Pluto! Haumea had suspicious discovery circumstances!), so poor Makemake stayed in the shadow of its more famous contemporaries. It was so overlooked that, in the hastily called press conference in which we announced the discoveries, I couldn’t even remember the official designation of Makemake when asked (it was 2005 FY9, of course; how could I have forgotten that?).
In the nearly 8 years since its discovery, we’ve learned more about Makemake, and it has become a critical Rosetta stone for understanding Kuiper belt atmospheres and outer solar system chemistry. It’s not overlooked anymore, but it still hasn’t had the massive scientific attention of Eris, with its large size and density, of Haumea, with its crazy spin and family of shards that it spread throughout the Kuiper belt, of Sedna, with its distant orbit unexplainable by anything currently known in the solar system. So I was very excited to recently find a scientific paper describing the first observations that would have a chance of discovering something truly special about Makemake: that –perhaps – it has the largest atmosphere in the Kuiper belt after Pluto’s. The observations were a stellar occultation (which I describe in detail earlier for Eris), which happens when Makemake passes in front of a star. When it’s in front of the star, Makemake essentially makes a shadow on the Earth of its size. The shadow races across the Earth as Makemake moves across the sky. If you can measure the size of the shadow, you can measure the size of Makemake. And, even better, if you can determine whether the shadow has hard or soft edges, you can detect whether or not there is an atmosphere. There is no way to observe the whole shadow at once, but, instead, telescopes watch the star blink out and then reappear as the shadow crosses across them. If you have a network of telescopes you have a good guess of the shape of the shadow. It’s an excellent technique, but hard, as Makemake doesn’t occult stars very often.
Alas, in this case, the observations were not quite sensitive enough to detect an atmosphere – that is, to tell whether or not the sides of the shadow were fuzzy – but, instead, they did allow a very precise measurement of the shadow’s size and shape. I started reading the paper already knowing that no atmosphere was detected, but curious about the measured size and shape. And then a funny thing happened. The paper said that they had measured the density of Makemake.
I was a bit surprised. Density, as you remember, is the mass of the object divided by its volume. I have been working hard for a lot of this decade trying to measure densities of dwarf planets because, in the end, densities tell you what an object is really made out of. In the convenient units that we use in the field, frozen water – which is abundant in the outer solar system – has a density of about 1 gram per cubic centimeter. Rock is close to 3 g/cc. The density of the Earth is about 5.5 g/cc, telling you that there’s something heavy down below – a core of iron in our case. The density of Saturn is only 0.7 g/cc showing that it really is gas almost all the way down. In the Kuiper belt it has long been known that Pluto has a density of about 2 g/cc – meaning that it is composed of something like half rock and half ice. Sure, there are other things around, but, really, if you want to understand the basics about what an object is and where it came from, knowing how much rock and how much ice is truly the main thing that you want to know.
When we first started discovering objects in the Kuiper belt, we expected them all to be made of the same stuff and thus have the same densities as Pluto. One of the quick surprises, though, was how much variation there is. Haumea and Eris are both closer to 2.6 g/cc (much more rock!), small objects are closer to 1 g/cc (almost all ice). Objects in between are… well… that’s one of the critically important questions. Do densities smoothly increase as objects get bigger? Are densities modified by giant impacts of the types that form satellites? I would love to know the answer. One way to help answer the question would be to measure the density of an object that didn’t have a satellite. If giant satellite-forming impacts blast away ice and leave rock and thus make high densities, then objects without satellite will have systematically lower densities. It’s a nice little easily testable hypothesis. Except for one problem: we can’t easily measure the densities of objects without satellites.
Density, again, is mass divided by volume. Volume we get by measuring the size – as in these occultation measurements. Mass, to date, we have only gotten in one way: find a satellite and see how long it takes to go around the Kuiper belt object. Simple high school physics then tells you the mass of the central object. A massive central object will cause satellites to whip around quickly, or keep even distant satellites in orbit. A less massive central object will cause slow lazy satellite orbits. The measurements, in this case, can be very precise. This is how we know, for example, that Eris has a mass that is 28% greater than that of Pluto.
Makemake has no known satellites, and I’ve looked hard with both the Keck telescope and the Hubble Space Telescope. If there is anything there it must be really really small. And without a satellite, there is no way to directly measure the mass of Makemake. With no mass there is no density. Yet, right here, in the abstract of the paper, the value of the density is measured and it is 1.7 g/cc with an uncertainty in either direction of 0.3 g/cc. So there is the answer. Makemake has lower density than the objects with satellites. So it must be that giant impacts do blast ice away and leave dense rocky cores, and Makemake hasn’t had one of these giant impacts or it would have left satellites – which it doesn’t – and also a high density, which we now know it doesn’t.
Except. Wait. How exactly did they measure the density when there is no satellite? In fact, how did they measure the density from these stellar occultations at all? In fact, what?
It’s probably not worth going into detail the ways in which the alleged density measurement in the paper is both flawed and wrong. Many papers get published that are flawed and wrong. If you have never written a scientific paper that is flawed and wrong you have probably not written many, if any, scientific papers. It’s OK. That’s the way science works. Someone writes something that they think is right, someone else disagrees, the conversation continues. The next time someone writes a paper about the subject they might mention an alternative interpretation. We all move on.
The density measurement wasn't even a major focus of the paper. The atmosphere was. And, to me, that analysis looks solid. But I couldn't overlook two things. First, the paper claimed a density measurement which was, quite simply, wrong. Not just the wrong value, but a value that we currently have absolutely no way of knowing. Second, the density of Makemake – the thing that they claimed to have measured but could not possibly have – matters immensely. I don’t think they knew how important this measurement was or perhaps they would have been more careful, but, to me, the density of Makemake is one of the key measurements that explains how the whole outer solar system put itself together.
I concluded that I really needed to write a paper correcting the flawed paper. It's an unpleasant task, for although everyone has written papers before that are flawed and wrong, no one likes someone else pointing that out. Scientists are human. Scientists get feelings hurt, get embarrassed, get mad, hold grudges, do all of the things that humans do. The good news, if there was any, was that in reanalyzing the data myself, I came up with some pretty cool statistical methods that no one had yet applied to occultations. So the paper wouldn't just be a "these guys are wrong" paper, it would also be a reanalysis describing a better way of analyzing any such data in the future.
And then for the really bad news. The first author of the paper was Jose Luis Ortiz. In case you have forgotten your recent outer solar system history, Ortiz was the alleged perpetrator in the Great Haumea Caper (I just made that name up, but I kind of like it. I should mention that, still, nearly 8 years later, I can't really tell you what actually happened for sure. Wish I could.) And it is really really easy to believe that I would take great delight in publicly slamming a scientific paper that Ortiz wrote. And that I would take tiny flaws and try to make them sound worse than they are. But none of this is true. I would have been much happier had the original paper simply been correct. I would have been happier if he had realized himself that his data contained no density measurement. I would have been happier working on my real projects rather than having to correct this one. But none of that happened. And I care what the density of Makemake is. And I don't want anyone to look up the Ortiz paper, read about the density measurement, and go off making up hypotheses for the outer solar system based on a measurement that is not just wrong but, with the current data, impossible.
So I wrote the paper. It came out in the Astrophysical Journal Letters the other day. If you don't have a subscription, you can read the archived version for free on arXiv, the place that most astronomers and physicists post papers so that all can read. I wish I could also give you links to the Ortiz paper, but, alas, there is nothing publicly posted (I was wrong; see below!). So you'll just have to take my word for it. I find that unfortunate, because I think it would make interesting reading to compare them side-by-side and find where Ortiz went wrong. But, for most of you: tough luck.
What next? Well, while thinking hard about all of this, it occurred to me that there is a way to measure the density of Makemake. It will not be easy. It'll take observations from the ALMA millimeter array in South America, more stellar occultations, and a precise determination of Makemake's rotation speed. But I think, in principle, all of those could happen. Stay tuned. Eventually this flawed claim of a density measurement may have been the thing that spawned the seeds that allow the density to finally be known. Stranger things have definitely happened in the outer solar system.
Update:
@Vagueofgodalming in the comments below points out that you can indeed read the Ortiz paper in a version archived by ESO. Which is great. Sadly, however, you still don't get the whole story because almost all of the real details of the paper are hidden in the "supplementary information" that isn't included here (UPDATE #2: You CAN get the supplement directly from Nature. Which is just plain weird. But here it is). The journal Nature, in which this appears, has stringent length limits in the actual journal, but lets you add as much as you want in the supplementary information. I can imagine that at one point the information in these supplements was really supplemental. You could learn more by reading it, but you didn't have to read it to have the point of the paper proved. This is no longer the case. The supplements are critical to the arguments in the paper, so, as you either have to keep flipping back and forth to figure out what is going on, or, worse, you never bother to read the supplement so all you are really reading is an extended abstract that never really proves the points the paper is trying to prove. Or the supplement is just wrong but the fact is missed in peer review because the reviewer didn't bother to take the supplement that seriously. As you might have guessed by now, I find papers with "supplementary information" that is actually just "required information that doesn't fit in our print journal" to be irritating. I don't blame the paper at this point, though, I blame the journal for propagating this poor style upon us.
Update:
@Vagueofgodalming in the comments below points out that you can indeed read the Ortiz paper in a version archived by ESO. Which is great. Sadly, however, you still don't get the whole story because almost all of the real details of the paper are hidden in the "supplementary information" that isn't included here (UPDATE #2: You CAN get the supplement directly from Nature. Which is just plain weird. But here it is). The journal Nature, in which this appears, has stringent length limits in the actual journal, but lets you add as much as you want in the supplementary information. I can imagine that at one point the information in these supplements was really supplemental. You could learn more by reading it, but you didn't have to read it to have the point of the paper proved. This is no longer the case. The supplements are critical to the arguments in the paper, so, as you either have to keep flipping back and forth to figure out what is going on, or, worse, you never bother to read the supplement so all you are really reading is an extended abstract that never really proves the points the paper is trying to prove. Or the supplement is just wrong but the fact is missed in peer review because the reviewer didn't bother to take the supplement that seriously. As you might have guessed by now, I find papers with "supplementary information" that is actually just "required information that doesn't fit in our print journal" to be irritating. I don't blame the paper at this point, though, I blame the journal for propagating this poor style upon us.
Obviously scientists are human and feel and do all that you say (and much worse things sometimes) but something that most don't do is to reach out to the rest of the World and share with us what's going on in the academic sphere. You are one of the rather few exceptions and for that I am very thankful.
ReplyDeleteMike, I respect and understand your feelings about Ortiz, complete with your need to correct the data while wishing you didn't have to. That's a tribute to your humanity, and it's the reason why people like you - beyond respecting your work and generosity of spirit.
ReplyDeleteAs for 'the Great Haumea Caper,' that's just a great name. A shenanigan with a side order of pineapple? Who wouldn't love that?
I love reading your posts, and will look forward to the resolution to this MakeMake question. Thanks for sharing!
The Ortiz et al paper is at (pdf) http://www.eso.org/public/archives/releases/sciencepapers/eso1246/eso1246a.pdf
ReplyDeletehttp://www.eso.org/public/news/eso1246/ is the news release it's linked from.
Dr Brown,
ReplyDeleteI am a fan of both your blog and your book, and both my 10-yr-old daughter and I have a keen interest in astronomy. I'm not sure this is really an appropriate place to ask this question, but it's been something I've wondered about for a long time, so what the heck...
Why is it that astronomers can know things about objects very far away - distant galaxies from almost the dawn of the universe, planets around other suns, etc - but you have such a hard time finding stuff in our own solar system? It seems to me that closer things would be much easier to see than distant ones. I'm amazed we're stilling finding stuff in our solar system.
I appreciate your attention and time, if you get around to answering this.
Thank you.
Imagine you had a small fly held out 100 feet away. You might not even know it were there. Now imagine a big billboard advertisement 1,000 feet away. You can definitely know a lot more about the billboard than you can the fly.
DeleteGalaxies have angular diameters and brightnesses that are much, much greater than Kuiper Belt objects.
I have been subscribed to "Mike Brown's Planets" RSS feed for a few months, as well as various NASA feeds and my local "Oxford Mail"(UK) news, this latter coming from just 3-5 km distant. To my surprise Mike's feed arrives before my local news, despite being from Pasadena(?). Any explanation as to the location of the RSS feed, please? OR has Pasadena got a really REALLY fast feed?
ReplyDeleteThis comment has been removed by the author.
DeleteIf you're talking about how long a feed takes to load when you try to view it: it's probably different server loads, speeds, and locations. Mike's blog is hosted by Blogspot, and his feed is on FeedBurner (and both services are run by Google), so maybe they have more, closer, and/or faster servers than whoever hosts NASA's and the Oxford Mail's feeds.
Delete