Chris Lintott’s Universe

I’m on top of a mountain in Spain; if you’d like to join me on my observing run then I’ll be posting updates on the Galaxy Zoo Blog.

Last week’s Nature had a series of stories about the feature by Eric Hand. My reaction on reading it was a sense that he’d got it absolutely right, hitting a tone that is best described as qualified success.

There is no doubt that Phoenix was a success. It was a cheap (especially when compared to the next generation of Mars landers, the American MSL and European Exomars) spacecraft, and yet in landed and produced results from each of its instruments. The headlines were taken by the chemistry of the soil – perfect for growing asparagus, if only it weren’t sprinkled with oxidising perchlorates – but there were other, ground-breaking results too. Flying an atomic force microscope to another planet and getting it to work is breathtaking, for example, and I’m sure we’ll be surprised again and again as the teams publish their results.

And yet, and yet…Phoenix was beset with problems. The team did manage to get icy soil into Phoenix’s ovens, but never a pure sample of ice. As a result, what was for me the most important single result expected from the mission – a measurement which would have told us whether the ice that Phoenix was sitting on had melted recently or whether it had been frozen in place for billions, not millions of years – will never come. According to the article, there are hints in the data that might be the signature of organic materials, but as the team ran out of time to run their blank, comparison sample we will probably never know for sure.

The fact that these problems were avoidable – the Nature article mentions that the problems with TEGA could have, should have been spotted before launch – is ultimately irrelevant. All problems with spacecraft should probably have been foreseen and corrected, but 100% success is achievable only in the dreams of bureaucrats. What matters is the overwhelming need for the mission to be seen, and talked about, as a resounding success and nothing but that. What Phoenix was trying to do was hard, it put in a huge amount of work in a very nasty environment and we know more about Mars than we did. We shouldn’t be afraid to admit that it didn’t quite deliver everything it could have.

Where does this fear come from? It’s pervasive, and it’s not necessarily NASA’s fault. I know that when reporting on Phoenix my instinct was to tell the positive story, and I felt guilty for mentioning the problems that the team – some of the nicest people I’ve met in the five years I’ve been doing Sky at Night – were having.

In an accompanying editorial, Nature have an answer. They point out that because even cheap missions are rare and (compared to, say, funding an Earth-based research project or experiment) expensive, the cost of failure is too large. A single mishap can derail a program for years. Their solution is for there to be more recycling of material between missions – shared landing systems, for example.

The amazing rovers, Spirit and Opportunity, they point out, bounced to the surface in an airbag, whereas Phoenix touched down, and MSL, the next in line, will land via something very scary called a sky crane. (If you haven’t seen it, it’s found it’s way onto youtube).

Why didn’t all the missions use the same technology? What Nature is missing is that they couldn’t have done; if I understand the situation correctly, the trajectory needed to land Phoenix in the martian arctic wouldn’t have slowed the lander enough in the atmosphere to allow airbags to be used, and MSL is simply too heavy for airbags to support it. The spirit of their call is right, though – more probes, costing less (and therefore each individual mission doing less) and a more honest approach to success or failure is exactly what’s needed. And not just on Mars, either.

Hat tip : Emily at the Planetary Society.

Whoever set the Guardian’s Everyman crossword this week was obviously thinking astronomically.

1 down is ‘Astronomer sees trail rising over one constellation (7)’ which I got straight away, but 2 down – ‘Rhea spinning, orbiting close to host planet (5)’ took me most of an evening for the penny to drop.

Answers to follow when their competition closes on Saturday; all hail the Guardian for making their excellent crosswords – certainly the most amusing and personable of the major cryptics – free online.

The Carnival of Space is open for your reading pleasure at the astroblogger’s blog.

Roll up, roll up etc.

If you only read one book about science in your life, then make it this one by Bad Science’s Ben Goldacre. For those who don’t know, Ben is a medical doctor who has been writing in the Guardian for years about the use and misuse of (mostly) health information in the media.

This is serious stuff, and his book covers much of the same territory as the blog and column; a relentless focus on the truth, with much more humour and lightness of touch than that description implies. It’s essential reading for anyone who ever glances at the stories in the paper that talk about which diets to choose, whether vitamin pills have any effect or which new treatments are available for cancer. Not because this book contains the answers, but rather because it will teach you how to find the answers by thinking statistically.

As a result, it should be essential reading for anyone who exists in the modern world, where the single most important thing you could do is develop a healthy respect for the sheer weirdness and counterintuitive nature of the probabilities that affect our judgement of risk. (Simple example: Goldacre points out we find it much easier to judge the significance of a result which says ‘Six instead of three people out of 1000 got cancer when they didn’t exercise than we do a headline that says ‘Chance of getting cancer doubled for the lazy’).

Mostly, then, the point of this post is to tell you to read the book. But as it climbed through the sins and stupidities of the alternative medicine industry, nutritionists, the pharmaceutical industry and the media (no-one gets off very lightly) I couldn’t help wondering where it was going. Goldacre has useful insights, among them the realisation that any science story that becomes big news is almost invariably removed from those most qualified to write about it, but while reading I couldn’t help feeling that having diagnosed our problem Goldacre was about to offer a grand solution.

He doesn’t. Worse that that, what he does say is defeatist and wrong-headed. He argues that science reporting should be reserved for the features section of the paper, where it could be given space to breathe. I get the sense he’d agree that this would reduce the audience for it to those who are already interested, but I’m not ready to give up on everyone else just yet. His clarion call, when it comes, is for scientists to blog more, and find their own audience that way.

It is true that blogs can provide in-depth, detailed coverage of a level that it’s impossible to imagine appearing in the paper. From my field, take Emily’s work at the Planetary Society as your canonical example. The problem is that by communicating only through the medium of the blog, rather than via the book or the column (both of which Goldacre has done brilliantly) or the news we speak almost exclusively to the converted. We do need to be louder and more strident about the need for scientific literacy (or rather, the statistical literacy that it requires), but to stop trying to communicate to a mass audience? That’s just giving up.

I’ve been negative often enough on this blog and in print about the scientific potential of the International Space Station, and had some interesting arguments as a result. 

If they’re going to produce spin off technology like this, though, combining space and wine, I may have to reconsider.

I’ve been trying to use art in my lectures and talks for a while now; I’ve written here before about how Anthony Gormley’s sculpture Blind Light helps me to think about the cosmic microwave background, and Escher’s Cubic Space Division is still the best way I know of explaining why we conclude that space itself is expanding.

A conversation today added to my list. I read on the Guardian’s website this morning that Keith Tyson was producing what he calls History Paintings; these prints, of red, green and black stripes and made available over the net, are apparently generated by an algorithm based on a roulette wheel.

I’ve been a fan for a while – his work is often scientifically inspired for one thing (and I tried quite hard to have him adopted as the artist for Bang! of all things.) And so, although only 5000 are (were by now, I should think) available, I managed to get in quickly enough to grab one. Here’s the result – History Painting (December, Oxford) :

I was quite disappointed there wasn’t any green. I wonder what the odds of that happening are? The algorithm isn’t clear from the information given in the paper, but we might guess that the odds of getting a green stripe in a particular position are the same as the odds of landing on green on a British roulette wheel; 1 in 38. That means the chance of getting no green at all would be just over 1 in 3. So my print, while pleasing, is nothing particularly special.

As the article in the paper notes, though, it’s possible that someone got an all-green image. According to my guess as to how likely green is, this would happen 1 time in 10^61 – that’s one in ten million billion billion billion billion billion billion. If I’d got an all green image, would I have celebrated my amazing luck?

I’m not sure. For starters, this run of roulette spins is exactly as likely as any other – it just looks special because it’s all green. I’d have to be thinking particularly mathematically to remember that, though, and maybe instead I would have suspected that the algorithm was not working as advertised (perhaps everyone’s is green, or the odds of getting green are less than I estimated).

Talking about this today in the department, I realised it’s a good analogue for looking out at the Universe; if we see an unusual or unlikely feature on large scales we have two choices. Take, for example, the presence of a large cold spot in the data from WMAP, looking at the early Universe as revealed by the cosmic microwave background.

If our models are correct, such a feature is fantastically unlikely, like an all-green History Painting. But there it is, so what do we make of it? We can either celebrate our ‘luck’ – of all the ways the roulette wheel of the Universe could have spun, it just happened to produce this interesting feature for us to look at. Or we can conclude that our understanding of the rules that govern the Universe is wrong, just as I might have concluded that an all-green image was a sign that the art-algorithm was broken.

Knowing when to do which is almost impossible, and one of the joys and frustrations of modern cosmology. Unlike artists like Keith Tyson, who can continually reinvent and respin the wheel, we only have one Universe to look at and must make do with whatever it shows us.

It’s (almost) Christmas, and buying presents is a nightmare.

Fortunately, here at Oxford Astrophysics we have a solution. Beautiful vintage prints of Palomar Sky Survey plates are available from Operation Skyphoto, set up to raise money to help treat Alexander Thatte, the son of two Oxford Physicists who has leukaemia. All money raised will go toward Alexander’s treatment or to research into leukaemia, so treat the astronomer in your life this Christmas.

For a little extra, you can even have your edition autographed by Jocelyn Bell-Burnell, or ‘gift wrapped by real astrophysicists’.