Cosmology Conundrum

March 26, 2013

Some colleagues have suggested that my posting on Planck Day was overly frivolous, given the huge importance of what Planck has achieved. (Nicely written up by Andrew Jaffe.) Other colleagues have suggested that Planck Day was bad press, as it was such a huge public non-event, with a big fuss about mild parameter adjustment. I find both these things true, leaving a weird sense of tension and excitement. I’d better explain myself.

First, lets be clear about the technical achievement. Planck is an absolute triumph of technology, engineering, management, and organisation. An amazing machine that has worked beautifully. It also represents a stunning scientific achievement. The lamda-CDM model, and its beautifully articulated engine of prediction, is a conceptual and analytic triumph. Of course this triumph belongs not just to the Planck team, but represents the accumulated achievement of many scientists all over the world over a number of years.

Lets just look at that power spectrum fit. It is not just a question of the theoretical curve going vaguely up and down, in more or less the same way as the data. The detailed agreement is gob-smacking – multiple peaks, their positions, their sizes, their widths, their second and third order curvature. Any scientist will look at this and think “no way is this a fluke”. Don’t let doubters trot out that coffee-time stuff about being able to fit anything with enough parameters. Firstly, that old chestnut is largely nonsense, and secondly, the fit quality is way beyond that.

Parfait. Everybody love parfait

And yet – outside the world of the CMB, the CDM paradigm has problems, as we were reminded here at ROE the day after Planck Day, in a nice wee coffee talk by Jorge Penarrubia.The best known problems are that CDM predicts far too many dwarf galaxies, and galaxy profiles that are much cuspier than observed,  but there are other claims, such as the existence of a very unlikely polar structure of dwarfs surrounding the Milky Way,  and of suspicious uniformities in galaxy rotation curves. All these problems may still get patched up by astrophysical fixes to do with feedback, IMF games, etc… but we don’t know yet.

Its common to hear people say that Particle Physics and Cosmology are in a similar situation – a model rather than a theory – a perfect fit but with no explanation. Why do all those parameters have those particular values? It seems so arbitrary. And what are dark matter and dark energy anyway? In this view, even the boringly successful fit is tantalising because it tells us there must be a deeper theory in waiting that will explain the perfect fit.

That may describe particle physics, but it doesn’t describe cosmology. Rather, what we have is perfect but fishy. How can the model be so perfect in some places and so poor in others? Have we missed something simple? It may well be that the astrophysical fixes do the trick, and then cosmology does look indeed like particle physics – successful but unsatisfying. Or it could be that some conceptual change is needed, and a revolution is waiting.

I’d give the revolution 2:1 against. But thats good enough odds that your eyes are glued to the table… Mesdames et Messieurs, faites vos jeux.


The book next door

January 4, 2009

Today I went out for a haircut and came back with a book. I got the haircut ok, but walking back I stumbled across Bells Books, one of those lovely musty second hand bookstores. I emerged with A Sourcebook in Astronomy 1900-1950, in which Harlow Shapley collected original articles by all the right people – Ed Hubble, Annie Cannon, Tommy Gold, Albert Einstein, Abbe Lemaitre and many others, as well as many others you won’t have heard of. I guess I was in the mood for the history of astronomy. My teacher and good colleague Mary Bruck died in December, and I just finished her excellent book on Mary Agnes Clerke.

Carrying my lucky find, I remembered a discussion I had in the Kings Buildings branch of Blackwells last year. The shop had big piles of Course Textbooks but very little else. The guy on the till explained how Amazon was killing real bookshops – why walk to the store, not knowing if they will have the book you want, when you know you can order it from your living room ? We’re not talking about the latest Dan Brown. We are talking about unnusual books, just the right book. The giant warehouse wins when you are hunting in the long tail. But we miss those magical hours  wandering the bookshelves, browsing. Wandering the shelves is more than just entertaining; it maximises serendipity. As the man in Blackwells said to me, three times out of four, the book we buy is the one next door to the one we came in to look for.

Scientific progress can be like this too. You are hunting for one thing and find another. There are several famous examples in modern astronomy. One is the building of the Jodrell Bank telescope to look for meteor showers, only to find persistent radio sources, which would turn to be the most powerful objects in the Universe. Another is Giacconi’s rocket shot of 1962, funded to look for solar X-rays reflected from the moon; instead it found Sco X-1 and the X-ray background. I have heard some X-ray astronomy old timers say that Giacconi always wanted to find something new and unexpected; the lunar fluorescence thing was just a funding excuse. It took another twenty eight years before X-ray astronomers really did detect the Moon. X-ray image of the Moon (ROSAT, June 1990)

Here is the picture, taken by ROSAT in June 1990. This is such a cool picture. Not only can you see the sunlit side of the Moon, but the other half is dark, because it blocks the cosmic background.  You can just see that the sky shines in X-rays coming from behind the Moon.

Sometimes the Snark is a Boojum after all. Decades of struggling to measure the expected deceleration of the universe, which would tell us its mean density, and finally the supernovae show us something entirely new – dark energy – and the universe becomes a more mysterious place.

I wonder what unexpected corner we will turn down next ?