Tuesday, October 4, 2011

Congratulations Cosmology

The news today had an almost poetic structure to it, as if it was part of a well crafted story.

Firstly, for all those rock dwellers out there, the Nobel Prize in physics, announced today, went to two groups who, in 1998, turned the world of cosmology upside down. Their discovery subsequently planted a great big question mark in the rest of the world of fundamental physics, for it was the observations made by today's Nobel prize winners that caused the entry into the standard cosmological model of that most mysterious of concepts:
Dark Energy
The observation these groups made is all the more beautiful for how mundane it actually is. They simply made the observation that a bunch of distant supernovae were dimmer than had previously been expected. But, still today, nobody has any good idea what causes this dimming. Subsequent cosmological observations have made it clear that the dimness of these supernovae probably means the expansion of the universe is accelerating, but why? And if it isn't accelerating, what profound property of our universe is so cleverly mimicking this acceleration.

This was great news for cosmology and if the fact that observations of exploding stars can completely change our understanding of the make up of the universe wasn't poetic enough, today also happened to be the day that ESA announced its next wave of big experiments. One of these was Euclid (artist's impression below). The confirmation of Euclid's eventual launch is also great news for cosmology. On the same day that one beautiful cosmological experiment wins the most prestigious prize available to a scientist, another beautiful cosmological experiment is announced.

What is ESA's stated goal for Euclid? Nothing else but:
To understand the nature of dark energy and dark matter by accurate measurement of the accelerated expansion of the Universe through different independent methods.
And so, while today the big news headlines were of the Nobel Prize being awarded to the two groups who pointed out an enormous cosmological mystery, it is not outside the realm of possibility that hiding subtly in the news background the experiment that will solve this mystery was also today finally made a certainty.

You couldn't write a better script if you tried.

An artist's impression of the Euclid satellite


  1. Of course, Euclid could easily end up just making the mystery even worse, or even get scooped by some other experiment. But who would write a story like that?

  2. And, of course, here in Australia the headline read "Local astronomer..." There was some mention of the universe expanding at an accelerating rate, mixed in with the fact that he was a joint US-Australian citizen, living in Australia.

  3. Thank you so much for following the blog, John. It is very much appreciated. I'm really disappointed that I'll miss your visit to NZ in November by just a month.

    I think Brian Schmidt counts as a local astronomer for Australia. In fact, Aus and NZ have a joint bid on at the moment for the "Square Kilometre Array" (SKA) radio telescope which is going to be a pretty impressive experiment. It would be pretty cool for the two countries if they won the bid. Maybe a new Nobel prize laureate floating around the astronomy community might help the cause a little!

  4. Was it a visual or a numerical observation when the Nobel group suggested that distant supernovae were "dimmer" than they had though?

    The word reminds me of your comparison of some cosmological work as tracing the after-effects of a light event, in a sense that a photograph makes a image out of a light event.

  5. This question keeps coming up in the comments and I keep promising to answer it. I keep not answering it because I keep having too much to say. I promise to write a careful post about it eventually.

    But to be brief and not particularly comprehensive. In the end it always comes down to a numerical observation. Maybe occasionally someone might notice something by eye at first, but it will be translated into a number by the time it gets published. With the supernovae though it was definitely numerical right from the beginning. Too much science about the supernovae themselves had to go into it before one could even say how "bright" one particular supernova was, such that the eye alone would tell you nothing except that you saw a supernova.

    The reason why it has to become a number in the end even when the eye noticed it first is because the number can quantify something unambiguously. If you see something and tell me it was red, this gives me a little bit of information. If someone else measures the light and tells me its frequency I know so much more about that light.

    That isn't to say that the image tells you nothing. The image definitely helps point the way. Remember my post about "pink elephant" clusters. The second most extreme galaxy cluster (or even possibly the most extreme - it depends on how you define extreme) that has ever been observed was actually discovered serendipitously. An X-ray telescope had photographed something else in the sky and someone was looking through this telescope's image database and saw something interesting in the field of view (we can't "see" X-rays, but we can see an X-ray photograph) that he rightly guessed might be a cluster. When the quantitative analysis was done it turned out to be a very big cluster that hadn't been seen before because it was very far away. (small caveat: this is the story I was told. I can't say with certainty it is correct. In the absolute flukish chance that someone involved in that discovery ever reads this comment... I'm referring to XMMU J2235.3-2557 please verify or deny)

    So in that case at least, the image and the eye started the discovery. But I'm sure many other people have "seen" clusters in X-ray photographs only to find out that their eyes had tricked them when the quantitative analysis was done.

    And definitely with the supernovae, these guys won't have *seen* their result coming until the numbers were spat out at the end. There is also the fact that no individual supernova that they saw would have been on its own "too dim", the statistical significance required a number of supernovae to be detected.

    I hope that helps. I will definitely, one day, write a proper post on the topic of how images (moving and still) help discoveries in a quantitative science. In the meantime, feel free to keep asking questions and tease al the info out of my bit by bit.

  6. I've been told before that "brief" means something else to the rest of the world than what it does to me.

  7. Maybe I'll end up teasing it all out of myself anyway...

    One type of image that we make copious use out of are figures that plot the relationship between one parameter and another. These can vary from a simple line plot to more complicated curves that show in a two dimensional space what parameters are and are not excluded by measurements. These aren't replacements for the numbers so much as ways to visualise the numbers. When we discuss a project we will often be drawing these figures on the black/white board explaining what we expect the final result to look like, etc.

    From this perspective, the supernovae result was very much a visual result. They have a final figure where they plotted two variables that fill the place of brightness and distance. I haven't been able to find a good example of this plot on the internet (readers, again, if you know of a good plot, please add a link). I would say this plot was what convinced the community of the result at the time than the bare numbers it is illustrating.

    OK, if you're feeling ambitious, here: A nobel prize winning paper is one of the two original papers. Look at pg 23, Figure (b). The solid black line that moves downward and to the right is what would have been the most popular model before this paper. The solid line that is an exact horizontal line is very close to what is now the most popular model. It is relatively clear that the modern model fits the red data points better than the one from 1998. Congratulations if you've survived this long.

    The type of exclusion plots I was referring to as well in this comment can be also be seen in this paper. For example all the plots on pg 27.

    So, long story short, even if it is just the numbers that matter in the end and not an actual picture of a supernova, we still find it very useful to visualise these numbers in figures.