Tuesday, June 4, 2013

Cosmological perturbations post Planck (CPPP) I

This week Helsinki is hosting a conference on the theoretical implications of the recent results from the Planck satellite. The official theme of the conference is cosmological perturbations post Planck. This is alluding to the fact that on large distance scales and at early times, the universe is very homogeneous (it is almost the same everywhere) but has small perturbations in things like its temperature and its density. Planck measures the temperature of the cosmic microwave background (CMB) today, which is an almost direct measurement of the density of the universe soon after the big-bang. This is because the CMB that came from the more dense bits of the universe lost a bit of energy climbing away from that little bit of extra matter and vice versa, the CMB gained energy falling out of less dense regions. Therefore, Planck has made an accurate measurement of the perturbations of the density of the early universe.

But what, are (some of) the implications of this measurement..? Hopefully this conference will elucidate that a little.

I'm going to do my best to describe what is said as the conference proceeds...

Planck's results

The conference started this morning with Helsinki's Mr Planck, Hannu Kurki-Suonio giving an overview of specifically what Planck found in its measurements. If you want a more detailed summary of this you can read some of my posts from when the data was released. The essence is, however, that the standard cosmological model, which had been settled on by most of the community as the simplest model that fits all the pre-Planck data works very, very well in a post-Planck world. There are a number of things about this model that are uncomfortable from a theoretical perspective, but it fits the data we measure extremely well.

But, there are some anomalies (which Hannu ran out of time to cover), which means there are some aspects of the data that aren't predicted by this simplest cosmological model. The anomalies are anomalies because they aren't overwhelmingly statistically significant. This simplest cosmological model only predicts the statistical properties of the perturbations in the universe and all of these anomalies are technically possible, they're just somewhat unlikely. They also don't have obvious explanations from well-motivated new physical effects. They could be statistical flukes. If you have a big enough set of data and look at it in enough different ways you will find anomalies, that's just what noise is. There are two questions that need asked when considering these anomalies:

  • Are there actually more anomalies than we would expect?
  • For each anomaly, is there a well-motivated model that can generate the effect seen without changing all of the many other things that aren't anomalous (either by completely replacing this simplest cosmological model, or by tweaking it in some way)?

The first question is almost impossible to answer. There are too many ways of looking at a data set this big and it's just too hard to quantify all the ways in which is isn't anomalous. This leaves us with just the second question. The reason why these anomalies are called anomalies is that we weren't expecting things like this and the reason for that is that none of the things we thought were well-motivated deviations from the simplest cosmological model predicted these things.

That's the playing field at the beginning of the conference.

The Curvaton

Many of today's talks were on the topic of the curvaton.

It's going to be hard for me to describe to you what the curvaton is given that I haven't ever properly told you what the inflaton is, but I'll give you a quick whirlwind introduction of both. The inflaton is the field that drove something called inflation. Inflation is a (hypothetical) period early in the history of the observable universe when the universe's expansion accelerated. This period is thought to have happened because it would have smoothed out any pre-exisiting inhomogeneities in the universe (in its density, in the curvature of space-time, in the number density of exotic types of matter, etc). There are issues with this because inflation needs the universe to be somewhat homogeneous even to get started, but despite that inflation still definitely leaves the universe more homogeneous than it found it, so at the very least it helps.

But, the thing that is most interesting about this potential inflationary period is that it would also seed very small perturbations in the otherwise homogeneous universe it left behind. This doesn't sound like much of a gain. Without inflation the problem was that there might be too much inhomogeneity, why should we celebrate this small amount of inhomogeneity inflation leaves behind? The answer to that is that, for a given inflationary model we can actually predict the statistical distributions of these post-inflation perturbations. This gives us something to measure and then compare to theory. In other words, we can gain evidence for or against inflation through observation. There may have been inhomogeneities around in the universe before inflation but we have (almost) no way of predicting them. Inflation lets us make predictions and test them.

So, what is the curvaton in all of this? Well, in the simplest models of inflation there is only one thing other than space-time that is around during inflation. This is the inflaton, the field driving this accelerated expansion. In a curvaton model, there is still an inflaton driving this expansion, but there is also as least one other thing around, the curvaton. And, in these models, it is the curvaton that produces the perturbations that we observe today. The inflaton still produces perturbations, but they decay over-time and the curvaton's don't (as quickly).

Why is this interesting? Why should one study a curvaton model?

That's a very good question. The first, not-quite-completely-joking answer I can give is because you can. It is a possible reality for the universe. This is what theoretical physics is about, thinking about what is possible and exploring the observational consequences if the possible were real. So, from that perspective, why just assume that, if inflation occurred, that a curvaton field wouldn't be present? The counter to this perspective is that it adds complexity to inflationary models.

 Unfortunately, it is now past midnight. More to come tomorrow...


  1. Hi Shaun -- Happy to see that you've decided to blog another conference. The last one made for interesting reading. Certainly looking forward to your posts from the rest of the week.

    Too bad the discussion of the anomalies got cut short -- are any of the speakers making an attempt to comment on them specifically? Doesn't seem obvious from glancing at the slides posted so far.

    1. Hi Daniel, thanks for the comment. I appreciate the encouragement, a lot.

      Regarding Hannu's discussion of the anomalies, I don't think it is too bad that he was cut short. He was speaking mostly as a Planck representative (I think) so he would have just made the same comments about them as people did back in April.

      Regarding other people speaking about the anomalies, Dominik Schwarz spoke entirely about the anomalies, so you can/should check out his slides. David Lyth gave a talk mostly about how the curvaton and a super-horizon fluctuation could explain the power asymmetry (whereas inflation and a super-horizon fluctuation couldn't). But other than that, nobody has yet made mention of them.

      I'm just about to try and write a more detailed post (more detailed than yesterday's slightly poor attempt above) about yesterday and today's talks.