Does inflation theory have any confirmed predictions?

[acesuv]

Inflation theory, as far as I know, is completely ad-hoc (that is, created to fit already available data). What testable predictions does inflation theory make, and which of these have been confirmed to be accurate?

 

[Vanadium 50]

That the universe is flat, homogeneous and isotropic. Flatness has been verified at the few percent level, and homogeneity and isotropy to the few 10^-4 level.

 

[George Jones]

According to recent grad/research level texts, the generation of perturbations via quantum fluctuations:

The most serious of the above three problems is the horizon problem. As we have seen, there are possible solutions of the flatness and monopole problems that do not rely on inflation.

The most exciting aspect of the inflationary cosmological theories described in chapter 4 is that they provide a natural quantum mechanical mechanism for the origin of the cosmological fluctuations observed in the cosmic microwave background and in the large scale structure of matter, and that may in the future be observed in gravitational waves.

In the modern view, by far the most important function of inflation is to generate the primordial curvature perturbation ... It may generate other primordial perturbations too, including the isocurvature and tensor perturbations ... However, the historical motivation for inflation was rather different, and arose largely on more philosophical grounds concerning the question of whether the initial conditions required for the unperturbed Big Bang seem likely or not.

Originally inflationary scenarios were suggested as a pseudo-solution to certain pseudo-problems; these are only of historical interest today and the only reason to take the possibility of an inflationary phase in the early universe seriously is because it provides a mechanism for generation the initial perturbations.

 

[acesuv]

That the universe is flat, homogeneous and isotropic. Flatness has been verified at the few percent level, and homogeneity and isotropy to the few 10^-4 level.

I thought the homogeneity of the universe (and perhaps its isotropy) were actually problems that the inflation theory set out to remedy. Are you saying that the theory has made predictions about the *amount* of homogeneity in the universe?

 

[Mordred]

I thought the homogeneity of the universe (and perhaps its isotropy) were actually problems that the inflation theory set out to remedy. Are you saying that the theory has made predictions about the *amount* of homogeneity in the universe?

Inflation is needed to solve several problems. flatness, homogeneous universe, horizon problem and the monopole problem. Prior to the inflation model. Observations made us realize that these problems would need to be solved in order to have a valid observation to theory model. Key questions however remained. The most important is how much inflation? how many e-folds in a given amount of time? We now know that we need at least 60 e-folds in a very short period of time. Luckily the inflation models did predict that.

Another aspect people often forget about inflation is that it also leads to precise high energy particle physics processes. Due to various energy densities of the universe, temperatures would simply be too hot for some reactions to occur. Such as hydrogen and lithium. These can only form when the universe cools off sufficiently due to an increase in volume.
When we observed the CMB and took measurements of the elements available at the CMB we found those %'s in strong agreement with what the inflation model and particle physics predicted at those temperatures.

So more accurately early observations made us realize those first 4 problems needed a solution. Inflation models provided a viable solution. Later and far more accurate observations confirmed the accuracy of the inflation models. Try to keep in mind prior to the inflation models, observations were far more limited than it is today. The observations at the time were only sufficient to support an expanding universe, with a finite beginning. At that time there was no workable theory on why the universe is as homogeneous and isotropic as it was being observed at that time.

 

[bapowell]

Yes, there are predictions that were made well before data was available. Superhorizon temperature and polarization anisotropies in the CMB (particularly the latter) are important predictions of inflationary expansion. There are currently no other causal mechanisms that produce correlated polarization anisotropies across superhorizon scales. Also, inflation predicts that the density fluctuations should be primarily adiabatic - this has been confirmed to high degree by recent CMB measurements (it didn't have to be this way - the other popular account of structure formation for a long time was cosmic strings which produce very different perturbations). Lastly, the simplest models of single field slow roll inflation predict a Gaussian field of anisotropies with a nearly flat spectrum. This too is consistent with current observations. The presence of B-mode polarization in the CMB (indicative of primordial gravity waves) would be another striking piece of corroborating evidence for inflation. There are currently several experiments underway to observe B-modes.

 

[Mordred]

. Lastly, the simplest models of single field slow roll inflation predict a Gaussian field of anisotropies with a nearly flat spectrum. This too is consistent with current observations.

Are you implying that the simpler single field slow roll inflation models are more accurate to observations than the multi- field slow roll models(or other multifield models)?

I've read numerous papers that suggest that but nothing I found that was conclusive. So if I understood that statement correctly I would be interested in any related material reflecting it.

 

[Chronos]

That's what Planck data suggests, preferring a single slow role model. Here is a good review paper; http://arxiv.org/abs/1312.3720, Inflation after Planck: and the winners are

 

[bapowell]

Sure, there are mutlifield models consistent with data (for example, any multifield model that can be written as an effective single field model). The problem with multifield models is that they generically create isocurvature perturbations which have been tightly constrained by CMB data.

At the end of the day, it is generally possible to find many classes of model that simultaneously fit current data (I've written a few papers on this "degeneracy problem", see for example http://arxiv.org/abs/arXiv:1011.0434). One way to resolve the degeneracy problem is to take a model selection approach, where you basically formally apply Occam's razor to select the most predictive, parsimonious model as the best fit. In this case, single field models that fit the data as well as multiple field models would win out because of their simplicity and economy of parameter space.

 

[Mordred]

Nice, both articles neatly summarizes and supports the single field roll with low kinetic term. Nice to see the reference and reduction of Encyclopædia Inflationaris down to 17 compatible models. lol now I'll have to be curious and go through that reference and see if I can distinguish which ones would qualify.

I'll be studying both articles in more detail later on. Lots to grasp, lol I'm still working on another paper you sent me Bapowell

 

[skydivephil]

If you read Alan Guths account of the history of inflation, it was invented to solve one problem and one problem only: the mono pole problem. It was only realized afterwards that it then solves the flatness and horizon problems and gives an explanation to the origin of structure. That to me is the very opposite of ad hoc.
Accroding to WMAp and PLanck teams, inflation has passed several experimental tests , in particular detecting the red tilt in the power spectrum , observing a number called Ns=.96 , just as simple inflationary models predict.
However I think the comment made by Planck spokesman George Efstathiou was interesting. he said something a long the lines of inflation looks good after Planck but there are anomalies that could mean that inflation didnt happen at all and we should keep an open mind.
There is more data to be mined in the polarisation maps and this may require a new mission like these:
http://arxiv.org/abs/1102.2181
http://arxiv.org/abs/1306.2259
or even this very unlikely mission to directly detect primordial gravity waves.
http://en.wikipedia.org/wiki/Big_Bang_Observerver

My take, and I don't know if others agree, is that inflation is not an ad hoc theory but is probably not the whole story, the data so far looks good for inflation but it is not definitive enough yet that we can't consider rival theories e.g VSL, CCC, ekpyrotic, higgs cycles, Horava gravity etc.
Even if we can establish that some form of inflation there are still many different inflationary models that could be proposed and different ideas for what powered inflation to begin with.

 

[phyzguy]

Inflation theory, as far as I know, is completely ad-hoc (that is, created to fit already available data). What testable predictions does inflation theory make, and which of these have been confirmed to be accurate?

I understood that one firm prediction of inflation (at least of the simplest models) was a red-tilted power spectrum. I believe this was predicted before the measurements were made, and the prediction has been borne out by observation. The current Planck data shows a primordial tilt (ns) of about 0.96. Can anyone comment as to whether or not it is correct that this red tilt was actually predicted before the measurements were made?

 

[Vanadium 50]

I thought the homogeneity of the universe (and perhaps its isotropy) were actually problems that the inflation theory set out to remedy. Are you saying that the theory has made predictions about the *amount* of homogeneity in the universe?

At the time of the Guth 1981 PRD, the universe was known to be flat to within about an order of magnitude. Inflation predicted that it be exactly flat (well, at least to many decimals). We know now that it is flat to a couple of percent. I would call that a prediction, and a successful one: after all, we might have discovered that Omega was 2 or 0.5 and that would have falsified this model.

 

[skydivephil]

http://map.gsfc.nasa.gov/news/7yr_release.html

"The 7-year data provide compelling evidence that the large-scale fluctuations are slightly more intense than the small-scale ones, a subtle prediction of many inflation models."

 

[bapowell]

I understood that one firm prediction of inflation (at least of the simplest models) was a red-tilted power spectrum. I believe this was predicted before the measurements were made, and the prediction has been borne out by observation. The current Planck data shows a primordial tilt (ns) of about 0.96. Can anyone comment as to whether or not it is correct that this red tilt was actually predicted before the measurements were made?

Yes, the earliest models of slow roll inflation (Linde's "chaotic" models based on polynomial potentials V\propto \phi^p and those based on Mexican hat potentials with V\propto 1-\phi^p near the maximum) predict red spectra. These were studied in the 90's well before WMAP was able to weigh in on the problem. It's only really with Planck that we can say with much confidence that n_s < 1 when there is no running of the spectral index.