Beyond Precision Cosmology
Raul Jimenez (right) gave an overview of the current status of cosmology and the future prospects. In the past 20 years have made great strides in the understanding of cosmology. We have six numbers to describe the Universe composition and evolution. However, we do not know what the energy density of the Universe is, and we do not know what dark matter is. We also do not know what the inflaton is, and have not discovered gravitational waves from inflation.
There is a big difference between modeling and understanding. Cosmology is special, inasmuch as you can do measurements, but no experiments. We have to use the entire Universe as our laboratory. We can use cosmological data to test standard physics.
There is a wonderful agreement of new data with the Lambda-CDM model, with some notable exceptions. The Universe is for 73% dark energy, and 23% dark matter. This is a nice model, but it is incomplete. Neutrinos have mass; and it is unsatisfactory, as we have the cosmological constant problem – what is dark energy ?
As spectacular as the CMB is, it has saturated the information content. The next challenge is the large-scale structure. A forthcoming avalanche of new data will enable precision tests beyond the standard model, and neutrinos, which contribute at least to 0.5% of the total matter density, will be a part of the picture.
The current paradigm describes well all observations, however, there is a tension between the local value of H0 and the one derived from the CMB, and this is significant. This can be alleviated if there is extra radiation in the form of N_eff=0.4; however, this solution is disfavoured by Planck polarization data. Model independent measurements of the expansion history are going to stay with us even as we evolve our model. There is already a normal ordering preference from cosmology, with odds of 50 to 1.
The trouble with H0 is that if you measure the expansion rate is 73.24+-1.74 km/s/Mpc, while from Planck you get 67.8+-0.9 km/s/Mpc. Formally it is a 3.4 sigma effect, but maybe the model is wrong. The error of the Planck measurement is dominated by cosmic variance, given by the fact that you can only measure a finite number of modes at a given scale.
A combination of data can be used to figure out what is the expansion history of the Universe is. One can use supernovaes, Baryon acoustic Oscillations (BAO), and Planck data. When Planck will release their final analysis, if it stays what it is we will still not know how to reconcile the fact that the expansion history does not match.
At half of the age of the Universe (redshift 1) you can find that the expansion history is compatible with the Planck-CMB lambda_CDM model. The problem is at the edges, either at the early time or at the other end, z=0 or z=1100. In summary, the H0 trouble is a mismatch of anchors, not evidence for strange expansion history.
Jimenez finished his talk by talking about the constraints from cosmology on the neutrino mass hierarchy. By using a Bayesian analysis, he claimed that the normal hierarchy is already favoured at 50:1 odds.