Friday, February 12, 2016

Gravity Waves from Inflation???

Congrats to the LIGO team for the (likely) detection of gravity waves. It looks like they will be presenting us with many more candidates in the near future. I look forward to see the results.

But that leaves us with the question:  can we detect the tensor gravity waves from inflation?
The problem right now is that the data of B-polarized modes from the CMB has some large error bars and/or contamination from non-primordial sources.

For example, below is a plot of those experiments that have released data on the auto-correlation and/or self-cross-correlation between frequencies.
Yellow =  Planck 2015 low l
Light Blue = WMAP 9year
Light Green = Planck 2015 Mid-range l with foreground removed (see prior blog post)
Light Grey = SPTPol 100 day  95GHz x 150GHz
Orange = BICEP2 x KECK   95GHz x 150GHz
Black = Polar Bear1
Brown = Output from CAMB using BestFit Planck2015 TT+TE+EE+lowP+lensing+ext plus r =1
Blue = Output from CAMB using BestFit Planck2015 TT+TE+EE+lowP+lensing+ext plus r =0.1
Grey = Output from CAMB using BestFit Planck2015 TT+TE+EE+lowP+lensing+ext plus r =0.0
where r = initial tensor-to-scalar ratio (which is an input for CAMB)

I think that we can safely say that r = 1 is ruled out. But, from this graph alone, it's hard to rule out r=0.1 or r = 0. So, below is a ZoomIn around the region in which BICEP2/KECK are focused. This time, I plot CAMB output for r = 0.1, r = 0.01, and r = 0. The data points are almost always above the lines, which means that there is likely contamination from foreground or sources of B-modes other than primordial + lensing of E-polarized modes. Also note in the graph below that, if there is no lower error bar, then this means that the error bar goes into negative values. (This didn't happen in the Linear scale plot above.)

Note that the CAMB Output has no foreground added to it. The Planck Mid Range data supposedly has dust removed, but it seems to suffer from erroneous data near l = 225. The BICEP2/KECK data has some dust contamination because there is some dust contamination even at 95 GHz.

All of this just means that we'll have to wait a little bit longer before we can say definitively that we have detected primordial gravitational waves. We should be getting results some time this year from BICEP3+KECK at 95GHz. If this data plus the BICEP2/KECK data at 150GHz is cross-correlated with a B-lensing map and a foreground map to remove lensing+foreground, then one should be able to make some meaningful constraints on the tensor-to-scalar ratio, r. Also, the Planck collaboration is expected to make another release of their data this year (and a final release in 2017.) This data should have BB modes vs. l (for all l), which is something that they have not yet published.
I'm excited to see the results when they are announced.
In the somewhat near future, we should be expecting results from PolarBear2, SPT-3G, and CLASS. This is an exciting time for studying gravity waves produced during inflation.

No comments:

Post a Comment