Header 1

Our future, our universe, and other weighty topics

Wednesday, May 14, 2014

Why a New Planck Paper Casts Grave Doubts on the “Epic” BICEP2 Study

In March the BICEP2 study announced observations of what is called b-mode polarization. The study was declared to be “direct evidence” for the theory of cosmic inflation, the theory that the universe underwent a period of exponential expansion during its first second. But in recent days rumors have been swirling that the study (hailed an an epic breakthrough) may not hold up. Today National Geographic published a piece entitled “Big Bang Discovery Comes Under Fire.”

Let me explain exactly why a new scientific paper by the Planck team throws grave doubt on the BICEP2 study, by making it seem rather likely that the “epic discovery” may be merely the observation of ordinary, run-of-the-mill dust. I will use scientific visuals rather than rumors.

First let us look at a graph from a scientific paper by another scientific team (POLARBEAR) that is doing studies very similar to the BICEP2 study. It is a graph with the same scale and legends as the key graph of the BICEP2 paper. The graph is below.

Graph 1, from the POLARBEAR scientific paper

The graph above shows some purple lines. The lower dark-dashed purple line shows b-mode polarization observations we would expect to see from galactic dust, if a parameter called the polarization fraction (represented by the letter p) is equal to 1.5%. The higher light-dashed purple line shows b-mode polarization observations we might expect to see from galactic dust, if a parameter called the polarization fraction (represented by the letter p) is equal to 10%.

Below is the key graph from the BICEP2 paper. Notice that this diagram and the first diagram are plotting the same thing on the same scale (the only difference being that the POLARBEAR graph goes down slightly lower on the scale).

Graph 2, from the BICEP2 scientific paper

Because these two graphs plot the same thing on the same scale, it is very easy to take the 10% dust polarization line from the first graph and move it to the second graph. When we do that, we get the following graph:

Graph 3, combining the BICEP2 graph with one line from the POLARBEAR graph

This graph should be very worrying for anyone who thinks that the BICEP2 observations are from the Big Bang or cosmic inflation. The graph shows that if there was dust polarization of about 10%, then that could explain the BICEP2 observations (as the purple line in the graph above cuts right through the black dots representing the BICEP2 observations). I don't have a graph showing 13% dust polarization, but that would be a purple line higher than the purple line above, and in such a case all of the BICEP2 observations could be explained from a combination of dust polarization and gravitational lensing, shown by the solid red line in the graph.

So the key issue is: how high is this dust polarization fraction? If it's only 1.5% then the BICEP2 team has little to worry about, but if it's 10% or 15% then all the BICEP2 observations can be explained by ordinary dust and gravitational lensing, and the claims of an “epic Big Bang breakthrough” would seem to crumble completely.

In the BICEP2 paper the scientists refer to a preliminary pdf which they apparently used in estimating the polarization fraction, a PDF with the following visual showing dust polarization fractions in various regions of the sky:

Graph 4, preliminary Planck map used by BICEP2

Notice the legend at the bottom that indicates the dust polarization fraction, which extends from 0% (deep blue) to 20% (red). The BICEP2 observations were from a region around the bottom middle of this map. Based on the map above, you might have estimated that there was a dust polarization fraction between about 5% and 7%. That would translate to a purple line uncomfortably close to the 10% line in my combined graph above (graph 3), but it would still leave a little breathing room.

However, in the past two weeks the Planck team has published a scientific paper that includes a revised version of the graph above. The graph (which excludes certain areas) is shown below:

Planck dust map

Graph 5, the later Planck map published in a scientific paper

As you can see, this is basically the same graph as the previous graph, but there's one big difference. Many of the regions which used to be yellow are now red. This means many of the regions listed as having only about 10% dust polarization are now shown with a much higher level of polarization – a polarization of almost 20%.

The map above does not show the region at the bottom middle from which the BICEP2 observations were made. But the best guess one can make from these two maps (keeping in mind that the second one is the later and more accurate one) is that the polarization fraction in the area of the sky observed by BICEP2 is at least 10%. In fact, table 2 of the above Planck paper tells us that the average dust polarization fraction (considering all regions) is 19%.

Such a result casts very grave doubt on the pretentious cosmic claims of the BICEP2 team. If the dust polarization is greater than 10% in the area observed by BICEP2, it means the dust polarization line is even higher than the purple line I have drawn for graph 3. In such a case all the BICEP2 observations can be explained by ordinary dust and gravitational lensing, not something special that happened in the Big Bang.

It will be quite an irony if this turns out to be true. The scientists who thought they were seeing something extraordinarily special (an echo from the first second of the universe's creation) may really just have detected polarization from something as ordinary and common and lowly as the dust on your shoes.