As mentioned in the post by Josquin elsewhere in this blog, weak lensing has great potential for cosmology and one which we still have to wait to be fully realized. A blessing one might almost think then.
Not always so though ...
In fact, from some perspectives it is merely a nuisance. This is the case, for example, for the hunt for a primordial gravity waves imprint onto the Cosmic Microwave Background (CMB) anisotropies pattern. CMB is roughly a sea of photons left over after the very hot and dense phase of the evolution of the Universe. Those photons traveled nearly freely for most of the time and some of them are reaching our detectors just now. They come to us from all directions and their properties are remarkable direction-independent, reflecting the fact that our Universe is on average (though today only on sufficiently large scales) isotropic and homogeneous. Small deviations, called anisotropies, from the average properties have been observed and attest to a presence of primordial seeds which started off the growth of the structures, e.g., galaxies, and their clusters, as seen so abundantly at this time.
It has been known for over a decade now that a rather peculiar mode of the polarization of the microwave photons, the one where the polarization directions are arranged in patterns reminding swirls, could be generated by the gravity waves. So if the "swirls" in the polarization are indeed found that could be a tale-telling sign that those were indeed present in the very early Universe, i.e., at the time when the initial conditions for the subsequent evolution of the Universe were set. The search for this polarization signal, commonly referred to as a B-mode polarization (as opposed to the E one, not A, as one might expect) is nowadays one of the major drivers for the experimental effort going on in the CMB field. The task is daunting, if only because of the minute amplitudes of the B-modes as expected from our current theories of the early Universe. It is even more daunting, if all the 'real-life' effects, ranging from instrumental systematics to polarized emissions of galactic and extragalactic origins, are, and will be for any actual experiment, added to the mix from which the primordial signal is to be extracted.
And then there is also lensing ...
The CMB photons on their way to our detectors are subject to a gravitational pull due to growing mass overdensities, which they pass by. It bends, or lenses, somewhat the photons trajectories as compared to the original ones, changing the statistical properties of the CMB measurements. This affects both the total intensity of the CMB photons and its polarization. The most insidious feature of this effect, at least from the perspective of the B-mode science, is a conversion of some of the E-mode polarization to the B-mode. The effect is strong enough that this converted part swamps the primordial B-part by orders of magnitude over a range of angular scales. At first sight this seems to make prospects for the primordial B-mode hunters rather bleak.
Don't let a despair to take hold of you though. There is a bright(er) side.
Let us first consider what options are we left with. One way out of the problem could be to correct the effect 'on average'. The averaging here is meant over an entire set of possible Universes. Simple as it is, this approach has one serious drawback. Though, by construction, on average we will indeed do well, at the end we have only one Universe to look at and we should rather care about the correction which works well enough in our specific case. In our CMB jargon we say that this method suffers due to the extra cosmic variance, which boosts up the uncertainty of the result in the case of our specific Universe.
We may like therefore to think of some better ways of doing the correction, ones, which are better tailored to our specific observed Universe. Such methods have indeed been proposed. Nonetheless even for these refined methods, one may wonder if we do not suffer some extra loss of precision. If such loss is unavoidably present, it could give rise to some ultimate limit defining how well we can ever detect the B-mode polarization, and how low a level of it we can in principle see. Presence of such a limit was indeed claimed in some earlier papers, but later contested by follow-up research. The authors of this recent paper support the latter conclusion, using a simple and intuitive, even if not strictly exact, reasoning, building up our confidence that no limit is indeed present. The fact, by far non-trivial, is no doubt good news for the prospects of B-mode lensing cleaning.
The remaining message of the paper is somewhat less optimistic. The authors show that the proper "de-lensing" can be only achieved with the high quality, high resolution, and high signal-to-noise polarization measurements. Neither external data, say galaxy clustering etc, nor total intensity CMB observations can provide much help here. So clean we will, but it will take a lot of work and time, both on experimental and data analysis levels.
But yet again we'd better stay positive and think of that as yet another exciting opportunity ...
No comments:
Post a Comment