Here is the second entry of my “Top arXiv papers of the week” column. The usual warnings I already pointed out in the first entry hold here too, in other words your mileage may vary. With these caveats in mind, hope you enjoy!
#1 2001.02451: Is the Hubble tension a hint of AdS around recombination? by Gen Ye and Yun-Song Piao
By far the most interesting paper I read this week! I’m guessing everybody who reads this column is familiar with the H0 tension, i.e. the discrepancy between local distance ladder and high-redshift (CMB+BAO+SNe) estimates of H0. What not everybody might be familiar with is the concept of (Anti-de Sitter) AdS vacuum, but you can basically think of it as a negative cosmological constant (Wikipedia explains this quite well). AdS vacua are very strongly motivated by string theory, especially given all this business with the swampland conjectures (with my colleagues Luca Visinelli and Ulf Danielsson, the latter a very well-known string theorist and probably one of the best-known Swedish physicists, we looked for signatures of a negative cosmological constant at late times, see our paper in Symmetry). It is by now understood that constructing late-time solutions to the H0 tension is very challenging, because BAO and SNe data (assuming you believe in them, which not everybody would sign up for) constrain the expansion history too much to allow much wiggle away from LCDM. One of the most promising solutions was proposed by Marc Kamionkowski’s group in Poulin et al., and envisages a phase of early dark energy (EDE) behaving as a cosmological constant at early times, before redshifting away faster than radiation, before recombination. In this way you reduce the sound horizon, which results in you inferring a higher H0 to keep the angular scale of the CMB first peak fixed. However, EDE does not completely solve the H0 tension. Long story short, the local measurement gives H0 of about 74 (in appropriate units) whereas EDE gives you H0 of about 70-71ish at best.
How can one improve this? Well, the name of the game is to have more energy injection by EDE (the parameter both Poulin and the authors call f_EDE), but this also means that you then have to dilute this energy much faster, or risk screwing up recombination. Ye and Piao found a clever way of doing this, by using a potential wherein the initially frozen scalar field rolls down to the bottom of its potential, which is a negative plateau, i.e. it enters an AdS phase. Shortly after recombination, the field still has enough energy to climb out of the AdS vacuum and enter the dS vacuum of LCDM. The key point is that during the AdS phase, the equation of state of the field is w>1 so the energy stored in the EDE field is diluted faster than in the original EDE model (although many would argue that an equation of state w>1 might be unphysical). Within this model, the authors manage to get H0 of about 72ish, which is probably one of the best solutions to the H0 tension I have ever seen so far, at least at face value. What troubled me a bit upon reading this paper is that in the MCMC analysis the authors did not vary a certain parameter called alpha_ads and related to the depth of the AdS vacuum, because it was numerically troublesome (basically the chains would prefer to lower this parameter rather than converge). I worry that if one actually found a way to vary this parameter, it might lead to a lower value of H0 or at least to larger error bars (although I am not totally sure, and the only way to find out would be to try). The potential chosen by the authors is quite phenomenological but I would say that, as a proof-of-principle, it’s perfectly fine. Still, this was by far the most interesting paper I read this week and unless I am missing something obvious, this paper provides one of the most promising solutions to the H0 tension I have ever seen.
#2 2001.01301: Echos from the braneworld black holes by Ramit Dey, Sumanta Chakraborty, and Niayesh Afshordi
For those of you not familiar with the concept of echos in black hole (BH) physics, an echo-like pattern in the ringdown phase of a binary BH merger (i.e. the phase where the composite object relaxes) can be thought of as a post-merger repetition (with a given time delay) of the gravitational wave (GW) signal. This can be caused by partial reflection of the incoming wave due to Planck-scale structure around the event horizon, as motivated by several proposed solutions to the BH information paradox (such as fuzzballs or firewalls), as well as tentative resolutions to the cosmological constant problem (CCP). One of the authors, Afshordi, is well known for having found tentative evidence of echos from the first three GW events detected by LIGO, that is, GW150914, GW151226, and LVT151012. Mathematically speaking, echos arise from so-called reflective boundary conditions at the horizon (whereas the usual boundary conditions entail purely out/in-going waves at spatial infinity/at the horizon). The question is then whether, besides these proposed solutions to the BH information paradox/CCP, there are models beyond GR where these reflective boundary conditions are a necessity and thus one would expect echos on completely general grounds.
The authors find one such well-motivated model, i.e. brane-world models. For a popular account of how brane-world models work, you can read this post I wrote on the OKC blog 2 years ago, but long story short is that our 4D spacetime is restricted to a brane living in a higher-dimensional spacetime known as bulk. If we localize a BH on the brane and we believe in the AdS/CFT conjecture, then gravity-CFT coupling on the brane should quantum correct this BH, which leads one to expect that the boundary conditions at the BH outer apparent horizon (which does not necessarily coincide with the BH event horizon) should not be purely ingoing, but partially reflective, hence echos. The authors also find that current LIGO/Virgo data can place limits on the tidal charge of brane-world BHs (heuristically speaking, the tidal charge is the result of gravitational effects from the bulk being projected onto the brane, so it records information on the extra dimensions) of order Q~M^2, with M the BH mass. I found this paper very interesting and well written (Chakraborty, who was a PhD student of the well-known Thanu Padmanabhan, has written many important papers on extra dimensional models and their observational signatures), although personally it left me with two open questions: a) what happens when we include angular momentum (all BHs we know are rotating)? and b) there seems to me to be a slight inconsistency in that the authors use what looks to be a classical BH line element, but wouldn’t one expect that to be corrected by quantum effects? Anyway, a very interesting read which I would definitely recommend to anyone interested in BHs, GWs, extra dimensions, echos, or near-horizon Planck-scale structure.
#3 2001.00948: Speeding up Dark Matter with Solar Neutrinos by Yue Zhang*
I’ll assume most readers of this column are familiar with the idea of dark matter (DM) direct detection, where you look for the energy deposited by DM recoiling off target nuclei. If you think about the usual exclusion plots in the mass-vs-cross-section-plane produced by experiments such as XENON, LUX, and so on, you might recall that the exclusion limits quickly get crappier at low masses (usually below ~GeV) because if the DM is too light it can’t deposit enough energy in the target nuclei, hence it is not detected. The question Zhang asks is then whether one can think of a mechanism which speeds up sub-GeV DM (or at least part of it) so that it might leave a signature in direct detection experiments?
The model constructed relies on a relatively simple neutrino portal interaction, which allows neutrinos to interact with DM. Crucially for the model to work, the DM needs to form stable bound states (which is achieved via a simple Yukawa interaction with a light scalar). What then happens is that solar neutrinos can be energetic enough to dissociate the bound state, speeding up one of the two final state particles which then reaches Earth at much higher velocities. The differential flux of this DM component wrt velocity, of course, very closely resembles the solar neutrino energy spectrum, with a broad tail at high velocities. Zhang then uses this model to reinterpret recent PICO results, showing that they in principle (within this semi-model-dependent framework) constrain DM masses as low as 0.05 GeV. The paper is very interesting, although one thing bugged me upon my first reading: is all of this still allowed given that, thanks to surveys such as GAIA, we are starting to understand the local DM velocity distribution much better?
*who should not be confused with Yue Zhao from the University of Utah
#4 2001.01407: Gravitational waves from binary black holes as probes of the structure formation history by Tomohiro Nakama
This paper is based on what in hindsight is a simple and obvious but rather powerful idea. Basically, the BH-BH merger rate depends on the star formation history, which in itself depends on the halo mass function, which in itself depends on the matter power spectrum. The latter is something we would like to understand much better as any deviation from the expected quasi-scale-invariant shape can tell us more about either primordial fluctuations or the nature of DM (especially in relation to the so-called small-scale crisis, which might be solved if the power spectrum is suppressed on small scales). The idea is then that if we manage to measure GWs from binary BH-BH mergers from high-redshift sources, we should be able to constrain the high-redshift structure formation history and hence the matter power spectrum (and by extension possibly the nature of DM).
Nakama shows this explicitly by considering two scenarios for DM: the usual cold DM, and fuzzy DM, where DM is composed of ultra-light bosons and the uncertainty principle results in structure being washed out on small scales. The resulting BH-BH merger rate can be substantially different across the two scenarios and can be probed by future GW observatories such as DECIGO, the Einstein Telescope, and in principle also Pre-DECIGO. As a proof-of-principle the study is really interesting, although the procedure going from the power spectrum to the BH-BH merger rate for alternative DM models would probably need some refining. Nakama used a semi-analytical approach, but most likely a simulations-based approach would be necessary as there is a lot of “gastrophysics” involved therein. This also raises the question as to whether we will in the future be able to disentangle the gastrophysics from more “primordial” (read: interesting) physics? Of course, this is a question which is relevant not only to the paper at hand but to many other astrophysical and cosmological probes.
#5 2001.03120: Field theoretic interpretations of interacting dark energy scenarios and recent observations by Supriya Pan, German S. Sharov, and Weiqiang Yang
When we talk about interacting dark energy (IDE) we usually refer to models where DE and DM interact non-gravitationally. In recent years these models have gained a lot of interest especially on the observational side. I myself have worked quite a bit in this direction. Usually the way one treats these models in cosmological studies is not from first-principles (i.e. starting from a well-motivated Lagrangian) but simply by modifying the DM and DE conservation equations, i.e. the DM and DE energy-momentum tensors are no longer separately covariantly conserved but only their sum is. Their separate non-conservation is governed by a coupling function (usually called Q). The question is then what functional form should we give to Q? Usually people take something like Q~H*rho, where H is the Hubble parameter and rho is something with dimensions of energy density (usually the DM or DE energy density, or some combination of them). The motivation, however, is simply that these phenomenological coupling functions are simple to handle. Personally, I (and I guess many others working on IDE) always had this worry in the back of my mind that such couplings were purely phenomenological and do not come from any motivated field theory.
Today’s paper instead tries to find a well-motivated field theoretical interpretation for these couplings. Pan et al. work with a field theory description where the role of DM and DE is played by two separate scalar fields, and they include a DM-DE interaction potential which is symmetric in the two. Then, they basically show how from relatively well-motivated potentials one can extract 3 functional forms for Q which have been widely used in the literature. Interestingly, this study also addresses something else which had always bothered me (which was first brought up in a referee report I received), i.e. how should a quantity determined from microscopic physics such as Q know about something macroscopic such as the expansion rate of the Universe H? Today’s paper addresses that issue explicitly showing how relatively well-motivated potentials can lead to Q scaling as H. One thing I am still concerned about (again, a concern first raised in this same report I received) is whether all this can work when we factor in the expectation that DM should be composed of heavy fields and DE of light fields. In this case, the two sectors should decouple and not interact very efficiently (see these papers by Guido D’Amico et al., and David Marsh - the Swedish David Marsh). Introducing strong couplings by hand means that the DE scale will be renormalized by exchange of DM quanta and the DE potential will become very steep, in conflict with observations. This issue is not addressed in Pan et al.’s paper, because they only work at tree level and they do not, as far as I can see, introduce any explicit scale for the DM and DE scalar fields. In principle, if both the DM and DE were composed of ultra-light axions, then the aforementioned problem might not hold any longer (to my great relief). Regardless, I found this paper very interesting and I am glad that it at least partially addressed an issue which had always been in the back of my mind every time I have been working on IDE models!