This week’s entry looks at a new H0LiCOW analysis which quantifies the impact of the mass-sheet transform, the possibility of nailing down H0 from dark sirens following future upgrades to LIGO/Virgo, and a novel class of shadows from asymmetric thin-shell wormholes. Enjoy!
#1 2007.02941: TDCOSMO IV: Hierarchical time-delay cosmography -- joint inference of the Hubble constant and galaxy density profiles by Simon Birrer et al.
Recently, measuring the Hubble constant H0 from time-delay cosmography, as done for instance by the H0LiCOW collaboration, has opened a new window onto the H0 tension business. In short, there is now a new class of H0 measurements to reckon with if one wants to solve the H0 tension, a class which appears to be more consistent with the local Cepheid-calibrated SNeIa value rather than with the inverse distance ladder inference of H0 (by this I mean the combination of CMB, BAO, and high-redshift SNeIa). In my Week 4 entry, I discussed the mass sheet transform (MST) as being the first and foremost source of uncertainty in time-delay cosmography. In short, the MST is a degeneracy direction along which lensing observables remain unchanged, but the overall time delay does change, changing the inferred value of H0. This degeneracy is usually broken by making some assumptions concerning the lens mass density profiles, but this introduces some theoretical assumptions which propagate to the final value of H0, which could lead one to question how model-independent this value is. Another source of uncertainty in time-delay cosmography is the assumption that the errors of each individual lens are uncorrelated. In this week’s paper, Birrer et al. revisit these assumptions and improve the H0LiCOW analyses by making use of a Bayesian hierarchical framework, which ends up quantifying the importance of the MST. Importantly, in order to better constrain the deflector mass density profiles, an important ingredient towards breaking the degeneracy associated to the MST, Birrer et al. use imaging and spectroscopy data from the SLACS sample. This includes 33 lenses, 9 of which have resolved kinematics which allow one to constrain stellar anisotropy. It is worth pointing out that all the analyses in this paper are blind.
This is a long and technical paper, but the money plot is undoubtedly Fig. 18, and in particular the four new constraints on H0 appearing in the lowermost panel. Overall, it is clear that the four measurements in question are consistent with each other. This is at least partially due to the uncertainties being enlarged with respect to e.g. the earlier 1907.04869 by Wong et al. by about a factor of 3. This is mostly due to the fact that virtually no assumption on the radial mass density profile of the lens galaxy has been made (and I’m guessing also partially due to the fact that covariance between lenses has been accounted for), and the uncertainty associated to the MST is totally accounted for. With these larger uncertainties and overall statistical consistency between different analyses in mind, it is nonetheless curious that the analyses using 33 SLACS lenses with SDSS spectroscopy find a central value of H0~67 (as opposed to the previous 74ish), more in line with the inverse distance ladder inference rather than with the local Cepheid-calibrated SNeIa value. However, it is also worth pointing out the underlying assumption that the TDCOSMO and SLACS galaxies are drawn from the same parent population, the validity of which I cannot judge. Again, I cannot stress enough how it is important to remember that the uncertainties are large. What I think is really important is that this paper finally quantifies the contribution of the MST to the uncertainty budget (in this sense, I wonder whether this paper addresses to some extent the concerns raised by Kochanek in 1911.05083). Again, this paper is extremely interesting and I don’t want to write anything which might be “politically inconvenient” for H0LiCOW, but one might indeed wonder whether these results (see again the lower panel of Fig. 18) might temporarily “remove” so to say H0LiCOW from the H0 tension debate, at least until a smaller uncertainty can be reached? I leave the question to readers of this blog (and would love to hear your thoughts, if you have any, in the comments section!).
#2 2007.02883: Dark Sirens to Resolve the Hubble-Lemaître Tension by Ssohrab Borhanian et al.
I’ve written about the H0 tension quite a bit in this blog, and while I’ve mostly focused on solutions to this tension, an equally worthwhile direction to pursue is that of arbitrating the tension. In other words, independently determining H0 to increasing accuracy, to figure out whether one or more of the measurements in question might be, to put it shortly, incorrect. One interesting possibility in this sense is the use of gravitational waves (GWs) as standard sirens (SS). The idea, in short, is that fitting a binary system merger GW signal one can obtain the luminosity distance to the binary. If one can identify an electromagnetic (EM) counterpart, one might obtain a spectroscopic redshift for the host. Armed with redshift and luminosity distances, one can essentially construct a Hubble diagram from GW SS, which can be used to constrain the low-redshift expansion history, including H0. The possibility of arbitrating the H0 tension in this way had been studied by Feeney et al. in 1802.03404, focusing on binary neutron star (BNS) systems, for which EM counterparts are to be expected. There it was found that ~50 BNS mergers could be sufficient to arbitrate the H0 tension.
The problem is that BNS are not the only binary systems which produce GWs, and many other systems are not associated to EM counterparts. An example are binary black hole (BBH) mergers, of which LIGO has detected several. Nonetheless, one can still determine the luminosity distance to the BBH host even in the absence of an EM counterpart, using a more complex statistical approach formulated for instance in 1108.1317. Heuristically, this statistical method requires identifying all possible hosts by making use of overlapping galaxy surveys, and weighing these hosts accordingly. Due to their possible use as SS even in the absence of an EM counterpart, BBHs are sometimes referred to as dark sirens. For instance, in 1901.01540, a combination of LIGO/Virgo and DES data was used to measure H0 from the BBH GW170814, although the resulting uncertainty is, not surprisingly, very large, and unable to arbitrate the tension.
The question Borhanian and collaborators address in this week’s paper is then how can one improve constraints on H0 from dark sirens in light of planned sensitivity upgrades to LIGO/Virgo. One of the main motivations from the present work are the recent detections of GWs from the compact mergers GW190412 and GW190814 (I wrote about the latter in my Week 26 entry). Thanks also to their unusually large mass ratios, these were the first systems for which higher-order multipoles beyond the usual (l,m)=(2,2) were detected with high significance. The importance of higher-order multipoles lies in their ability to break the distance-inclination degeneracy, hence improving distance constraints from dark sirens. The larger sensitive volume to follow from these upgrades would also help the localization of the dark sirens, which would in turn help nail down the host and get a spectroscopic redshift from EM follow-up. The final result is that in the future one should be able to measure H0 from dark sirens to 2%, which could help arbitrating the H0 tension. This conclusion relies on being able to control the amplitude calibration of the detectors to below 1%. Overall, it looks like the future of cosmology from dark sirens could be, oxymoronically so, bright.
#3 2007.03327: A novel shadow from the thin-shell wormhole by Xiaobao Wang et al.
It’s been a while since I wrote about black hole (BH) shadows, which nonetheless remain for me an active area of research, so let me just recall that a BH shadow is the apparent (i.e. gravitationally lensed) image of the photon sphere (the region of space-time where gravity is so strong that photons travel in unstable, not necessarily circular, orbits). In other words, the BH shadow is the closed curve on the sky separating capture from scattering orbits, and is not a direct image of the event horizon. Thanks to the image of M87* delivered by the Event Horizon Telescope, we have now entered an era where we can start to test 1) the geometry in the BH vicinity, i.e. the BH metric itself or, even more excitingly, 2) distinguish BHs from other dark, compact, and possibly more exotic objects. One such example is that of wormholes, a mathematical solution to the Einstein equations which heuristically connects two widely separated space-time regions (anyone is a fan of the Netflix series Dark, as I am?). Not all wormholes cast shadows, but if they do, the latter can be an useful way to distinguish them from BHs (see for example 1803.11422, 1805.11591, or 1806.07782). In this week’s paper, Wang and collaborators focus on one particular class of wormholes, usually referred to as asymmetric thin-shell wormhole (ATSW). ATSWs, first introduced by Visser in 0809.0927 (though the article appeared about 20 years earlier than the preprint, no this is not because of a violation of causality due to a wormhole, see the arXiv comments), basically consist of a static thin shell connecting two separate Schwarzschild BHs with different masses. Visser refers to the process of constructing ATSWs as surgically grafting two separate Schwarzschild spacetimes in such a way that no event horizon forms, which makes sense if one thinks of the mathematical procedure for constructing ATSWs (outlined in this week’s paper) as basically a cut-and-glue operation.
By studying geodesics in this spacetime, Wang and collaborators find novel shadows which had not been studied previously. Heuristically, the main point is that while BHs are one-way objects, for ATSWs some photons can actually “bounce back” to the original spacetime. This means that, in general, ATSW shadows will be smaller than that of corresponding Schwarzschild BHs with the same mass. Moreover, for the mechanisms identified by the authors to work, the asymmetry between the Schwarzschild masses on either side of the thin shell plays a key role. Interestingly, such a shadow can be cast even if there is no photon sphere on the observer’s spacetime (in my Week 16 entry I already discussed the possibility of a shadow being cast even without photon sphere). Fig. 1 of this week’s paper is particularly useful in this sense, outlining the conditions necessary for the novel ASTW shadow to exist. At the moment I would say the prospects of distinguishing ASTWs from BHs using their shadows is weak, mostly because one would need to determine the mass of the object in question (which uniquely fixes the expected size of the shadow for Schwarzschild BHs) to high fidelity, while unfortunately, current BH mass determinations suffer from >100% systematic uncertainties. If you want to find out why, you can read Section 3A of my recent paper with Cosimo Bambi and Luca Visinelli 2001.02986. Nonetheless, this remains a very interesting paper, particularly in light of planned upgrades to the Event Horizon Telescope and future VLBI projects.