New measurement of the Hubble constant is consistent with the CMB value

[phyzguy]

TL;DR Summary

New measurement of the Hubble constant is consistent with the CMB value.

This paper just came out with a new measurement of the Hubble constant based on the technique of gamma ray attenuation. The result is consistent with the lower (CMB-based) value. Interestingly, they also do a joint analysis of several non-CMB techniques (BAO+BBN+SN+γ-ray attenuation), and find a value completely consistent with the CMB value. Clearly, the jury is still out on this issue, and recent pop-science articles about a "crisis in cosmology" are a bit premature.

 

[Arman777]

I want to ask something off-topic. Is there a some site or some journal that you are following so that you can learn about these articles or its just from random searches/news ?

Both the γ-ray measurements and the EBL models may be subject to hidden systematic uncertainties. We account for these effects by fitting a systematic error in δτ /τ as an additional nuisance parameter
We assume that the systematic error is independent of the statistical uncertainties in the measurements of γ-ray attenuation. In order to exhaust all possible trends, we also assume that the systematic error is a power-law function of γ-ray energy and redshift, where power-law indices are additional free parameters fitted to the data. All cosmological constraints presented in our work are marginalized over the nuisance parameters describing the systematic errors. For both EBL models, the obtained systematic errors are smaller than or comparable to the statistical uncertainties. All fits point to a statistically significant dependence on energy, with the systematic error increasing with decreasing energy.

And something seems odd to me. They are fixing the $\Omega_{m} = 0.32$ and this leads to $H_0 = 67.5 \pm 2.1$ but when they fix the Hubble constant at $H_0 = 68$ They get $\Omega_{m} = 0.21\pm 0.08$

I don't know statistics so it just seemed odd to obtain such matter density value.

I guess this is an another experiment that takes side to the CMB. If the true value is obtained from the CMB then there should be some major systematic errors in the SN (also Gaia etc.) measurements. I think the crisis is still going on.

 

[phyzguy]

I usually go to the new astrophysics papers section on the arXiv and see what is there. On your other question, I think the most relevant result is when they have fit for both H0 and Ωm, which you see in the graph below. At least in this paper, they conclude that the CMB measurements (red contours) and the combined other measurements (black contours) are consistent. Identifying the unkown systematic errors in each of these measurements is always the hard part.

 

[Arman777]

Thanks.

Identifying the unkown systematic errors in each of these measurements is always the hard part.

Well yes I agree. I am doing a project about Hubble Tension. These recent papers are great for me. Something new is happening every day

 

[kimbyd]

And something seems odd to me. They are fixing the $\Omega_{m} = 0.32$ and this leads to $H_0 = 67.5 \pm 2.1$ but when they fix the Hubble constant at $H_0 = 68$ They get $\Omega_{m} = 0.21\pm 0.08$

What this says, essentially, is that the gamma ray attenuation measurements have very little sensitivity to the matter density. This is relatively clear on the plot that phyzguy posted above, where you see that the measurements by this work alone (the green triangle-shaped contour) permits matter density to go pretty much as low as you like. And as matter density drops, the constraints on the Hubble parameter get looser and looser. Thus this analysis can't really be used to measure matter density.

But if you use other data to constrain the matter density, you also find a relatively narrow measurement of the Hubble parameter, because the matter density is constrained to be near the peak of the triangle. This suggests, at least on the surface, that if the matter density had been lower, this technique would be less able to measure the Hubble parameter.

 

[kimbyd]

Summary: New measurement of the Hubble constant is consistent with the CMB value.

This paper just came out with a new measurement of the Hubble constant based on the technique of gamma ray attenuation. The result is consistent with the lower (CMB-based) value. Interestingly, they also do a joint analysis of several non-CMB techniques (BAO+BBN+SN+γ-ray attenuation), and find a value completely consistent with the CMB value. Clearly, the jury is still out on this issue, and recent pop-science articles about a "crisis in cosmology" are a bit premature.

Very interesting! Definitely leans towards the CMB estimate being more likely to be correct. I'll be curious to see further developments in this space.

 

[Arman777]

What this says, essentially, is that the gamma ray attenuation measurements have very little sensitivity to the matter density. This is relatively clear on the plot that phyzguy posted above, where you see that the measurements by this work alone (the green triangle-shaped contour) permits matter density to go pretty much as low as you like. And as matter density drops, the constraints on the Hubble parameter get looser and looser. Thus this analysis can't really be used to measure matter density.

But if you use other data to constrain the matter density, you also find a relatively narrow measurement of the Hubble parameter, because the matter density is constrained to be near the peak of the triangle. This suggests, at least on the surface, that if the matter density had been lower, this technique would be less able to measure the Hubble parameter.

So this this technique is not good for measuring matter density since it fluctuates but its good for $H_0$...

 

[kimbyd]

So this this technique is not good for measuring matter density since it fluctuates but its good for $H_0$...

I wouldn't say it's good for $H_0$. It's not great for $H_0$, but more important than that is it's another independent way of measuring the parameter. The error bars really aren't anything to write home about, but the independence of the technique makes it valuable. Presumably further effort might reduce the error bars further, making it more useful in the future.

 

[Arman777]

I wouldn't say it's good for $H_0$. It's not great for $H_0$, but more important than that is it's another independent way of measuring the parameter. The error bars really aren't anything to write home about, but the independence of the technique makes it valuable. Presumably further effort might reduce the error bars further, making it more useful in the future.

I find another recent article that is model- independent
https://arxiv.org/pdf/1908.04967.pdf
Error bars a bit high but seems good enough I guess. There are so many contradicting independent measurements. Its really crazy. Every independent way is great indeed but there are alwats two side of the coin.

 

[kimbyd]

I find another recent article that is model- independent
https://arxiv.org/pdf/1908.04967.pdf
Error bars a bit high but seems good enough I guess. There are so many contradicting independent measurements. Its really crazy. Every independent way is great indeed but there are alwats two side of the coin.

Eventually they'll figure out what's causing the discrepancies. 99% of the time it's some error in the design of one of the observations.

 

[Wrichik Basu]

I want to ask something off-topic. Is there a some site or some journal that you are following so that you can learn about these articles or its just from random searches/news ?

I have subscribed to alerts from APS and Springer. They periodically send me emails with the list of latest publications.