Part of the Hubble discrepancy can be explained by local underdensity

[phyzguy]

TL;DR Summary

The fact that we live in a local underdensity of matter (just by chance) can explain most of the discrepancy in the Hubble constant.

Interesting paper on the arXiv today. The authors claim that there is observational evidence that we live in a region with slightly lower density than the universe average, just by chance. Taking this into account can explain as much as 5.5% of the discrepancy in the Hubble constant between the local measurements and the CMB measurement. If you shift the local measurements to lower values by 5.5% (about 3.8 km/sec/Mpc), then the discrepancy is within the experimental errors.

 

[Mordred]

Interesting article thanks for sharing, the article has an interesting argument.

 

[mfb]

Shouldn't this be part of the uncertainties already?

If you shift the local measurements to lower values by 5.5% (about 3.8 km/sec/Mpc), then the discrepancy is within the experimental errors.

The 5.5% are only for quite local measurements, not for the more distant supernova measurements.

 

[kimbyd]

Shouldn't this be part of the uncertainties already?

In an ideal world, sure. In practice it's difficult to account for this type of systematic error.

The basic way to understand this is that if we are in, say, a 2-$\sigma$ underdensity, then it stands to reason that estimates which assume our local density is drawn from a random ensemble of possible local densities will be biased. If the underdensity was less severe, say 1-$\sigma$, then the random sampling of local densities should average to provide the correct result, since 1-$\sigma$ is the expected deviation in such an ensemble.

If the magnitude of the local density can be determined and definitively shown to be below the average in the observable universe, then it would shift a lot of cosmological parameter measurements quite a bit. There'd probably be less dark energy, for instance. Such an underdensity might cause estimates of spatial curvature to be off. And yes, the Hubble constant would be different.

 

[Arman777]

In the conferance videos that mfb shared ( about the Hubble discrepancy), Adam Riess stated that we do not live in a "giant" local void.

https://arxiv.org/abs/1810.03526

İf you want to listen start from the 19:00

http://online.kitp.ucsb.edu/online/enervac-c19/riess/

 

[atyy]

Summary: The fact that we live in a local underdensity of matter (just by chance) can explain most of the discrepancy in the Hubble constant.

Interesting paper on the arXiv today. The authors claim that there is observational evidence that we live in a region with slightly lower density than the universe average, just by chance. Taking this into account can explain as much as 5.5% of the discrepancy in the Hubble constant between the local measurements and the CMB measurement. If you shift the local measurements to lower values by 5.5% (about 3.8 km/sec/Mpc), then the discrepancy is within the experimental errors.

I don't think 5.5% is enough to be consistent with experimental errors. The authors say "This is a differ-ence of about 9.6%, much larger than the combined error. The local estimates for the Hubble constant have been obtained well within a distance less than 100 Mpc, inside the region for which we found a significant underdensity in the matter distribution.Therefore our finding can at least explain part of the difference. But the discrepancy is larger than what could plausibly be accommodated by our observations. Thus one has to look in addition for other reasons for this discrepancy. There could wellbe further systematic effects which may have been overlooked or have been underestimated so far. "

 

[phyzguy]

I don't think 5.5% is enough to be consistent with experimental errors. The authors say "This is a differ-ence of about 9.6%, much larger than the combined error. The local estimates for the Hubble constant have been obtained well within a distance less than 100 Mpc, inside the region for which we found a significant underdensity in the matter distribution.Therefore our finding can at least explain part of the difference. But the discrepancy is larger than what could plausibly be accommodated by our observations. Thus one has to look in addition for other reasons for this discrepancy. There could wellbe further systematic effects which may have been overlooked or have been underestimated so far. "

What I meant was that if you remove the 5.5%, then the remaining discrepancy could just be statistical errors. The 9.6% discrepancy is claimed to be 4.4σ, if you believe all systematic errors have been accounted for ( for the record, I don't believe this). If 5.5% is due to the local underdensity, then the remaining 4.1% would just be just 1.9σ, which could just be statistical errors.