Scientists make further step towards understanding dark energy | Hacker News Reader

Scientists make further step towards understanding dark energy

phys.org

56 pointsdnetesn32 comments

It looks like the major contribution of this paper is in a new signal processing technique. They have something they call the 'chained power spectrum' which gets used to boost a 4 sigma signal into an 11 sigma one.

I don't have all morning to dig into their math but the results are going to live and die on their modeling assumptions. (However I do note that in 4.2 they admit that the choice of "PPP" (Eq 22, section 3.5, all references to the arxiv version) fails in simulated data and of the alternatives that don't fail that particular test, they went with "PQP" instead of the more conservative "QQP" for reasons that I have trouble as interpreting as 'it made the results flashier'. But I'm just some weirdo on the internet, and it's not my field, and I've got to get back to work).

Youtube visualization of the additional new data the study contributed (eBOSS):

https://www.youtube.com/watch?v=UTlYUxucEZA

The linked phys.org article btw didn't say very much. Would be nice with a summary of the actual scientific impact.

As far as I could tell, for models of cosmological growth, it tightened the limits on dark energy to match the standard GR type even more but no new odd things appeared.

I read these findings rather differently. The "expanding balloon" spacetime predicted from GR may absolutely be in question. It appears every type of measurement confirms a discrepancy in the Hubble Constant with respect to distance. Though none of the results from different spectra agree with each other. As new observations from next gen wide array sources such as the Vera Rubin Observatory and James Webb Space Telescope come online, the risk is that the overall picture becomes less clear with higher accuracy ;)

New Distance Measurements Bolster Challenge to Basic Model of Universe

https://public.nrao.edu/news/challenge-model-of-universe/

so cool

"The eBOSS observations are consistent with the dynamical dark energy probed by our team using the BOSS survey four years ago."

What they are referring to here is a model which is an alternative to having a constant cosmological constant. The takeaway is that they are very confident that dark energy is something whose density can vary in time and space. There has been mounting evidence for this for some time, but it's a departure from the original formulation of general relativity. Here is some more information (2012) https://phys.org/news/2012-11-dark-energy-static-dynamic.htm...

Arxiv link (presumably not final version): https://arxiv.org/abs/2007.09011

This seems to be saying they measured Dark Energy acceleration in galaxies only a billion years old and light years away, not 13 billion light years old and correspondingly distant.

Correct, the data is regarding an age/area of the universe that has not been probed prior. Also, light years are a measure of distance, not time.

It's a measure of both, since you're looking back in time.

In this specific case perhaps but not in general - consider for example a circle with a radius of one light year centered on the Earth.

In the context of astronomy generally then ;)

Not sure what you mean - distance directly away from an observer correlates with time (things further away look younger due to the finite speed of light), but in general it's just a measure of distance. It's right there in the units - light (speed) years are m/s * s; that is, metres. I'm not aware that astronomers use ly as a unit of time, unless that's changed since I studied it for my undergrad, although that was admittedly some time ago.

While astronomers are obviously very much aware that light-years is a measure of distance, "seeing further away" is a stand-in for "seeing further back in time." New telescopes are rated by how much further away they can resolve in terms of light-years because that is a direct translation into how close to the Big Bang they can resolve.

What I'm saying is that it is extremely useful that being able to resolve objects 13.2 billion light-years away means being able to see back 13.2 billion years in time. The units are 1:1 convertible, and astronomers therefore frequently interchange them in casual conversation.

I don't think it is so simple. As a rule of thumb, sure.

But space is supposed to have been expanding while the light was in transit, so 13 Gy-old light has travelled way more than 13G light-years, and the object that emitted it is "now" dizzyingly farther away even than that. (Scare quotes, because simultaneity is meaningless at such a distance; and it must be outside our light-cone, so can't really meaningfully be said to exist in our universe anymore.)

I am not clear at all on how light experiences expansion of space, as that seems to require time, and light travelling in vacuum doesn't experience time.

Light emitted only a billion years ago has traveled only a little more than a billion light years. But what's a few million light-years, extra, among friends?

Well you're touching on some deep and difficult stuff here. There are many others who understand and can explain all this better than I, but I'll do my best to comment usefully, however before I do - none of what you said here has any bearing on the original question of whether 'light years' are units of time. They're not. Anyway, all this stuff is fascinating, so here goes:

> But space is supposed to have been expanding while the light was in transit, so 13 Gy-old light has travelled way more than 13G light-years, and the object that emitted it is "now" dizzyingly farther away even than that. (Scare quotes, because simultaneity is meaningless at such a distance; and it must be outside our light-cone, so can't really meaningfully be said to exist in our universe anymore.)

Close but not exactly - the point where the light was emitted is now 13G light years away, because by definition, a light year is the distance light travels in a year. The object emitting that light, however, was accelerating away from us and "now", that is, following the expansion that occurred during the time the light was in transit, is more like 42G light years away. When the light was emitted, the object in question was much closer than that. The expansion of space has changed the definition of a metre and therefore effectively moved us and the object apart. A (crappy) visualisation of an object A emitting a photon P at point X towards an observer B (first, later, and now are of course in implied scare quotes):

first: AXP-B

later: A----X-P--B

now: A-----------X-------PB

The distance XB is the distance P has travelled in total, 13G light years in our earlier example.

> I am not clear at all on how light experiences expansion of space, as that seems to require time, and light travelling in vacuum doesn't experience time.

This is a great question and took some research for me to get close to being able to answer, but the best I can do here is to link some related discussions [0][1][2], and summarise that photons don't experience anything because they have no frame of reference, so it's meaningless to ask whether or not or how they experience time, or indeed anything else. Photons always move at C in every frame of reference (remember there is no such thing as an absolute velocity for an observer, it's always relative to a reference frame), so there can be no frame of reference comoving with a photon in which time can be measured (and indeed, in every reference frame, time appears to move at the normal rate - time dilation only occurs between two frames). What we see on Earth is that a photon is "stretched" by the expansion of space and therefore red shifted - and as I guess you know, that's a key marker astronomers use to calculate the distance a photon has travelled - more red shift means more distance. As very distant objects are accelerated out of our light cone, they are red-shifted so far that they eventually disappear - in fact, far far into the future, the same will be true of everything in the universe, meaning that any being looking up hundreds of billions of years from now will see no galaxies outside their own.

> Light emitted only a billion years ago has traveled only a little more than a billion light years. But what's a few million light-years, extra, among friends?

As explained above, not so. Light travels at one light year per year. Expanding space is changing the meaning of distance but not the speed of light.

[0] https://physics.stackexchange.com/questions/54162/how-does-a...

[1] https://physics.stackexchange.com/questions/29082/would-time...

[2] https://physics.stackexchange.com/questions/332473/if-photon...

Well that diagram was crappier than even I thought; this one is much better:

https://blogs-images.forbes.com/startswithabang/files/2018/0...

(sourced from https://www.forbes.com/sites/startswithabang/2019/02/26/how-...)

Suffice to say, then, light emitted a billion years ago has travelled substantially farther than a billion x 9.46e12 km, according exactly to how much the bit of universe it traversed expanded on the way.

No, the other way around - the distance light would have travelled if the universe had stopped expanding at the time it was emitted would have been a lot less than a billion light years. The light ended up travelling a billion light years because of the expansion.

Try a thought experiment. You're driving at 60mph between two towns on a straight road that's 10 miles long when you set off but is growing in the direction of travel. You arrive an hour later - how far have you travelled?

You're both right and talking past each other. The light moved between two points that if measured at the start was fewer light-years apart than the time it took to cross the gap, or a greater number of light-years if measured now.

I think the most correct and relevant thing to say is that the light moved 1 light-year per year and covered exactly the specified distance, if you integrate the distance over the trajectory of the light as it moved across the universe. Inflation simply changes the geometry of the space before (or after) the light passes through.

> I think the most correct and relevant thing to say is that the light moved 1 light-year per year and covered exactly the specified distance, if you integrate the distance over the trajectory of the light as it moved across the universe.

Yes, so light travels X light years in X years, not substantially more than that which is how I read (or misread) the parent comment. Language is fun I guess.

You contradict yourself. In that billion years, either the light travelled (1) a distance we call "a billion light years (a measure adjusted to account for expansion of space during transit)", or (2) it travelled "more than a billion light years (a measure that neglects expanding space)". In either case, substantially greater than 9.46e24 meters.

If the road grew while I was on it, then it would take longer than an hour to get there. If it took an hour to get there anyway, I went faster than 60 mph. If I went 60 mph for an hour and got there, then the road did not grow.

You can fool with "light years" as a unit all you like ("instantaneously 9.46e12 m, but more as travel time increases"), but you don't get to fool with meters. If no numbers change, expansion is meaningless, the only thing that changes is light wavelengths. Then you are just talking about tired light.

> If the road grew while I was on it, then it would take longer than an hour to get there. If it took an hour to get there anyway, I went faster than 60 mph. If I went 60 mph for an hour and got there, then the road did not grow.

I'll start with this because it's the easy one and I think you've maybe just misread. The road is 10 miles at the outset. Your speed is 60mph. If the road doesn't expand, the journey would take 10 minutes. Instead the journey takes an hour - 60mph for one hour is 60 miles. So in total you've travelled 60 miles. To switch back to our 'light travelling through space' discussion, by analogy - light travels at C; if light travels for 1 billion years, it travels a distance of C * 1 billion == 1 billion light years. When it set out on its journey towards us, the distance between there and here was less. Therefore it has travelled 1 billion light years to reach us only because space was expanding. It hasn't travelled more than that, it can't - the speed of light is a constant in every reference frame. X years at light speed is necessarily X light years.

> You contradict yourself. In that billion years, either the light travelled (1) a distance we call "a billion light years (a measure adjusted to account for expansion of space during transit)", or (2) it travelled "more than a billion light years (a measure that neglects expanding space)". In either case, substantially greater than 9.46e24 meters.

(note that I think you mean 9.46e21 meters, since conventionally, 1 billion is 1e9; I'll run with that figure for consistency).

For the record, option 1) is what I'm saying. However, one billion light years is 9.46e21 meters, so I'm totally stumped as to what you mean when you say "In either case, substantially greater than 9.46e2[1] meters". Those are the same thing. What am I missing? Light travels one billion light years, 9.46e21 metres, in one billion years. That's it. Where am I misunderstanding you or contradicting myself? I'm honestly confused.

Well, OK.

  $ /bin/units
  You have: 1 Gly
  You want: m
   * 9.4607305e+24
   / 1.0570008e-25
  You have: 1 Gly
  You want: km
   * 9.4607305e+21
   / 1.0570008e-22
  $

But I got mixed up on starting the road length, reading it as 60 mi. Fair's fair.

I got mixed up between m and km so I guess we're even :-)

Typo. "...13 billion years old and correspondingly distant."

This might resolve my doubts that the phenomenon could be an artifact of different conditions in the early universe.

11 sigma sounds pretty high to me (non-physicist), is that kind of standard?

That's insanely accurate. For comparison most discoveries are usually only considered proven when it's 5 sigma.

For example they waited years to announce the Higgs bosson because they wanted to reach 5 sigma.

Thanks, that's what I thought. I was doubting it was even the same metric :)

> That's insanely accurate.

This is misleading, and it's a misconception arising from the way that even physicists sometimes misspeak when talking about statistics. The "sigma" is a reference to a standard deviation, and we use it to measure evidence against a null hypothesis (which is generally defined relative to a model of some type). An 11 sigma result means that there's strong evidence against (hypothesis + model), not anything in absolute terms about "accuracy".

The meaning of such a result depends on the model being used; sometimes it's as simple and easy to defend as there being some sort of distribution with a mean which exists, and other times there's a complicated model which may itself be incorrect (producing significant results).

You can trivially get to 11 sigma by making your null hypothesis sufficiently garbage.

That's called "p-hacking" and is unfortunately very common. It is encouraged whenever someone innocently reads meaning into a sigma value without evaluating whether or not the null hypothesis is any good.

If I saw someone going around saying "The sky is reddish green, 11 sigma result!" my first suspicion would not be that their numbers were incorrect but rather that their null hypothesis was something silly like "the sky is entirely the solid sRGB color #0000FF." Obviously this garbage null hypothesis can be rejected with enormous confidence, that's where the 11 sigma came from, but the confident rejection of a garbage null hypothesis should not be taken to support their proposed alternative theory unless you have reason to believe the null hypothesis was well modeled, either because you checked it or because someone else did. It all comes back to trust, in other words.

You could call it a deviant standard.

This is really quite above us, but still so amazing!

Reply on news.ycombinator.com