“The heart can get really cold if all you’ve known is winter.” -Benjamin Alire Sáenz
If you didn’t read everything this week on Starts With A Bang, it’s hard to blame you. Each week, on practically a daily basis, we’ve got new pieces coming at you, sharing the wonders of the Universe in our own unique fashion. The past seven days alone saw the following:
- Hubble vs. the Big Bang (for Ask Ethan),
- The horrible conditions of s’more factory farms (for our Weekend Diversion),
- NASA’s greatest observatories view the galactic center (for Mostly Mute Monday),
- What happens to light when it hits the Sun? (a wonderful Astroquizzical from Jillian Scudder),
- Bring deep space home (a review of Govert Schilling’s latest, great book),
- What does “scientific consensus” mean, and
- The physics of Happy Gilmore (for Throwback Thursday).
In addition to all that, I had a couple of new pieces appear over at Forbes:
- How We Know How Many Galaxies Are In The Universe, Thanks To Hubble, and
- Dark Matter Is Necessary For The Origin Of Life.
There were some excellent conversations that took place, but one comment was the most important and informative of all. So for the first time, we’ll have a one-comment-dominated edition (with the rest following the big one) of our Comments of the Week!
From Seshadri Nadathur on the topic of the “cold spot” in the CMB: “I quite like your blog, and I appreciate the lengths you go to to ensure that the science you present is correct. This is why I am disappointed to see that you are in this instance giving credence to ridiculous hype.”
I’m going to separate this comment up into seven major chunks, so that we give each one of the major points the attention it deserves. To start, in the original article I wrote, I talked about three things that happen in the Universe that make what we see as the Cosmic Microwave Background (CMB) and its fluctuations today different from what the initial fluctuations that were left over after inflation ended:
First, the density fluctuations grow-and-shrink as matter and radiation interact after the Hot Big Bang takes place. This leads to an initially scale-invariant spectrum of density fluctuations (solid lines, above) becoming a “wiggly” spectrum of CMB imperfections.
Second, there is material — both hot and fast-moving — in between the CMB and our eyes, for the CMB photons to interact with. When the photons pass through this material, their energy spectrum shifts, due to the Sunyaev-Zel’dovich (SZ) effect. This will result in spots appearing to be the wrong temperature, particularly on quite small scales, unless we account for it.
And third, when structure forms in the Universe, and CMB photons enter that gravitational well, oftentimes the well that they’re leaving is either deeper (for clusters) or shallower (for voids) than the well they entered, making the temperature warmer where there’s more structure and colder where there’s less. This is the Integrated Sachs-Wolfe (ISW) effect.
This is all correct, and no one disputes this. The question is, can this picture — scale-invariant spectrum of fluctuations from inflation, matter-radiation interactions, the SZ effect and the ISW effect — account for everything we see?
Let’s go.
Let me summarise some facts regarding this supervoid “explanation” of the Cold Spot:
1. In November last year, a detailed study of the possible ISW effect due to exactly such a supervoid was published (by me and others) in Physical Review D, explicitly showing that this supervoid fails to account for more than 10% of the total temperature discrepancy at the centre of the Cold Spot unless all of GR is wrong. This reference is here: http://ift.tt/1IyYATT [Link fixed.]
I agree with what’s stated in the paper almost certainly. The void is just under 300 Megaparsecs (or nearly a billion light years) in size, which is huge! The authors (including Seshadri, who commented here) claim that this is a 10% less-dense region of space, although judging by galaxy counts, I find that it’s a 20% less-dense-than-average region of space. That’s a small detail, however. They (correctly) claim that a void of this size and density is nothing special; one that was the same size and 20% less dense than normal would also be expected in the Universe, albeit there should only (by my calculations) be about two-to-three of them in the local Universe, not twenty.
But they are correct with the possible exception of the numbers: they claim entire fluctuation is an extra “coldness” of about ~150 microKelvin, and the ISW for a supervoid of 10% lower density on that scale only contributed about ~20 microKelvin of that. (If it turns out to be 20% lower density, that’s a contribution more than ~20 microKelvin, but somewhat less than ~40 microKelvin, according to their paper.) Regardless, without the supervoid effect, this is approximately a five-sigma fluctuation: a rarity where we’d only expect one in the Universe at most, but not an impossibility.
2. We also showed that in order to explain all of temperature discrepancy at the centre of the Cold Spot one would require a void that is so large and so empty that the chance of it existing in our Universe would be less than 1 in 1 million. The comparable probability of the Cold Spot just being a random fluctuation that requires no special explanation is roughly 1 in 1000, even according to the most pessimistic estimates (and likely much larger).
This part is important, because there was a paper by Istvan Szapudi and his collaborators that claimed that 100% of this effect could be caused not by the ISW, but by a second-order term known as the Rees-Sciama effect. After looking at the Nadathur et al. paper and Szapudi’s paper, I am well convinced that the second order effect is much, much smaller and negligible, and that the Szapudi analysis is incorrect. This means that the cold spot is actually there, and actually quite cold. It’s not catastrophically cold for the scale-invariant fluctuations, just unlikely. What’s ridiculously unlikely, however, is that a supervoid accounts for 100% of this cold spot.
The temperature contours for a given region size, a given density contrast and the number of regions expected per-local-Universe is given below, by Nadathur’s figure 5.
3. The Cold Spot is not unusually cold at the centre. If one selects the coldest spot in the CMB map, naturally it is cold compared to the average point in the same map. This is just a selection effect. The question is whether it is cold compared to the coldest spot in CMB maps in alternative versions of the Universe. The answer is it is not: 100% of the coldness of the Cold Spot at the centre can be explained by the selection effect (Figure 6 in the paper above), compared to just 10% from the supervoid.
When they say “selection effect”, by the way, they mean — and here’s what I think they mean — the coldest spot in any Universe (mock Universe or real Universe) is always going to be much colder than average. Well, so what? Unless it’s so cold that it would be like a one-in-a-huge-huge-HUGE-number chance of such a spot existing, then that’s just the Universe we’re given. In our Universe, according to the Nadathur paper, that happens to be on a relatively large (~15 degree) scale, and happens to be a ~150 microKelvin fluctuation, possibly some (but not a lot) of which is due to this supervoid’s presence.
What they propose looking at as an “interesting” thing, however, is what the temperature from the nearby regions around it are doing, and how they differ from the “coldness” of the region we’re examining. That’s what figure 6, below, shows.
4. What is unusual about the Cold Spot is that there is a hot ring surrounding the central cold region, which does not happen in random CMB maps. The supervoid in no way accounts for this, because supervoids do not produce hot rings, unless all of GR is wrong.
This is unusual, I agree, but not important. The ring is… well, it’s not really a ring. I mean, look for yourself at the Planck data: do you see a ring?
No; you see a region where it’s hotter around a region that’s very, very cold. Although this is unusual in the sense that it’s rare, I don’t think it’s significant in that I don’t believe it means anything other than, “yup, this is what the temperature patterns around this spot look like.”
5. This “supervoid” is not particularly super. It’s rare, but not so very rare: theory predicts roughly 20 or so such voids should exist in our local Universe, and simulations agree. Several such voids have already been seen in other parts of the sky, some of which are larger than this “supervoid” and all of which are emptier. Yet we do not see 20 Cold Spots! Can it be that GR fails only in one region of the sky? Or is it more likely that the supervoid and the Cold Spot are unrelated?
Basically, the point that Seshadri is hammering on is this: the authors of the study that claimed this cold spot was entirely due to the supervoid were in gross error; the supervoid is not really super in any sense, and only contributes at the 10-to-30% (at most) level. Most of the “coldness” of this cold spot is due to the actual density fluctuations in the CMB itself.
But I think the thing the commenter says next is the most damning and accurate thing about the entire episode, and this is really where I blame myself the most.
To be honest the most disappointing thing about this whole episode is that these results were published months in advance of the current press release, and the authors of this paper and the press release are well aware of them. I’m sorry to say that in my opinion the fact that they went ahead with demonstrably incorrect and over-hyped claims reflects very poorly on them. Publicity in news media who know nothing about physics is one thing, but it is much more worrying that such pseudo-scientific claims receive prominence on your own well-informed blog too.
The whole science-communication-by-press-release business is one of the things I’m most strident in my opposition to. They’re often sensationalized, and I often take the short-cut of my own professional assessment of the author(s) in figuring out how closely to scrutinize a claim. I almost went to grad school at the University of Hawaii, and — if I had, based on my interests — Istvan Szapudi likely would have been my Ph.D. supervisor. So I’ve always thought well of him professionally (in addition to his works that I knew about), and that bias that I had prevented me from digging in more deeply and finding the (rather large) flaws in the study.
So while I do think I don’t have anything to apologize for when it comes to the general explanation of the CMB fluctuation, the SZ and the ISW effects or any of that, I did not do as good a job as I ought to have when it came to the scrutiny of the results as respects the cold spot/supervoid issue. The cold spot really is very cold, and almost all of the coldness has nothing to do with the supervoid or the ISW (or the Rees-Sciama effect, for that matter) at all.
And finally, some (brief) other comments.
From magnocrat on the multiwavelength Milky Way center: “It is mindboggling in the extreme but we must remember these are not true visual pictures. We could never , however close see things that way.”
Your eyes will never be this awesome, not unless you had panchromatic EM receptors hooked up to your eyes. But that’s kind of the point about coloration in astronomy and data visualization of all types: if you want to see what’s there, you need to present it in a way that maximizes both the information presented and the ease of processing it. When you look at this image, the data — from the X-rays (hot gas) to stars (points) to gas and dust (IR) — is easy to take in, simply through your eyes.
It’s not a “true visual” picture; it’s much better.
From Sean T on photons traveling through the Universe: “Actually, what Denier says is true, but is only true because space is expanding. The photon’s wavelength increases because of this spatial expansion. In a static universe, light does not lose energy as it propagates through the vacuum.”
And this is important when we consider light traveling to us from locations — like globular clusters, stars within the Milky Way or, honestly, anything in our local group — that aren’t caught up in the expansion of the Universe. Since the space between ourselves and these locations isn’t expanding, there’s no stretching of the wavelength of photons, and hence, no loss of energy. The only redshift/blueshift effects are from doppler (motion) shifts and changes in the gravitational properties of space from the emitter to the observer.
From David Helson on “scientific consensus” and expertise: “What sort of degree do you need to understand what it means when ‘scientists’ are caught rigging the peer review process? What does it suggest when a ‘scientist’ refuses to share his data and refuses to obey legal FOI requests? What sort of degree does it take to recognize that the mathematical weather models have all been wrong for several decades? Is it really uneducated to recognize which side of the grant process is unfairly weighted? Does human nature count for nothing? How much power is in it for the UN, and how much profit? What kind of scientist isn’t disturbed by a rigged peer review process and what kind of scientist leaves that out of his discussion?”
This is what we call “a slew of conspiracy theories.” If you have a problem with the science, then go and learn the science and expose it. The problem with your idea of what’s going on is that everyone who’s gone and done it — even extreme skeptics like Richard Mueller — wind up reaching the same conclusion, independently, as the scientific consensus you decry.
If you’re not going to do that work yourself, you’re just some guy on the internet with an inadequately informed opinion. Honestly, that’s most of who weighs in on climate change on the internet.
Do better.
And finally, from John H on the physics of Happy Gilmore: “This is definitely fun to try, but Padraig who, in addition to his incredible golf skills, also trained as a dancer when younger, makes it look a lot easier than it is. Even long drive competitors don’t use the technique, because it’s near impossible to master consistently and they still have to keep the ball in a target grid used to determine legal attempts. For regular play it just doesn’t add anything useful. Closer is always better than further.”
Here’s the thing: I imagine it’s all about practice. Golfers practice their swing… what, tens of thousands, hundreds of thousands of times, before they’re ready to compete professionally at the highest levels? What if, instead of stationary drives (iron and putter play is different), golfers simply learned how to hit tee shots with a running start? Could that just up the risk/reward game? I’d like to see. Hell, if I didn’t find golf so dull, maybe — since I have practically no experience being a golf-ball-whacker-guy — I’d do the experiment myself! Maybe someday… maybe someday.
Thanks for a great week of comments, and see you back here soon for more about the Universe!
from ScienceBlogs http://ift.tt/1JgvJ55
“The heart can get really cold if all you’ve known is winter.” -Benjamin Alire Sáenz
If you didn’t read everything this week on Starts With A Bang, it’s hard to blame you. Each week, on practically a daily basis, we’ve got new pieces coming at you, sharing the wonders of the Universe in our own unique fashion. The past seven days alone saw the following:
- Hubble vs. the Big Bang (for Ask Ethan),
- The horrible conditions of s’more factory farms (for our Weekend Diversion),
- NASA’s greatest observatories view the galactic center (for Mostly Mute Monday),
- What happens to light when it hits the Sun? (a wonderful Astroquizzical from Jillian Scudder),
- Bring deep space home (a review of Govert Schilling’s latest, great book),
- What does “scientific consensus” mean, and
- The physics of Happy Gilmore (for Throwback Thursday).
In addition to all that, I had a couple of new pieces appear over at Forbes:
- How We Know How Many Galaxies Are In The Universe, Thanks To Hubble, and
- Dark Matter Is Necessary For The Origin Of Life.
There were some excellent conversations that took place, but one comment was the most important and informative of all. So for the first time, we’ll have a one-comment-dominated edition (with the rest following the big one) of our Comments of the Week!
From Seshadri Nadathur on the topic of the “cold spot” in the CMB: “I quite like your blog, and I appreciate the lengths you go to to ensure that the science you present is correct. This is why I am disappointed to see that you are in this instance giving credence to ridiculous hype.”
I’m going to separate this comment up into seven major chunks, so that we give each one of the major points the attention it deserves. To start, in the original article I wrote, I talked about three things that happen in the Universe that make what we see as the Cosmic Microwave Background (CMB) and its fluctuations today different from what the initial fluctuations that were left over after inflation ended:
First, the density fluctuations grow-and-shrink as matter and radiation interact after the Hot Big Bang takes place. This leads to an initially scale-invariant spectrum of density fluctuations (solid lines, above) becoming a “wiggly” spectrum of CMB imperfections.
Second, there is material — both hot and fast-moving — in between the CMB and our eyes, for the CMB photons to interact with. When the photons pass through this material, their energy spectrum shifts, due to the Sunyaev-Zel’dovich (SZ) effect. This will result in spots appearing to be the wrong temperature, particularly on quite small scales, unless we account for it.
And third, when structure forms in the Universe, and CMB photons enter that gravitational well, oftentimes the well that they’re leaving is either deeper (for clusters) or shallower (for voids) than the well they entered, making the temperature warmer where there’s more structure and colder where there’s less. This is the Integrated Sachs-Wolfe (ISW) effect.
This is all correct, and no one disputes this. The question is, can this picture — scale-invariant spectrum of fluctuations from inflation, matter-radiation interactions, the SZ effect and the ISW effect — account for everything we see?
Let’s go.
Let me summarise some facts regarding this supervoid “explanation” of the Cold Spot:
1. In November last year, a detailed study of the possible ISW effect due to exactly such a supervoid was published (by me and others) in Physical Review D, explicitly showing that this supervoid fails to account for more than 10% of the total temperature discrepancy at the centre of the Cold Spot unless all of GR is wrong. This reference is here: http://ift.tt/1IyYATT [Link fixed.]
I agree with what’s stated in the paper almost certainly. The void is just under 300 Megaparsecs (or nearly a billion light years) in size, which is huge! The authors (including Seshadri, who commented here) claim that this is a 10% less-dense region of space, although judging by galaxy counts, I find that it’s a 20% less-dense-than-average region of space. That’s a small detail, however. They (correctly) claim that a void of this size and density is nothing special; one that was the same size and 20% less dense than normal would also be expected in the Universe, albeit there should only (by my calculations) be about two-to-three of them in the local Universe, not twenty.
But they are correct with the possible exception of the numbers: they claim entire fluctuation is an extra “coldness” of about ~150 microKelvin, and the ISW for a supervoid of 10% lower density on that scale only contributed about ~20 microKelvin of that. (If it turns out to be 20% lower density, that’s a contribution more than ~20 microKelvin, but somewhat less than ~40 microKelvin, according to their paper.) Regardless, without the supervoid effect, this is approximately a five-sigma fluctuation: a rarity where we’d only expect one in the Universe at most, but not an impossibility.
2. We also showed that in order to explain all of temperature discrepancy at the centre of the Cold Spot one would require a void that is so large and so empty that the chance of it existing in our Universe would be less than 1 in 1 million. The comparable probability of the Cold Spot just being a random fluctuation that requires no special explanation is roughly 1 in 1000, even according to the most pessimistic estimates (and likely much larger).
This part is important, because there was a paper by Istvan Szapudi and his collaborators that claimed that 100% of this effect could be caused not by the ISW, but by a second-order term known as the Rees-Sciama effect. After looking at the Nadathur et al. paper and Szapudi’s paper, I am well convinced that the second order effect is much, much smaller and negligible, and that the Szapudi analysis is incorrect. This means that the cold spot is actually there, and actually quite cold. It’s not catastrophically cold for the scale-invariant fluctuations, just unlikely. What’s ridiculously unlikely, however, is that a supervoid accounts for 100% of this cold spot.
The temperature contours for a given region size, a given density contrast and the number of regions expected per-local-Universe is given below, by Nadathur’s figure 5.
3. The Cold Spot is not unusually cold at the centre. If one selects the coldest spot in the CMB map, naturally it is cold compared to the average point in the same map. This is just a selection effect. The question is whether it is cold compared to the coldest spot in CMB maps in alternative versions of the Universe. The answer is it is not: 100% of the coldness of the Cold Spot at the centre can be explained by the selection effect (Figure 6 in the paper above), compared to just 10% from the supervoid.
When they say “selection effect”, by the way, they mean — and here’s what I think they mean — the coldest spot in any Universe (mock Universe or real Universe) is always going to be much colder than average. Well, so what? Unless it’s so cold that it would be like a one-in-a-huge-huge-HUGE-number chance of such a spot existing, then that’s just the Universe we’re given. In our Universe, according to the Nadathur paper, that happens to be on a relatively large (~15 degree) scale, and happens to be a ~150 microKelvin fluctuation, possibly some (but not a lot) of which is due to this supervoid’s presence.
What they propose looking at as an “interesting” thing, however, is what the temperature from the nearby regions around it are doing, and how they differ from the “coldness” of the region we’re examining. That’s what figure 6, below, shows.
4. What is unusual about the Cold Spot is that there is a hot ring surrounding the central cold region, which does not happen in random CMB maps. The supervoid in no way accounts for this, because supervoids do not produce hot rings, unless all of GR is wrong.
This is unusual, I agree, but not important. The ring is… well, it’s not really a ring. I mean, look for yourself at the Planck data: do you see a ring?
No; you see a region where it’s hotter around a region that’s very, very cold. Although this is unusual in the sense that it’s rare, I don’t think it’s significant in that I don’t believe it means anything other than, “yup, this is what the temperature patterns around this spot look like.”
5. This “supervoid” is not particularly super. It’s rare, but not so very rare: theory predicts roughly 20 or so such voids should exist in our local Universe, and simulations agree. Several such voids have already been seen in other parts of the sky, some of which are larger than this “supervoid” and all of which are emptier. Yet we do not see 20 Cold Spots! Can it be that GR fails only in one region of the sky? Or is it more likely that the supervoid and the Cold Spot are unrelated?
Basically, the point that Seshadri is hammering on is this: the authors of the study that claimed this cold spot was entirely due to the supervoid were in gross error; the supervoid is not really super in any sense, and only contributes at the 10-to-30% (at most) level. Most of the “coldness” of this cold spot is due to the actual density fluctuations in the CMB itself.
But I think the thing the commenter says next is the most damning and accurate thing about the entire episode, and this is really where I blame myself the most.
To be honest the most disappointing thing about this whole episode is that these results were published months in advance of the current press release, and the authors of this paper and the press release are well aware of them. I’m sorry to say that in my opinion the fact that they went ahead with demonstrably incorrect and over-hyped claims reflects very poorly on them. Publicity in news media who know nothing about physics is one thing, but it is much more worrying that such pseudo-scientific claims receive prominence on your own well-informed blog too.
The whole science-communication-by-press-release business is one of the things I’m most strident in my opposition to. They’re often sensationalized, and I often take the short-cut of my own professional assessment of the author(s) in figuring out how closely to scrutinize a claim. I almost went to grad school at the University of Hawaii, and — if I had, based on my interests — Istvan Szapudi likely would have been my Ph.D. supervisor. So I’ve always thought well of him professionally (in addition to his works that I knew about), and that bias that I had prevented me from digging in more deeply and finding the (rather large) flaws in the study.
So while I do think I don’t have anything to apologize for when it comes to the general explanation of the CMB fluctuation, the SZ and the ISW effects or any of that, I did not do as good a job as I ought to have when it came to the scrutiny of the results as respects the cold spot/supervoid issue. The cold spot really is very cold, and almost all of the coldness has nothing to do with the supervoid or the ISW (or the Rees-Sciama effect, for that matter) at all.
And finally, some (brief) other comments.
From magnocrat on the multiwavelength Milky Way center: “It is mindboggling in the extreme but we must remember these are not true visual pictures. We could never , however close see things that way.”
Your eyes will never be this awesome, not unless you had panchromatic EM receptors hooked up to your eyes. But that’s kind of the point about coloration in astronomy and data visualization of all types: if you want to see what’s there, you need to present it in a way that maximizes both the information presented and the ease of processing it. When you look at this image, the data — from the X-rays (hot gas) to stars (points) to gas and dust (IR) — is easy to take in, simply through your eyes.
It’s not a “true visual” picture; it’s much better.
From Sean T on photons traveling through the Universe: “Actually, what Denier says is true, but is only true because space is expanding. The photon’s wavelength increases because of this spatial expansion. In a static universe, light does not lose energy as it propagates through the vacuum.”
And this is important when we consider light traveling to us from locations — like globular clusters, stars within the Milky Way or, honestly, anything in our local group — that aren’t caught up in the expansion of the Universe. Since the space between ourselves and these locations isn’t expanding, there’s no stretching of the wavelength of photons, and hence, no loss of energy. The only redshift/blueshift effects are from doppler (motion) shifts and changes in the gravitational properties of space from the emitter to the observer.
From David Helson on “scientific consensus” and expertise: “What sort of degree do you need to understand what it means when ‘scientists’ are caught rigging the peer review process? What does it suggest when a ‘scientist’ refuses to share his data and refuses to obey legal FOI requests? What sort of degree does it take to recognize that the mathematical weather models have all been wrong for several decades? Is it really uneducated to recognize which side of the grant process is unfairly weighted? Does human nature count for nothing? How much power is in it for the UN, and how much profit? What kind of scientist isn’t disturbed by a rigged peer review process and what kind of scientist leaves that out of his discussion?”
This is what we call “a slew of conspiracy theories.” If you have a problem with the science, then go and learn the science and expose it. The problem with your idea of what’s going on is that everyone who’s gone and done it — even extreme skeptics like Richard Mueller — wind up reaching the same conclusion, independently, as the scientific consensus you decry.
If you’re not going to do that work yourself, you’re just some guy on the internet with an inadequately informed opinion. Honestly, that’s most of who weighs in on climate change on the internet.
Do better.
And finally, from John H on the physics of Happy Gilmore: “This is definitely fun to try, but Padraig who, in addition to his incredible golf skills, also trained as a dancer when younger, makes it look a lot easier than it is. Even long drive competitors don’t use the technique, because it’s near impossible to master consistently and they still have to keep the ball in a target grid used to determine legal attempts. For regular play it just doesn’t add anything useful. Closer is always better than further.”
Here’s the thing: I imagine it’s all about practice. Golfers practice their swing… what, tens of thousands, hundreds of thousands of times, before they’re ready to compete professionally at the highest levels? What if, instead of stationary drives (iron and putter play is different), golfers simply learned how to hit tee shots with a running start? Could that just up the risk/reward game? I’d like to see. Hell, if I didn’t find golf so dull, maybe — since I have practically no experience being a golf-ball-whacker-guy — I’d do the experiment myself! Maybe someday… maybe someday.
Thanks for a great week of comments, and see you back here soon for more about the Universe!
from ScienceBlogs http://ift.tt/1JgvJ55
Aucun commentaire:
Enregistrer un commentaire