Worth reading just for this :
The President’s budget reflects a consistent and fundamental vision about American strength that is fundamentally at odds with a vision presented by almost 50 years ago by the physicist Robert Wilson, the first director of the Fermi National Accelerator Laboratory near Chicago at which a large particle accelerator was being built. When testifying before Congress about the machine and its cost, Wilson was asked if it completion would aid in the defense of the nation. His answer is striking.
"No Sir…I don’t believe so…. It has only to do with the respect with which we regard one another, the dignity of men, our love of culture... It has to do with are we good painters, good sculptors, great poets? I mean all the things we really venerate in our country and are patriotic about. It has nothing to do directly with defending our country except to make it worth defending."
And yet...
Whether future historians will view the United States as a truly great nation will not depend upon our military strength or our ability to successfully assimilate immigrants, any more than we celebrate the greatness of ancient Greece or Rome by counting their military victories.
I suppose strictly speaking no, not by counting their military victories, that would be silly. But without military victories there would have been no Greece or Rome in the first place.
https://blogs.scientificamerican.com/guest-blog/killing-science-and-culture-doesnt-make-the-nation-stronger/
Tuesday, 28 March 2017
Monday, 27 March 2017
People want to give us money
First, participants were asked to estimate what percentage of the federal budget was spent on scientific research. Once they’d guessed, half of the participants were told the actual amount that the federal government allocates for nondefense spending on research and development. In 2014, that figure was 1.6 percent of the budget, or about $67 billion. Finally, all the participants were asked if federal spending on science should be increased, decreased or kept the same.
The majority of participants had no idea how much money the government spends on science, and wildly overestimated the actual amount. About half of the respondents estimated federal spending for research at somewhere between 5 and 20 percent of the budget. A quarter of participants estimated that figure was 20 percent of the budget — one very hefty chunk of change. The last 25 percent of respondents estimated that 1 to 2 percent of federal spending went to science.
When participants received no information about how much the United States spent on research, only about 40 percent of them supported more funding. But when they were confronted with the real numbers, support for more funding leapt from 40 to 60 percent.
https://www.sciencenews.org/blog/scicurious/most-americans-science-and-are-willing-pay-it
The majority of participants had no idea how much money the government spends on science, and wildly overestimated the actual amount. About half of the respondents estimated federal spending for research at somewhere between 5 and 20 percent of the budget. A quarter of participants estimated that figure was 20 percent of the budget — one very hefty chunk of change. The last 25 percent of respondents estimated that 1 to 2 percent of federal spending went to science.
When participants received no information about how much the United States spent on research, only about 40 percent of them supported more funding. But when they were confronted with the real numbers, support for more funding leapt from 40 to 60 percent.
https://www.sciencenews.org/blog/scicurious/most-americans-science-and-are-willing-pay-it
Tuesday, 21 March 2017
Ultra Diffuse Galaxies, "Still A Thing", Study Finds
Ultra-diffuse galaxies are, broadly speaking, about the same size as the Milky Way but a thousand times fainter. Although a few such objects have been known since the bygone days of yore (e.g. the 1980s), it's only in the last couple of years that they've been discovered in large numbers. Probably the biggest question is how massive these things are. It's easy enough to find their stellar mass - we can basically measure that directly from their brightness - but much harder to work out how much dark matter they have. For that, we need to measure the motions of their stars and/or gas, which is much harder than just measuring how bright they are.
There have been a handful of previous attempts. Beasley et al. 2016 looked at VCC 1287 and concluded it was of low mass, but still a very strange object because it doesn't fit the normal rotation-brightness relation. van Dokkum et al. 2016 looked at Dragonfly 44 and concluded it was of high mass, and though their estimate of the total mass is an extrapolation it's certainly more massive than VCC 1287.
This latest paper uses the ALFALFA hydrogen survey to search for UDGs. This means the sample is biased towards UDGs that are hydrogen-rich (which might not be the case for the population in general), but the advantage is that the the hydrogen measurements easily allow for estimate of the rotation velocities. Because the survey is very large, this means they have a decent sample of 115 UDGs. ALFALFA doesn't have a high resolution, so they also have VLA observations of 3 UDGs where they can measure the rotation curve accurately rather than just estimating it.
They find that the UDGs appear to be huge dwarfs. That is, they have very low masses but are very extended - they're around the same diameter as the Milky Way, but rotating 4-7x less quickly (30-50 km/s compared to 220 km/s), meaning they don't need nearly as much dark matter to hold themselves together. Many measurements in astronomy are not all that precise and errors of a factor of a few are not all unusual. Rotation velocities are an exception - this is much too large to be dismissed as an error, especially given their large sample size.
It's still too early to say what this means. In principle, the more dwarf galaxies found the better, since the standard model predicts far more than observed. Other papers have indicated that these UDGs are nowhere near numerous enough to explain the difference, but their existence implies that there are even more galaxies out there which are even fainter and harder to detect. But at the moment we only have these mass estimates for these ~100 gas-rich galaxies, whereas most UDGs may not have gas at all. So the conclusions are still suffering a huge bias, and another recent paper suggests that there might be populations of different types of UDGs. Further research is required... as in, years of further research, not, "it'll be done by Tuesday".
I like this paper very much, I think it does a good job of declaring the biases, uncertainties, definitions (importantly noting that there's no hard definition of UDG yet - the IAU's arm is not that long) and methodologies. The one thing I wish they'd done is compare their galaxies to the normal rotation-brightness relation (Tully Fisher) because it looks to me like there might be an interesting offset. There's a sample of the full table but I'm not sure where the entire thing is, perhaps it's not available until after acceptance. That's certainly something I'd sit on, since it easily has the potential to warrant another paper.
https://arxiv.org/abs/1703.05293
There have been a handful of previous attempts. Beasley et al. 2016 looked at VCC 1287 and concluded it was of low mass, but still a very strange object because it doesn't fit the normal rotation-brightness relation. van Dokkum et al. 2016 looked at Dragonfly 44 and concluded it was of high mass, and though their estimate of the total mass is an extrapolation it's certainly more massive than VCC 1287.
This latest paper uses the ALFALFA hydrogen survey to search for UDGs. This means the sample is biased towards UDGs that are hydrogen-rich (which might not be the case for the population in general), but the advantage is that the the hydrogen measurements easily allow for estimate of the rotation velocities. Because the survey is very large, this means they have a decent sample of 115 UDGs. ALFALFA doesn't have a high resolution, so they also have VLA observations of 3 UDGs where they can measure the rotation curve accurately rather than just estimating it.
They find that the UDGs appear to be huge dwarfs. That is, they have very low masses but are very extended - they're around the same diameter as the Milky Way, but rotating 4-7x less quickly (30-50 km/s compared to 220 km/s), meaning they don't need nearly as much dark matter to hold themselves together. Many measurements in astronomy are not all that precise and errors of a factor of a few are not all unusual. Rotation velocities are an exception - this is much too large to be dismissed as an error, especially given their large sample size.
It's still too early to say what this means. In principle, the more dwarf galaxies found the better, since the standard model predicts far more than observed. Other papers have indicated that these UDGs are nowhere near numerous enough to explain the difference, but their existence implies that there are even more galaxies out there which are even fainter and harder to detect. But at the moment we only have these mass estimates for these ~100 gas-rich galaxies, whereas most UDGs may not have gas at all. So the conclusions are still suffering a huge bias, and another recent paper suggests that there might be populations of different types of UDGs. Further research is required... as in, years of further research, not, "it'll be done by Tuesday".
I like this paper very much, I think it does a good job of declaring the biases, uncertainties, definitions (importantly noting that there's no hard definition of UDG yet - the IAU's arm is not that long) and methodologies. The one thing I wish they'd done is compare their galaxies to the normal rotation-brightness relation (Tully Fisher) because it looks to me like there might be an interesting offset. There's a sample of the full table but I'm not sure where the entire thing is, perhaps it's not available until after acceptance. That's certainly something I'd sit on, since it easily has the potential to warrant another paper.
https://arxiv.org/abs/1703.05293
Friday, 17 March 2017
Dark matter in the early Universe
There's this press release going around at the moment claiming that dark matter appears to be less important in galaxies earlier in the Universe than in the present day.
This is rather odd, since there's supposedly much more dark matter than normal matter in the Universe, with its total mass being responsible for forming the large-scale structures (filaments of galaxies and galaxy clusters) within the lifetime of the observable Universe. Without that extra mass it's hard to get these structures to form in the given time. Galaxy formation is often envisaged as being a process of how normal matter gets into dark matter halos and starts forming stars. Since those halos can merge over time this becomes tremendously complicated. So it's not outside the realm of possibility that earlier galaxies would have less dark matter, but it is odd.
Even odder is that the claims of the paper appear, for once, appear to be even stronger than in the press release : "we find that the dark matter fractions near the half-light radius for all our galaxies are modest to negligible." This does not necessarily mean they have no dark matter, just less than normal matter, but that's in complete contrast to nearby galaxies which are dominated by dark matter by a typical factor 5-10.
I'm less than convinced by this.
Below I reproduce the main part of figure 1 from their paper. Column a shows their observations of H-alpha (ionized hydrogen). Panel b shows the velocity of the H-alpha, which can be directly measured. Panel c shows the rotation curve of the galaxy - taking that direct velocity measurement and converting it into rotation. The points show the data, the red curves show their best fit to the data.
Galaxies which are dominated by dark matter have flat rotation curves at large distances. Their claim is that the galaxies in their sample show declining rotation curves, which is what you'd expect if there wasn't any dark matter. They have a larger sample of galaxies but they claim the six in the figure are their best candidates.
None of them look particularly convincing to me. The first one blatantly shows a flat curve, how they managed to fit a decline to it I don't know. The second has a clear asymmetry and looks like it might be interacting, with one side of the curve declining much more steeply than the others. The third is again flat, with the only hint of a decline being a single data point on the right hand side - again, how on earth they fitted such a nice symmetrical, steeply declining curve I don't know. Similarly the fourth and sixth appear to be dominated by individual data points. Such "wiggles" are well-known; to extrapolate like this is very dangerous. See also point 2 here, and note that other galaxies are well-known to show slight declines which then flatten off. Only their fifth galaxy looks at all plausible to me, and even that one hardly looks convincing.
They say :
We find interacting low mass satellites in three of our six sources and evidence for some tidal stripping in one, but the rotation curve is symmetric, even near the satellite.
Wait, what ? No it isn't ! Any fool can that just from looking at the second galaxy.
They also try "stacking" the rotation curves in their larger sample. This is essentially just adding them all up, which can give a much stronger signal (at the expense of losing information about individual objects). This produced a very nice, steeply declining curve. The problem is that if there are systematic errors in this they will only get larger, not weaker. Their model fits rely on some extremely questionable assumptions (neglecting atomic gas completely), the data is clearly messy (column a), and their rotation curve fit relies on a complicated model with many free parameters that seems so blatantly to be doing such a bad job in their best cases... sigh.
As a rule of thumb, if you can see a pattern visually it's worth investigating further to see if you can verify it rigorously. If you see a pattern, it might be real but it might not - your brain loves patterns in order to avoid getting eaten by stripy tigers in dark forests. It's had millions of years of practise over far longer timescales than any mathematical techniques humans have invented to detect trends. So if you can't see a trend in the data, chances are, there isn't one. I can't see a trend in this data, only dodgy fitting and rather extreme claims.
But heck, this is published in Nature, which is supposed to be about important, exciting results. Maybe their full, longer paper will be better, but I haven't read that yet .
https://arxiv.org/abs/1703.05491
This is rather odd, since there's supposedly much more dark matter than normal matter in the Universe, with its total mass being responsible for forming the large-scale structures (filaments of galaxies and galaxy clusters) within the lifetime of the observable Universe. Without that extra mass it's hard to get these structures to form in the given time. Galaxy formation is often envisaged as being a process of how normal matter gets into dark matter halos and starts forming stars. Since those halos can merge over time this becomes tremendously complicated. So it's not outside the realm of possibility that earlier galaxies would have less dark matter, but it is odd.
Even odder is that the claims of the paper appear, for once, appear to be even stronger than in the press release : "we find that the dark matter fractions near the half-light radius for all our galaxies are modest to negligible." This does not necessarily mean they have no dark matter, just less than normal matter, but that's in complete contrast to nearby galaxies which are dominated by dark matter by a typical factor 5-10.
I'm less than convinced by this.
Below I reproduce the main part of figure 1 from their paper. Column a shows their observations of H-alpha (ionized hydrogen). Panel b shows the velocity of the H-alpha, which can be directly measured. Panel c shows the rotation curve of the galaxy - taking that direct velocity measurement and converting it into rotation. The points show the data, the red curves show their best fit to the data.
Galaxies which are dominated by dark matter have flat rotation curves at large distances. Their claim is that the galaxies in their sample show declining rotation curves, which is what you'd expect if there wasn't any dark matter. They have a larger sample of galaxies but they claim the six in the figure are their best candidates.
None of them look particularly convincing to me. The first one blatantly shows a flat curve, how they managed to fit a decline to it I don't know. The second has a clear asymmetry and looks like it might be interacting, with one side of the curve declining much more steeply than the others. The third is again flat, with the only hint of a decline being a single data point on the right hand side - again, how on earth they fitted such a nice symmetrical, steeply declining curve I don't know. Similarly the fourth and sixth appear to be dominated by individual data points. Such "wiggles" are well-known; to extrapolate like this is very dangerous. See also point 2 here, and note that other galaxies are well-known to show slight declines which then flatten off. Only their fifth galaxy looks at all plausible to me, and even that one hardly looks convincing.
They say :
We find interacting low mass satellites in three of our six sources and evidence for some tidal stripping in one, but the rotation curve is symmetric, even near the satellite.
Wait, what ? No it isn't ! Any fool can that just from looking at the second galaxy.
They also try "stacking" the rotation curves in their larger sample. This is essentially just adding them all up, which can give a much stronger signal (at the expense of losing information about individual objects). This produced a very nice, steeply declining curve. The problem is that if there are systematic errors in this they will only get larger, not weaker. Their model fits rely on some extremely questionable assumptions (neglecting atomic gas completely), the data is clearly messy (column a), and their rotation curve fit relies on a complicated model with many free parameters that seems so blatantly to be doing such a bad job in their best cases... sigh.
As a rule of thumb, if you can see a pattern visually it's worth investigating further to see if you can verify it rigorously. If you see a pattern, it might be real but it might not - your brain loves patterns in order to avoid getting eaten by stripy tigers in dark forests. It's had millions of years of practise over far longer timescales than any mathematical techniques humans have invented to detect trends. So if you can't see a trend in the data, chances are, there isn't one. I can't see a trend in this data, only dodgy fitting and rather extreme claims.
But heck, this is published in Nature, which is supposed to be about important, exciting results. Maybe their full, longer paper will be better, but I haven't read that yet .
https://arxiv.org/abs/1703.05491
Thursday, 9 March 2017
A tautologous quote is a tautologous quote
"On the contrary, those UDGs found in the richest environments should be depleted of HI gas due to the removal of this component."
Translation : galaxies without much gas don't have much gas because they don't have much gas. Have you considered joining the tautology club in order to join the tautology club because you want to join the tautology club ?
Translation : galaxies without much gas don't have much gas because they don't have much gas. Have you considered joining the tautology club in order to join the tautology club because you want to join the tautology club ?
Friday, 3 March 2017
All Of Galaxy Formation Theory Is Wrong, Says Scientist
All Of Galaxy Formation Theory Is Wrong, Says Scientist
A legitimate headline - with the caveats of this being an arXiv-only article not submitted to any journal.
You might remember Mike Disney's last arXiv-only article in which he claimed that there were a huge number of optically faint or dark galaxies that had escaped detection in previous searches. That idea was more or less already vindicated by the discovery of huge numbers of extended, faint galaxies in clusters, though the other idea that they were missed in hydrogen surveys due to misidentifications looks to be very wrong.
Mike's latest offering turns that idea bizarrely on its head, assuming now (God knows why) that the very faint galaxies don't actually exist at all. In that case there are a number of strange "scaling relations" that the standard model of galaxy formation (hierarchical merging of smaller galaxies to build up larger ones) just can't explain. Some of these arguments might be interesting - e.g. all galaxies appear to have similar average brightnesses, gas density, and overall mass density (including dark matter). There's no obvious reason why that should be in a merging scenario, but if galaxies all formed by monolithic collapse in the uniformish-density early Universe it would make more sense.
Except that this completely ignores the newly-discovered ultra-faint galaxies, which certainly have much lower brightness densities and many probably have very different dark matter contents compared to normal galaxies of similar size and brightness. And that's one hell of a detail to overlook.
Personally I'm still rather skeptical of hierarchical merging - there are some things we know it can't explain, and monolithic collapse is a much simpler idea. If there was ever a case where I'd bet it'll turn out to be something in the middle, it'd be this.
There are two things I like about this article : 1) Mike throws caution utterly to the wind and considers an entirely different model of galaxy formation, which I think is an interesting speculative exercise to be encouraged more often provided it isn't taken to seriously; 2) The narrative writing style is easily ten times better than most papers. Instead of skimming the abstract and conclusions and figures and all that nonsense you can just start reading from beginning to end. Each section builds part of the narrative and the language is not the sadistically formal normally expected in refereed papers. And there's some very interesting stuff on using Bayesian statistics to assess evidence at the end, but that's a topic in itself (Mike's written but not published a book on that, which I've read - it's excellent).
The conclusions ? Sod 'em, it's a good read and I like the approach.
https://arxiv.org/abs/1703.00363
A legitimate headline - with the caveats of this being an arXiv-only article not submitted to any journal.
You might remember Mike Disney's last arXiv-only article in which he claimed that there were a huge number of optically faint or dark galaxies that had escaped detection in previous searches. That idea was more or less already vindicated by the discovery of huge numbers of extended, faint galaxies in clusters, though the other idea that they were missed in hydrogen surveys due to misidentifications looks to be very wrong.
Mike's latest offering turns that idea bizarrely on its head, assuming now (God knows why) that the very faint galaxies don't actually exist at all. In that case there are a number of strange "scaling relations" that the standard model of galaxy formation (hierarchical merging of smaller galaxies to build up larger ones) just can't explain. Some of these arguments might be interesting - e.g. all galaxies appear to have similar average brightnesses, gas density, and overall mass density (including dark matter). There's no obvious reason why that should be in a merging scenario, but if galaxies all formed by monolithic collapse in the uniformish-density early Universe it would make more sense.
Except that this completely ignores the newly-discovered ultra-faint galaxies, which certainly have much lower brightness densities and many probably have very different dark matter contents compared to normal galaxies of similar size and brightness. And that's one hell of a detail to overlook.
Personally I'm still rather skeptical of hierarchical merging - there are some things we know it can't explain, and monolithic collapse is a much simpler idea. If there was ever a case where I'd bet it'll turn out to be something in the middle, it'd be this.
There are two things I like about this article : 1) Mike throws caution utterly to the wind and considers an entirely different model of galaxy formation, which I think is an interesting speculative exercise to be encouraged more often provided it isn't taken to seriously; 2) The narrative writing style is easily ten times better than most papers. Instead of skimming the abstract and conclusions and figures and all that nonsense you can just start reading from beginning to end. Each section builds part of the narrative and the language is not the sadistically formal normally expected in refereed papers. And there's some very interesting stuff on using Bayesian statistics to assess evidence at the end, but that's a topic in itself (Mike's written but not published a book on that, which I've read - it's excellent).
The conclusions ? Sod 'em, it's a good read and I like the approach.
https://arxiv.org/abs/1703.00363
Making journals better
Here's my naive and simple idea as to how to reform peer review so as to reduce the "publish or perish" culture without stifling ideas, increase reproducibility of studies and generally make the world a nicer place without spending any money. I guess this one isn't particularly timely - it's been sat in the draft folder for quite a while. I've mentioned this before; I really just wanted to have a slightly more fully-developed and permanent record.
With fierce competition for jobs, it's inevitable that employers resort to using very simple methods to assess candidates : largely publication rate, hence the "publish or perish" guideline. Hence the obvious tactic : publish lots of mediocre papers. And a publication is a publication. There's no way to judge by glancing at a C.V. whether that research was really top-grade or just plain mediocre. But what if there was ?
What I'm proposing is that we try and label the papers as a guideline by which employers can quickly assess performance based on something more than sheer number of publications. I say "label" rather than "grade" because this can be a complex non-linear system. It might, for instance, be useful to label papers according to content. Most regular journals already have a main journal plus a "letters" section which publishes much shorter, timely articles. Why not extend this further ? Instead of just MNRAS Letters, also have MNRAS Observational Catalogues, MNRAS Numerical Simulations, MNRAS Serendipitous Discoveries, MNRAS Data Mining, MNRAS Clickbait, MNRAS Essays, MNRAS Breakthroughs, MNRAS Replication Studies, MNRAS Things I Just Thought Up Off The Top Of My Head While I Was On The Toilet, MNRAS Things Some Bloke In A Pub Said Last Tuesday, etc.
Papers could be labelled not only by content but also review rigour - not to be confused with research quality, because that's not the same thing. Indeed this might be necessary under the new system. If more complex papers are to be seen as more valuable, they'll need more careful review. All levels of peer review are going to need some basic guidelines, which will require some thinking about what we want journal-based peer review to actually mean. Currently, reviewers are given a free hand to request whatever changes they like.
For instance, the lowest level of review (for an "essay" paper, maybe) might be a single referee doing a check to make sure there are no internal inconsistencies, known problems with the methodology, factual errors etc. For the highest level, there might be three or more independent reviewers checking everything with a fine-toothed comb, and they'd be expected to check everything.
Replication studies are a time-consuming procedure, with a high potential just to confirm the previous findings and not learn anything new.One way to offset this would be to award replication studies an extra level of prestige : insist that these studies be subject to the highest level of review possible. Getting such a paper accepted would be a real challenge and recognised as such. So there would be a motivational balance of glory on the one hand, difficulty and low likelihood of new discoveries on the other. A successful replication study could also have a transformative effect on the original paper, changing it from a merely interesting result to one that deserves strong attention. That in turn encourages everyone to publish research which can be replicated in the first place.
https://astrorhysy.blogspot.com/2017/03/this-is-not-crisis-youre-looking-for.html https://astrorhysy.blogspot.com/2017/03/this-is-not-crisis-youre-looking-for.html
With fierce competition for jobs, it's inevitable that employers resort to using very simple methods to assess candidates : largely publication rate, hence the "publish or perish" guideline. Hence the obvious tactic : publish lots of mediocre papers. And a publication is a publication. There's no way to judge by glancing at a C.V. whether that research was really top-grade or just plain mediocre. But what if there was ?
What I'm proposing is that we try and label the papers as a guideline by which employers can quickly assess performance based on something more than sheer number of publications. I say "label" rather than "grade" because this can be a complex non-linear system. It might, for instance, be useful to label papers according to content. Most regular journals already have a main journal plus a "letters" section which publishes much shorter, timely articles. Why not extend this further ? Instead of just MNRAS Letters, also have MNRAS Observational Catalogues, MNRAS Numerical Simulations, MNRAS Serendipitous Discoveries, MNRAS Data Mining, MNRAS Clickbait, MNRAS Essays, MNRAS Breakthroughs, MNRAS Replication Studies, MNRAS Things I Just Thought Up Off The Top Of My Head While I Was On The Toilet, MNRAS Things Some Bloke In A Pub Said Last Tuesday, etc.
Papers could be labelled not only by content but also review rigour - not to be confused with research quality, because that's not the same thing. Indeed this might be necessary under the new system. If more complex papers are to be seen as more valuable, they'll need more careful review. All levels of peer review are going to need some basic guidelines, which will require some thinking about what we want journal-based peer review to actually mean. Currently, reviewers are given a free hand to request whatever changes they like.
For instance, the lowest level of review (for an "essay" paper, maybe) might be a single referee doing a check to make sure there are no internal inconsistencies, known problems with the methodology, factual errors etc. For the highest level, there might be three or more independent reviewers checking everything with a fine-toothed comb, and they'd be expected to check everything.
Replication studies are a time-consuming procedure, with a high potential just to confirm the previous findings and not learn anything new.One way to offset this would be to award replication studies an extra level of prestige : insist that these studies be subject to the highest level of review possible. Getting such a paper accepted would be a real challenge and recognised as such. So there would be a motivational balance of glory on the one hand, difficulty and low likelihood of new discoveries on the other. A successful replication study could also have a transformative effect on the original paper, changing it from a merely interesting result to one that deserves strong attention. That in turn encourages everyone to publish research which can be replicated in the first place.
https://astrorhysy.blogspot.com/2017/03/this-is-not-crisis-youre-looking-for.html https://astrorhysy.blogspot.com/2017/03/this-is-not-crisis-youre-looking-for.html
Scientists in society
Experts cannot compel civic engagement, and they must accept that their advice, which might seem obvious and right to them, will not always be taken in a democracy that may not value the same things they do. The job of mediating those values and policies lies with elected officials, not with scientists or other experts. The knowers cannot—and in a constitutional republic, should not—be the deciders.
At the same time, experts cannot withdraw from a public arena increasingly controlled by opportunistic demagogues who seek to discredit empiricism and rationality. Instead, the expert community must help to lead laypeople, who find the modern world intimidating and even frightening, back along the road to a better day when the citizens of the United States valued scientists and other professionals as essential parts of the American story. Experts must continue, as citizens, to advocate for those things they believe to be in the public interest, but the most important role they can play is defend a stark but empathetic insistence on science and reason as the foundation for public policy.
Indeed. But this will not help much unless experts advice is generally considered trustworthy : that is a necessary but not sufficient condition for following expert advice. Which in turn depends on how that information is communicated and disseminated :
http://astrorhysy.blogspot.cz/2016/09/would-i-lie-to-you.html
https://blogs.scientificamerican.com/guest-blog/how-does-the-public-rsquo-s-view-of-science-go-so-wrong/
At the same time, experts cannot withdraw from a public arena increasingly controlled by opportunistic demagogues who seek to discredit empiricism and rationality. Instead, the expert community must help to lead laypeople, who find the modern world intimidating and even frightening, back along the road to a better day when the citizens of the United States valued scientists and other professionals as essential parts of the American story. Experts must continue, as citizens, to advocate for those things they believe to be in the public interest, but the most important role they can play is defend a stark but empathetic insistence on science and reason as the foundation for public policy.
Indeed. But this will not help much unless experts advice is generally considered trustworthy : that is a necessary but not sufficient condition for following expert advice. Which in turn depends on how that information is communicated and disseminated :
http://astrorhysy.blogspot.cz/2016/09/would-i-lie-to-you.html
https://blogs.scientificamerican.com/guest-blog/how-does-the-public-rsquo-s-view-of-science-go-so-wrong/
Thursday, 2 March 2017
A weird thing n the Virgo cluster
This is a very interesting little object in the Virgo cluster. It consists of two connected clumps of hydrogen gas with a smattering of stars in or near to each clump. What's strange is that all these stars appear to be very young, < 50 Myr old, but it's far away from any other galaxies. It's almost like seeing a galaxy that's lighting up for the first time. Although its velocity dispersion (how fast the gas is moving around internally) is not especially large, it's high enough that it indicates it needs a dark matter component to hold it together.
The other interpretation, favoured by the authors, is that this is some form of debris either from galaxy-galaxy interactions or more likely ram pressure stripping. No dark matter, just random motions of gas that's dispersing. This might work in this case. The velocity dispersion of the object is low, easily low enough to be explicable by interactions. But it's also very isolated : at least 350 kpc (about 1 million light years) from the nearest galaxy. If the stars had stared forming as soon as the gas left its parent galaxy, it would have needed a velocity of almost 7,000 km/s to become so isolated within the maximum age of 50 Myr. That's about ten times faster than the typical velocities of galaxies in the cluster, which isn't really credible.
On the other hand perhaps the gas was removed much longer ago and has only now just started forming stars. That's much more plausible, though it implies there's a great deal of unseen gas streams lying around that have not yet collapsed to the point where star formation can occur. Which is fine by itself, since simulations show such gas streams could remain confined over the necessary timescales. My concern is : why are we only seeing this one very small patch of star formation ? Why aren't we seeing huge rivers of star formation throughout the whole stripped gas stream ? The team look for other objects and find two, which are also isolated patches of apparently young stars.
Now I might believe that we happen to be seeing one stream at almost the precise moment star formation was triggered - 50 Myr is a very short time. But in general, if we're really seeing star formation in the long tails of gas stripped by ram pressure, I'd expect to see star formation over much larger areas. To find three streams just at the moment star formation starts in very localised areas seems unlikely. And the velocity dispersion of the gas is so low that it would take about a billion years for the patch of newly-formed stars to double in size. This means that each star-forming region in the stream should remain small and easy to detect (lots of stars in a small area) for a long time. So why don't we see more of them ?
https://arxiv.org/abs/1702.07719
The other interpretation, favoured by the authors, is that this is some form of debris either from galaxy-galaxy interactions or more likely ram pressure stripping. No dark matter, just random motions of gas that's dispersing. This might work in this case. The velocity dispersion of the object is low, easily low enough to be explicable by interactions. But it's also very isolated : at least 350 kpc (about 1 million light years) from the nearest galaxy. If the stars had stared forming as soon as the gas left its parent galaxy, it would have needed a velocity of almost 7,000 km/s to become so isolated within the maximum age of 50 Myr. That's about ten times faster than the typical velocities of galaxies in the cluster, which isn't really credible.
On the other hand perhaps the gas was removed much longer ago and has only now just started forming stars. That's much more plausible, though it implies there's a great deal of unseen gas streams lying around that have not yet collapsed to the point where star formation can occur. Which is fine by itself, since simulations show such gas streams could remain confined over the necessary timescales. My concern is : why are we only seeing this one very small patch of star formation ? Why aren't we seeing huge rivers of star formation throughout the whole stripped gas stream ? The team look for other objects and find two, which are also isolated patches of apparently young stars.
Now I might believe that we happen to be seeing one stream at almost the precise moment star formation was triggered - 50 Myr is a very short time. But in general, if we're really seeing star formation in the long tails of gas stripped by ram pressure, I'd expect to see star formation over much larger areas. To find three streams just at the moment star formation starts in very localised areas seems unlikely. And the velocity dispersion of the gas is so low that it would take about a billion years for the patch of newly-formed stars to double in size. This means that each star-forming region in the stream should remain small and easy to detect (lots of stars in a small area) for a long time. So why don't we see more of them ?
https://arxiv.org/abs/1702.07719
Subscribe to:
Posts (Atom)
-
Of course you can prove a negative. In one sense this can be the easiest thing in the world : your theory predicts something which doesn...
-
Why Philosophy Matters for Science : A Worked Example "Fox News host Chris Wallace pushed Republican presidential candidate to expand...
-
In the last batch of simulations, we dropped a long gas stream into the gravitational potential of a cluster to see if it would get torn...