The optical observations of the UDF from 2004 were conducted with the Advanced Camera for Surveys/Wide Field Channel (ACS/WFC) not the predecessor to WFC3 (Wide Field/Planetary Camera 2, or WF/PC2). Also, the optical channel of WFC3 does offer a small improvement in pixel scale (40 milliarcseconds/pixel, versus 50 mas/pix for ACS/WFC. However, the near-infrared channel (where these images were taken) only has a pixel scale of 130 mas/pix, a factor of ~2.5 worse than ACS/WFC.
(The diffraction limit of HST varies from ~50 mas at 500 nm to ~150 mas at 1.5 microns, so the native resolution is worse as well. However, undersampling of images due to the detectors' oversized pixels is what dominates the fine details of its effective resolution. The true resolution is actually hard to estimate offhand for images like this. Each observation occurs with sub-pixel offsets with respect to the others, so if you subsample and stack them, you can recover much of the resolution that was lost due to undersampling. With many orbits' worth of images contributing to the UDF, they might have gotten back all of the lost resolution.)
Well, there's a lot of gloomy talk circulating among people whose Cycle 17 programs are directly affected by this. This means that either the PR people at STScI are being excessively optimistic or the liaisons to the community aren't quashing rumors very effectively.
The Solar Blind Channel (the least useful component of ACS, unless you happen to use it) is the only component coming back. The Wide Field Channel and High Resolution Channel, the real workhorses, aren't coming back until after the Servicing Mission. Even then, the ACS repair is on the bottom of their priority list since most of its functionality is duplicated (albeit not as well) by the optical channel of WFC3. This means that the ACS repair, perhaps along with the STIS repair, most likely will be crowded out of the schedule by replacement of the instrument control computer.
The original discovery was a radial velocity detection, not a transit detection. The "planet" wouldn't have transited because it was thought to have an almost face-on orbit, with an inclination close to that of the protoplanetary disk surrounding TW Hya. The star spots cause an apparent RV "wobble" because they reduce the flux from a single piece of the star's surface. As the star rotates, the missing flux shows up first in the blueshifted component (the side of the star coming toward us) and then in the redshifted component (the side of the star moving away). You can often identify this effect by measuring the time-dependent shape of the spectral line. Another good test (which these authors also used) is to measure radial velocities in the near-infrared, because spots have less contrast (and therefore lower RV variation) at redder wavelengths.
Also, for whatever it's worth, there have been rumors floating around since the original announcement that several groups have photometric data showing the variations in stellar flux due to these spots. The period of this variability was supposed to be consistent to the "planet's" period, a very strong argument that it was a rotation/spot effect.
Oh, I agree that Arecibo is still doing good science, and that achieving many of those other advances would also return good science. The problem, as I just mentioned in a different response, is that "good" might not be enough to justify funding as long as NSF funding remains flat and the cost of new facilities keeps rising. For example, do we expect any new results on pulsars in the next 5 years that really compete with weak lensing or planet searches in terms of impact?
(I only ask half-rhetorically, since I'm open to arguments for a compelling science case. It's not like we're short on other science programs that can be kicked off the lifeboat.)
Regarding surveys, I completely agree that expanding discovery space is key. However, SDSS and PANSTARRS represent fundamental shifts in the quantity or type of data available. (I'm a little skeptical of LSST at the moment, since I'd like to see a post-PANSTARRS discussion of what parameter space is left.) From what I understand, the latest generation of Arecibo surveys are what I call factor-of-3 programs - a factor of ~3 more in some quantity, like area or sensitivity. These incremental surveys may give you larger samples of known classes of objects, but they don't offer the same yield of unexpected discoveries like you get from order(s) of magnitude improvements. Am I selling the Arecibo surveys short?
In an era where every telescope or survey has lifetime costs of tens to hundreds of millions, do you really think we can afford to slice up the pie by wavelength and not pit wavelengths against each other? Some fields naturally rise while others fall (just ask the solar people!), and it doesn't make sense to maintain the same fractional allocation of money.
My argument, which also applies to Scott Ransom's post, is that there are so many science cases that are truly transformative, just doing reasonable science shouldn't be enough to guarantee funding out of the relatively flat NSF pool. We have to be active in examining existing programs to determine which are still contributing as much (per dollar) as a new telescope or survey would. Most of the results I've heard coming out of Arecibo lately seem to fall in the reasonable category, not in the transformative category. I'm certainly willing to be persuaded otherwise, though.
Finally, my list was largely optical/IR because those are what I'm most familiar with, but I'm certainly willing to include the new radio/mm/submm initiatives. For example, ALMA is obviously going to be huge, and I would happily kill a number of optical telescopes if it were necessary to keep ALMA alive.
Spitzer: Transitional disks. ULIRGs. Exoplanet secondary transits. Star formation, period. Direct imaging of free-floating planetary-mass objects.
See? It's not that hard, even if you don't stray too far outside your (or your colleagues') field of specialization. There really are a lot of important (and sexy) science cases floating around, they just don't really require Arecibo.
There is no compelling science case for Arecibo that can't be pursued with other telescopes, especially since the frontier of radio astronomy has mostly moved from sensitivity (requiring big apertures) to resolution (requiring long-baseline arrays), or to shorter mm/submm wavelengths that Arecibo can't handle.
They've actually moved a large fraction of Arecibo's time over to survey efforts: "We'll do the same piece of sky, but with a flux limit 3 times deeper!" Sorry, but there are too many programs with the potential for transformative new discoveries to keep a major observatory open purely for incremental science.
The ground-based astronomical budget is finite, and the only way we're going to finance 50 shiny new programs is by shutting down the old ones that aren't scientifically competitive. Arecibo hasn't been scientifically competitive in a decade, and it won't ever be competitive in the era when we want to build LSST, PANSTARRS, TMT, ATA, ATST, and a dozen other acronyms.
We've already had one near-miss, when Hillary Clinton tried to force some budget language funding Arecibo in the weeks before the Puerto Rico primary. She didn't earmark new funding, she just added a mandate that existing funding go there. Oddly enough, the legislation didn't mention which other ground-based program would be cut to free up the funds...
Yes, but if you want to engage the vast majority of humanity and get them interested in astronomy, it can help to anthropomorphize a bit. Precision matters in journal articles, but not so much in getting votes for JWST and TPF.
For empirical evidence, I'll note that one of us sounds like a pedantic nerd with a tweed fetish, while the other is currently modded at +5. Just saying.
I doubt if this post is high enough to net any karma, but oh well. I'll chalk it up as my outreach for the day...week...year...something. The important quantity isn't the average density, but the core density (as fusion only happens near the core). As stars evolve off the main sequence, their outer layers may expand, but they also become much more centrally condensed.
During the hydrogen burning phase, inert helium gradually builds up in the core and hydrogen becomes less common. This means the core has to contract and become hotter in order to produce enough energy to support itself and the surrounding envelope. The fusion rate depends on the square of the hydrogen density (since you need the hydrogen atoms to collide with each other), so if the hydrogen density goes down, the core has to become hotter and more generally dense in order to maintain the same energy production rate. (This is why stars gradually become more luminous over their main sequence lifetime, as the core actually has to produce more energy in order to support itself in its more compact configuration.)
As a star finishes exhausting its hydrogen, this actually reaches a very extreme configuration where the core becomes much more compact (and much hotter) trying to squeeze out the required energy with very little hydrogen remaining. The total energy being produced by the core (in order to keep itself from collapsing) increases very rapidly at this point, and the larger luminosity will then push the envelope outward, puffing it up. This is why stars expand into red giants, and this is the stage where the Earth will probably be engulfed.
For trivia purposes, the central core eventually runs entirely out of hydrogen and sits there as an inert clump while the upper edges of the core burn hydrogen. When the hydrogen is exhausted for a large enough fraction of the core, the center eventually becomes hot and dense enough to fuse helium into carbon. At this point, the overall luminosity drops again (because the star doesn't need to keep frantically burning just hydrogen to support itself) and the star contracts a bit. The process then starts over again, with a shell of helium fusion surround an inert carbon core that (for stars more massive than the Sun) eventually ignites to fuse into neon, oxygen, etc.
It actually doesn't even have an imager, just spectrographs. The term "adaptive optics" refers specifically to systems where a mirror in the light path deforms at very high rates (50-2000 Hz) to correct atmospheric distortions in the wavefront of the incoming light. TMT will have this, as do the VLT, Keck, Gemini, MMT, and Palomar. TMT is just the first that is being designed from the ground up with AO in mind.
Hobby Eberly is basically a very low-budget version of telescopes like Keck. It has the same mirror size (and therefore the same light collecting ability), but they made several design compromises to knock the cost down from $100 million (for Keck) to about $15 million. Most of these compromises reduce the image quality, so they don't even bother trying. They just mounted a bunch of spectrographs since somebody taking a spectrum of a single object usually doesn't care about the nonplanar focal surface and correspondingly tiny effective field of view.
Each of those objects appears to be a point source, meaning a star. The pair has roughly the same brightness and same color, so presumably they're similar-mass stars at the same distance. Just eyeballing the magnitude to be ~17 and the spectral type to be early M, they're probably located within a kiloparsec (~3000 light years).
It's very misleading to call WFC3 a successor to ACS. The NIR channel of WFC3 will be a significant upgrade over NICMOS in terms of throughput and field of view (and resolution, compared to NIC3). However, the UVis channel is clearly a compromise between the capabilities of the Wide-Field and High-Resolution channels of ACS, and as such it really can't compete with either (except insomuch as they're dead and it isn't).
WFC3/UVis is a clear step down on high-resolution imaging because the 40 mas pixels of WFC3 are larger than ACS/HRC and will undersample the PSF. It's an even larger step down on wide-field survey work because its throughput and field of view are both markedly inferior to the ACS/WFC (factors of 2/3 and 1/2 respectively). Completing a survey of the same area and depth with WFC3/UVIS will require 3 times as many orbits. Can anybody imagine the TAC approving 1800 orbits on something like COSMOS or the UDF?
This post brought to you by an astronomer who's procrastinating on rewriting his ACS-based HST proposal.
We've known for about a decade that the binary frequency among low-mass (early M) stars is only 30-35%. We've also known for at least that long that the general shape of the field mass function is weighted in favor of low-mass stars. It's a very short leap to draw the corresponding conclusion, and it's been done in plenty of other papers that actually present useful results at the same time.
For those who care about the background, the binary frequency has been shown pretty clearly to depend on mass. Solar-mass stars have binary frequencies of at least 60%, stars of 0.5 solar masses have binary frequencies of ~35%, and very low-mass stars and brown dwarfs (under 0.2 solar masses) have binary frequencies of around 10-20%. The binary frequency among more massive stars appears to be even higher than for solar-mass stars.
The popular reason to care about binary frequencies is to determine the frequency with which planetary systems could occur. If you're interested in habitable planets around solar-type stars, the higher binary frequency is one to care about. The frequency with which planets could form around lower-mass stars is intrinsically interesting since they're so common, but they're also much harder to detect any of these planets using existing indirect methods, so it's a harder question to actually answer. Once we have the ability to directly image planets, the problem will invert itself since it's easier to see planetary companions to faint stars than bright stars.
From what I know, magnetars radiate most of their energy on an extremely short timescale, of order tens of thousands of years or so. Considering how rare they are, the number of stars that are irradiated by SGR flares must be pretty small, and so any additional term in the Drake equation would be very, very close to unity.
If anyone wants to cruise for mod points, you could do an order-of-magnitude estimate of the fraction of irradiated stars using the age and total volume of the Milky Way, the mean time between SGR flares of this magnitude (call it a decade to a century), and the radius of OMG-We're-All-Gonna-Die that was specified in the article.
Of course, the supernova explosion that led to a magnetar's formation would would have already done quite a bit of damage to the surrounding area, so they aren't likely to have any meaningful impact on any planetary systems around them anyway.
Come on, give me a break. I've seen some of the science being done on this flare. There are enough cool things without being needlessly sensational, and invoking the Wipe-Out-All-Civilization radius definitely counts as sensational. After all, isn't the nearest magnetar something like 5 kiloparsecs away?
...there's no reason why you can't have an eigenvalue of 0.23. Of course, I can't see any case where you could solve that in your head in a sophmore physics class.
Heh. That's one of the reasons I just left Kansas. I like living in a Republican state and all, but I'd rather live someplace where the moderates run things instead of the reactionaries.
For those who follow this field, I'll remind you of the OGLE project, which has been doing the same thing from the ground. They found 60 likely planetary candidates (out of a similar number of stars monitored), but only two of those actually look like they could be planets. All the rest are either grazing-incidence binaries or blended binaries. The higher resolution of Hubble may help the blend problem to an extent, but I highly doubt the number of actual planets is anywhere near 100.
They also have little chance of confirming whether these are actually planets, as you need to do extremely high-resolution spectroscopy in order to confirm its existence via the radial velocity method. Even Keck can only do that for stars down to ~16th magnitude, and according to the observing proposal, this survey is going down to 23rd. They might be able to get precise-enough light curves to reject false positives based on color-curve changes, but I'd like to see it before I believe it.
The issue isn't about materialism at all. The space shuttle is a national asset. As such, the government has to weigh its value to the nation as a whole. It sucks for the astronauts, but after a point, you simply have to decide that the interests of 280 million people outweigh those of seven.
This is the same reason we don't leave Iraq in order to save hostages and the same reason we don't spend ten billion trillion dollars installing tons of high-tech armor on every humvee. Government is about assessing cost/benefit ratios, and when those in charge forget it, we all land in deep trouble.
The real issue here isn't astronaut safety, but asset safety. We have hundreds of astronauts, but only three shuttles. As such, we should be concentrating solely on how to maximize their survivability and not expending so many resources on crew survivability in the event of a catastrophic failure.
For those who travel cheap, a lot of KOAs are also being wired as hot spots. Unfortunately, the access charges tend to be rather steep. I was told at the KOA in Cedar City, Utah that it'd be $3.95 for one hour of access. I get the impression that flat-rate packages are a much better deal, though.
Actually, that's a problem both parties have suffered, particularly in the Senate. Zell Miller, a conservative Democrat from Georgia, often sides with the Republicans on issues. I believe he's even scheduled to speak at the Republican Presidential Convention this fall. Likewise, Jim Jeffords used to be a liberal Republican from (I believe) Connecticut, but he redeclared as an independent and allied with the Democrats in 2001 in order to give them control of the Senate.
This issue can also be seen in abortion debates. A lot of moderate and liberal Republicans are pro-choice, and more than a few conservative Democrats are pro-life. The Kansas Republican party has split into two wings (moderate and conservative) that have all but declared war on each other over this and a few other issues.
Slashdot Mirror
The optical observations of the UDF from 2004 were conducted with the Advanced Camera for Surveys/Wide Field Channel (ACS/WFC) not the predecessor to WFC3 (Wide Field/Planetary Camera 2, or WF/PC2). Also, the optical channel of WFC3 does offer a small improvement in pixel scale (40 milliarcseconds/pixel, versus 50 mas/pix for ACS/WFC. However, the near-infrared channel (where these images were taken) only has a pixel scale of 130 mas/pix, a factor of ~2.5 worse than ACS/WFC.
(The diffraction limit of HST varies from ~50 mas at 500 nm to ~150 mas at 1.5 microns, so the native resolution is worse as well. However, undersampling of images due to the detectors' oversized pixels is what dominates the fine details of its effective resolution. The true resolution is actually hard to estimate offhand for images like this. Each observation occurs with sub-pixel offsets with respect to the others, so if you subsample and stack them, you can recover much of the resolution that was lost due to undersampling. With many orbits' worth of images contributing to the UDF, they might have gotten back all of the lost resolution.)
Well, there's a lot of gloomy talk circulating among people whose Cycle 17 programs are directly affected by this. This means that either the PR people at STScI are being excessively optimistic or the liaisons to the community aren't quashing rumors very effectively.
The Solar Blind Channel (the least useful component of ACS, unless you happen to use it) is the only component coming back. The Wide Field Channel and High Resolution Channel, the real workhorses, aren't coming back until after the Servicing Mission. Even then, the ACS repair is on the bottom of their priority list since most of its functionality is duplicated (albeit not as well) by the optical channel of WFC3. This means that the ACS repair, perhaps along with the STIS repair, most likely will be crowded out of the schedule by replacement of the instrument control computer.
Also, for whatever it's worth, there have been rumors floating around since the original announcement that several groups have photometric data showing the variations in stellar flux due to these spots. The period of this variability was supposed to be consistent to the "planet's" period, a very strong argument that it was a rotation/spot effect.
(I only ask half-rhetorically, since I'm open to arguments for a compelling science case. It's not like we're short on other science programs that can be kicked off the lifeboat.)
Regarding surveys, I completely agree that expanding discovery space is key. However, SDSS and PANSTARRS represent fundamental shifts in the quantity or type of data available. (I'm a little skeptical of LSST at the moment, since I'd like to see a post-PANSTARRS discussion of what parameter space is left.) From what I understand, the latest generation of Arecibo surveys are what I call factor-of-3 programs - a factor of ~3 more in some quantity, like area or sensitivity. These incremental surveys may give you larger samples of known classes of objects, but they don't offer the same yield of unexpected discoveries like you get from order(s) of magnitude improvements. Am I selling the Arecibo surveys short?
My argument, which also applies to Scott Ransom's post, is that there are so many science cases that are truly transformative, just doing reasonable science shouldn't be enough to guarantee funding out of the relatively flat NSF pool. We have to be active in examining existing programs to determine which are still contributing as much (per dollar) as a new telescope or survey would. Most of the results I've heard coming out of Arecibo lately seem to fall in the reasonable category, not in the transformative category. I'm certainly willing to be persuaded otherwise, though.
Finally, my list was largely optical/IR because those are what I'm most familiar with, but I'm certainly willing to include the new radio/mm/submm initiatives. For example, ALMA is obviously going to be huge, and I would happily kill a number of optical telescopes if it were necessary to keep ALMA alive.
Spitzer: Transitional disks. ULIRGs. Exoplanet secondary transits. Star formation, period. Direct imaging of free-floating planetary-mass objects.
See? It's not that hard, even if you don't stray too far outside your (or your colleagues') field of specialization. There really are a lot of important (and sexy) science cases floating around, they just don't really require Arecibo.
They've actually moved a large fraction of Arecibo's time over to survey efforts: "We'll do the same piece of sky, but with a flux limit 3 times deeper!" Sorry, but there are too many programs with the potential for transformative new discoveries to keep a major observatory open purely for incremental science.
We've already had one near-miss, when Hillary Clinton tried to force some budget language funding Arecibo in the weeks before the Puerto Rico primary. She didn't earmark new funding, she just added a mandate that existing funding go there. Oddly enough, the legislation didn't mention which other ground-based program would be cut to free up the funds...
For empirical evidence, I'll note that one of us sounds like a pedantic nerd with a tweed fetish, while the other is currently modded at +5. Just saying.
During the hydrogen burning phase, inert helium gradually builds up in the core and hydrogen becomes less common. This means the core has to contract and become hotter in order to produce enough energy to support itself and the surrounding envelope. The fusion rate depends on the square of the hydrogen density (since you need the hydrogen atoms to collide with each other), so if the hydrogen density goes down, the core has to become hotter and more generally dense in order to maintain the same energy production rate. (This is why stars gradually become more luminous over their main sequence lifetime, as the core actually has to produce more energy in order to support itself in its more compact configuration.)
As a star finishes exhausting its hydrogen, this actually reaches a very extreme configuration where the core becomes much more compact (and much hotter) trying to squeeze out the required energy with very little hydrogen remaining. The total energy being produced by the core (in order to keep itself from collapsing) increases very rapidly at this point, and the larger luminosity will then push the envelope outward, puffing it up. This is why stars expand into red giants, and this is the stage where the Earth will probably be engulfed.
For trivia purposes, the central core eventually runs entirely out of hydrogen and sits there as an inert clump while the upper edges of the core burn hydrogen. When the hydrogen is exhausted for a large enough fraction of the core, the center eventually becomes hot and dense enough to fuse helium into carbon. At this point, the overall luminosity drops again (because the star doesn't need to keep frantically burning just hydrogen to support itself) and the star contracts a bit. The process then starts over again, with a shell of helium fusion surround an inert carbon core that (for stars more massive than the Sun) eventually ignites to fuse into neon, oxygen, etc.
As far as I can tell, it started in Los Angeles sometime in the last few weeks.
Hobby Eberly is basically a very low-budget version of telescopes like Keck. It has the same mirror size (and therefore the same light collecting ability), but they made several design compromises to knock the cost down from $100 million (for Keck) to about $15 million. Most of these compromises reduce the image quality, so they don't even bother trying. They just mounted a bunch of spectrographs since somebody taking a spectrum of a single object usually doesn't care about the nonplanar focal surface and correspondingly tiny effective field of view.
Each of those objects appears to be a point source, meaning a star. The pair has roughly the same brightness and same color, so presumably they're similar-mass stars at the same distance. Just eyeballing the magnitude to be ~17 and the spectral type to be early M, they're probably located within a kiloparsec (~3000 light years).
WFC3/UVis is a clear step down on high-resolution imaging because the 40 mas pixels of WFC3 are larger than ACS/HRC and will undersample the PSF. It's an even larger step down on wide-field survey work because its throughput and field of view are both markedly inferior to the ACS/WFC (factors of 2/3 and 1/2 respectively). Completing a survey of the same area and depth with WFC3/UVIS will require 3 times as many orbits. Can anybody imagine the TAC approving 1800 orbits on something like COSMOS or the UDF?
This post brought to you by an astronomer who's procrastinating on rewriting his ACS-based HST proposal.
For those who care about the background, the binary frequency has been shown pretty clearly to depend on mass. Solar-mass stars have binary frequencies of at least 60%, stars of 0.5 solar masses have binary frequencies of ~35%, and very low-mass stars and brown dwarfs (under 0.2 solar masses) have binary frequencies of around 10-20%. The binary frequency among more massive stars appears to be even higher than for solar-mass stars.
The popular reason to care about binary frequencies is to determine the frequency with which planetary systems could occur. If you're interested in habitable planets around solar-type stars, the higher binary frequency is one to care about. The frequency with which planets could form around lower-mass stars is intrinsically interesting since they're so common, but they're also much harder to detect any of these planets using existing indirect methods, so it's a harder question to actually answer. Once we have the ability to directly image planets, the problem will invert itself since it's easier to see planetary companions to faint stars than bright stars.
If anyone wants to cruise for mod points, you could do an order-of-magnitude estimate of the fraction of irradiated stars using the age and total volume of the Milky Way, the mean time between SGR flares of this magnitude (call it a decade to a century), and the radius of OMG-We're-All-Gonna-Die that was specified in the article.
Of course, the supernova explosion that led to a magnetar's formation would would have already done quite a bit of damage to the surrounding area, so they aren't likely to have any meaningful impact on any planetary systems around them anyway.
Come on, give me a break. I've seen some of the science being done on this flare. There are enough cool things without being needlessly sensational, and invoking the Wipe-Out-All-Civilization radius definitely counts as sensational. After all, isn't the nearest magnetar something like 5 kiloparsecs away?
...there's no reason why you can't have an eigenvalue of 0.23. Of course, I can't see any case where you could solve that in your head in a sophmore physics class.
Of course, what'd I do? I moved to California...
For those who follow this field, I'll remind you of the OGLE project, which has been doing the same thing from the ground. They found 60 likely planetary candidates (out of a similar number of stars monitored), but only two of those actually look like they could be planets. All the rest are either grazing-incidence binaries or blended binaries. The higher resolution of Hubble may help the blend problem to an extent, but I highly doubt the number of actual planets is anywhere near 100.
They also have little chance of confirming whether these are actually planets, as you need to do extremely high-resolution spectroscopy in order to confirm its existence via the radial velocity method. Even Keck can only do that for stars down to ~16th magnitude, and according to the observing proposal, this survey is going down to 23rd. They might be able to get precise-enough light curves to reject false positives based on color-curve changes, but I'd like to see it before I believe it.
This is the same reason we don't leave Iraq in order to save hostages and the same reason we don't spend ten billion trillion dollars installing tons of high-tech armor on every humvee. Government is about assessing cost/benefit ratios, and when those in charge forget it, we all land in deep trouble.
The real issue here isn't astronaut safety, but asset safety. We have hundreds of astronauts, but only three shuttles. As such, we should be concentrating solely on how to maximize their survivability and not expending so many resources on crew survivability in the event of a catastrophic failure.
For those who travel cheap, a lot of KOAs are also being wired as hot spots. Unfortunately, the access charges tend to be rather steep. I was told at the KOA in Cedar City, Utah that it'd be $3.95 for one hour of access. I get the impression that flat-rate packages are a much better deal, though.
This issue can also be seen in abortion debates. A lot of moderate and liberal Republicans are pro-choice, and more than a few conservative Democrats are pro-life. The Kansas Republican party has split into two wings (moderate and conservative) that have all but declared war on each other over this and a few other issues.