If you’re the type interested in peeking behind the curtains, read on:
In late September, the Hubble Space Telescope freaked out. The computer system responsible for transmitting data to the ground crashed. Thankfully, the telescope didn’t completely stop talking to the ground, but it took quite a while for the systems to come back on line. Now, thanks to NASA’s long held religious belief in redundancy, the telescope is once again taking data and transmitting the information back to hungry astronomers on the ground.
Now this failure happened just as NASA was making final preparations for the final servicing mission to HST. Since it wasn’t clear if they’d have to replace the communications electronics, NASA decided to delay the servicing mission again. (For reference, this servicing mission was originally scheduled to happen in 2003 and got sidelined in the aftermath of the Columbia disaster, so HST has been waiting for this visit for a *long* time!) Currently they’re guessing that the servicing mission will be sometime next spring.
Meanwhile, the last round of HST proposals for “Cycle 17” took place last winter and was explicitly expecting to use the new and refurbished instruments after the servicing mission. But that was assuming that the mission was going to be this last spring. It’s now been delayed at least a further year, and the observing queue for approved programs using the existing instruments is running dry. So they’ve put out an emergency call for proposals to fill in the time until the servicing mission. They are specifically looking for proposals which will either use a lot of time (>100 orbits) or “High risk/high gain” type proposals. So the gold rush is on to collect those photons which are about to “fall off the back of a truck,” and it’s time to dig up semi-crazy ideas for what to do with aging HST instruments.
Now one of the side effects of specializing in infrared observations of supernovae, is that while I’ve been able to come up with good uses for the large ground-based telescopes, it’s been much harder to come up with good HST programs. HST has an infrared instrument, but it’s frankly something of an underwhelming instrument. Its design was locked on the wrong side of the rapid development of infrared instrumentation in the 90s and by the time it was deployed, the detectors were well behind the standard for ground-based instrumentation. The field of view is pathetic, the detectors are tiny and noisy, and the sensitivity is nothing to write home about. It does have the advantage of being above the stupendous IR airglow, but even the image quality isn’t that much better than you can achieve with ground-based instruments and the new generation of laser-guide-star adaptive optics. Plus, my particular specialty has been largely spectroscopic science, and the spectroscopic capabilities of NICMOS are poor indeed. So, to date, I’ve largely dismissed NICMOS as useless, at least to me.
But now I’ve run into a different wall. It turns out that Type Ia supernovae have some very interesting behavior in the infrared at late times (roughly a year after the explosion). In particular, the late-time iron features show both a hollow (flat-topped) emission profile and are kinematically offset from the center of the explosion by several thousand kilometers per second. Unfortunately, observing these properties really pushes the sensitivity limits of the biggest telescopes on the planet. Even for supernovae in the nearest galaxies, I need to obtain spectra of objects which are several magnitudes fainter than the sky. These are observations so difficult it even impresses the guys trying to take spectra of type Ia supernovae halfway across the visible universe.
That’s all well and good, if it was easy it would already have been done, but now we’ve reached a point where it’s difficult to build on our successes. We have observed a small handful of objects using a fair bit of time on Subaru and Gemini. But now what we really need are observations of a couple of dozen objects to start looking at how these effects vary and broaden our results to the context of the general population of Type Ia supernovae. However, being limited to the nearest objects, we are stuck with asking for these observations an object or two at a time, and asking for a night or two of 8m time for each. And the TACs are understandably coming back with “what is one more spectrum going to do for you?.” So we’re trying to investigate other avenues. I’ve got a pilot proposal in with Rob Fesen & students to try and use optical data to get at the same science, which might help.
But while pondering these troubles at the SN meeting last week in Japan, I received the e-mail from SCScI announcing the new HST opportunity. So now the question is, can I use HST to learn something about more SNe Ia? My first crazy thought was to try and use the strange filter set on NICMOS to get “photometric redshifts”, using the flux ratios in neighboring filters to estimate the kinematic offsets of the iron lines. Fortunately, after a long night of sushi and sake (including the famously poisonous Fugu... it was a pretty fantastic meeting banquet), I came to my senses and remembered that NICMOS does have a rudimentary spectroscopic capability. The spectral resolution is pathetically low (about 1000 km/s) but in this case that’s actually a plus because it means I will be concentrating the faint emission into just a few pixels. So now it’s down to quantitative questions: (1) is NICMOS even sensitive enough to do this kind of observation, and (2) will I be able to push the observations out sufficiently far to measure an interesting number of supernovae?
To be continued....
In late September, the Hubble Space Telescope freaked out. The computer system responsible for transmitting data to the ground crashed. Thankfully, the telescope didn’t completely stop talking to the ground, but it took quite a while for the systems to come back on line. Now, thanks to NASA’s long held religious belief in redundancy, the telescope is once again taking data and transmitting the information back to hungry astronomers on the ground.
Now this failure happened just as NASA was making final preparations for the final servicing mission to HST. Since it wasn’t clear if they’d have to replace the communications electronics, NASA decided to delay the servicing mission again. (For reference, this servicing mission was originally scheduled to happen in 2003 and got sidelined in the aftermath of the Columbia disaster, so HST has been waiting for this visit for a *long* time!) Currently they’re guessing that the servicing mission will be sometime next spring.
Meanwhile, the last round of HST proposals for “Cycle 17” took place last winter and was explicitly expecting to use the new and refurbished instruments after the servicing mission. But that was assuming that the mission was going to be this last spring. It’s now been delayed at least a further year, and the observing queue for approved programs using the existing instruments is running dry. So they’ve put out an emergency call for proposals to fill in the time until the servicing mission. They are specifically looking for proposals which will either use a lot of time (>100 orbits) or “High risk/high gain” type proposals. So the gold rush is on to collect those photons which are about to “fall off the back of a truck,” and it’s time to dig up semi-crazy ideas for what to do with aging HST instruments.
Now one of the side effects of specializing in infrared observations of supernovae, is that while I’ve been able to come up with good uses for the large ground-based telescopes, it’s been much harder to come up with good HST programs. HST has an infrared instrument, but it’s frankly something of an underwhelming instrument. Its design was locked on the wrong side of the rapid development of infrared instrumentation in the 90s and by the time it was deployed, the detectors were well behind the standard for ground-based instrumentation. The field of view is pathetic, the detectors are tiny and noisy, and the sensitivity is nothing to write home about. It does have the advantage of being above the stupendous IR airglow, but even the image quality isn’t that much better than you can achieve with ground-based instruments and the new generation of laser-guide-star adaptive optics. Plus, my particular specialty has been largely spectroscopic science, and the spectroscopic capabilities of NICMOS are poor indeed. So, to date, I’ve largely dismissed NICMOS as useless, at least to me.
But now I’ve run into a different wall. It turns out that Type Ia supernovae have some very interesting behavior in the infrared at late times (roughly a year after the explosion). In particular, the late-time iron features show both a hollow (flat-topped) emission profile and are kinematically offset from the center of the explosion by several thousand kilometers per second. Unfortunately, observing these properties really pushes the sensitivity limits of the biggest telescopes on the planet. Even for supernovae in the nearest galaxies, I need to obtain spectra of objects which are several magnitudes fainter than the sky. These are observations so difficult it even impresses the guys trying to take spectra of type Ia supernovae halfway across the visible universe.
That’s all well and good, if it was easy it would already have been done, but now we’ve reached a point where it’s difficult to build on our successes. We have observed a small handful of objects using a fair bit of time on Subaru and Gemini. But now what we really need are observations of a couple of dozen objects to start looking at how these effects vary and broaden our results to the context of the general population of Type Ia supernovae. However, being limited to the nearest objects, we are stuck with asking for these observations an object or two at a time, and asking for a night or two of 8m time for each. And the TACs are understandably coming back with “what is one more spectrum going to do for you?.” So we’re trying to investigate other avenues. I’ve got a pilot proposal in with Rob Fesen & students to try and use optical data to get at the same science, which might help.
But while pondering these troubles at the SN meeting last week in Japan, I received the e-mail from SCScI announcing the new HST opportunity. So now the question is, can I use HST to learn something about more SNe Ia? My first crazy thought was to try and use the strange filter set on NICMOS to get “photometric redshifts”, using the flux ratios in neighboring filters to estimate the kinematic offsets of the iron lines. Fortunately, after a long night of sushi and sake (including the famously poisonous Fugu... it was a pretty fantastic meeting banquet), I came to my senses and remembered that NICMOS does have a rudimentary spectroscopic capability. The spectral resolution is pathetically low (about 1000 km/s) but in this case that’s actually a plus because it means I will be concentrating the faint emission into just a few pixels. So now it’s down to quantitative questions: (1) is NICMOS even sensitive enough to do this kind of observation, and (2) will I be able to push the observations out sufficiently far to measure an interesting number of supernovae?
To be continued....
No comments:
Post a Comment