Ten Things You Don’t Know About Hubble

On April 24, 1990, the Space Shuttle Discovery roared into space, carrying on board a revolution: The Hubble Space Telescope. It was the largest and most sensitive optical-light telescope ever launched into space, and while it suffered initially from a focusing problem, it would soon return some of the most amazing and beautiful astronomical images anyone had ever seen.

Hubble was designed to be periodically upgraded, and even as I write this, astronauts are in the Space Shuttle Atlantis installing two new cameras, fixing two others, and replacing a whole slew of Hubble's parts. This is the last planned mission, ever, to service the venerable 'scope, so what better time to talk about it?

Plus, it's arguably the world's most famous telescope (it's probably the only one people know by name), and yet I suspect that there are lots of things about it that might surprise you. So I present to you Ten Things You Don't Know About the Hubble Space Telescope.



Hubble has observed every planet in the solar system but one: Mercury.

So Hubble has observed the Sun, but it did so literally bass-ackwards. That was to protect its mirror; raw UV from the Sun can photochemically damage sensitive parts inside the 'scope, and of course can heat them up to dangerous levels. Also, as I pointed out before, some of the cameras would in fact be damaged by direct sunlight.

Because of that, Hubble is not allowed to point anywhere near the Sun, just to make sure no stray light seeps in. This "solar avoidance zone" is a circle 50 degrees in radius around our star. Anything closer than that is forbidden.

This directive has been broken by Hubble precisely once: to observe Venus, which gets about 45 degrees from the Sun at maximum. These observations were made using WFPC2 (shown in the image above; it was taken in the near-UV to see structure in the Venusian clouds) and the Goddard High Resolution Spectrograph. Astronomers were looking for sulfur dioxide in the atmosphere of Venus, a chemical which had been detected by an earlier probe and might be emitted by volcanism on the planet. All kinds of overrides had to be sent to the telescope to allow these observations, and it was so difficult that it hasn't been and probably won't ever be repeated.

But Mercury never gets even that far from the Sun; at most it is a mere 28 degrees from the Sun, far to close to ever be seen by Hubble. But that's OK: we have the MESSENGER spacecraft. It's zipped past Mercury twice already, and will fly by Mercury one more time in the coming months before falling into orbit around the innermost planet in 2011, where it will map the planet with far higher detail than Hubble ever could.

Not everything it sees is on purpose.

This one's a bit personal, so allow me to expound a bit here.

Hubble has several cameras on board. They sit in the very bottom of Hubble, in the wider portion below the mirror (unlike a normal telescope, the mirror for Hubble is located a third of the way up from the aft end). Each sees a slightly different region of the sky, separated by a few arcminutes (the Moon is 30 arcminutes across for comparison). So if one camera is being used to look at, say, the heart of the Andromeda galaxy, then the others are looking near the galaxy's center but not right at it.

Enter the Parallels Program. When a new solid state recorder was placed on board in 1997, it greatly enhanced Hubble's capability to record data (which was done using tape drives before then). The pipeline was fat enough to record data from three cameras at the same time, so when one was observing as the primary camera, the other two could take data as well.

Sometimes while observing some primary target Hubble would be rotated to point the other cameras at something interesting (like was done with the lunar observations I mentioned earlier), but sometimes they were simply allowed to record whatever the heck they saw. This procedure was called the Parallels Program, because the other cameras were used in parallel with the primary one.

In October of 1997, Hubble was pointed at the Large Magellanic Cloud, a small galaxy that orbits the Milky Way. WFPC2 was the primary instrument, but the camera I worked on, the Space Telescope Imaging Spectrograph, or STIS, was being used as a parallel instrument. That happened a lot, and at my office the first thing I would do every morning was go through the previous day's parallels using STIS and see if there was anything interesting in them.

Yes, part of my job was to look at Hubble images of regions of space no one had ever seen before and check them out. And yes, it was pretty damn cool.

The majority of the time there wasn't much to see: faint fuzzy galaxies, or a wisp of nebulosity. Sometimes the primary camera would observe a nearby galaxy many times over the course of months, and after a while just by glancing at the STIS image I could tell you what galaxy it was from the brightness and density of stars. Not a terribly marketable skill, but still. Cool.

Anyway, one day we got that LMC observation -- the one shown above -- and I noticed the fuzzy circle at the top. I knew right away it was a small planetary nebula, a blast of gas emitted from a dying star. You can see it in the image, and it's zoomed at the bottom left. To my disappointment it had been discovered before, so this wasn't new and I couldn't name it. But we did get good spectra, which allowed me to take some basic diagnostics of the nebula that hadn't been done before.

I was able to publish my results in a paper, which also was nice. My work on STIS was awesomely fun sometimes, but I rarely got to publish anything; my name was always way down the list of people who contributed to the work. So this was a nice perq.

The Parallels Program still continues. I don't know what it's found since I left the project. Maybe someday I'll poke around the archives and find out.

Introduction

On April 24, 1990, the Space Shuttle Discovery roared into space, carrying on board a revolution: The Hubble Space Telescope. It was the largest and most sensitive optical-light telescope ever launched into space, and while it suffered initially from a focusing problem, it would soon return some of the most amazing and beautiful astronomical images anyone had ever seen.

Hubble was designed to be periodically upgraded, and even as I write this, astronauts are in the Space Shuttle Atlantis installing two new cameras, fixing two others, and replacing a whole slew of Hubble's parts. This is the last planned mission, ever, to service the venerable 'scope, so what better time to talk about it?

Plus, it's arguably the world's most famous telescope (it's probably the only one people know by name), and yet I suspect that there are lots of things about it that might surprise you. So I present to you Ten Things You Don't Know About the Hubble Space Telescope.



Hubble took the deepest visible light image yet made.

In 2003, an astronomer (and friend with whom I worked on a Hubble project) named Tom Brown pointed Hubble at the outer fringes of the Andromeda Galaxy, a nearby large spiral like our own Milky Way. Using the Advanced Camera for Surveys, he commanded the space telescope to basically sit and stare at one spot for a total of three and a half days. His goal to was to be able to get good data on very faint stars in Andromeda, to characterize the way stars form in the galaxy.

He certainly was able to do that (and found many stars younger than expected; in Andromeda's halo the stars were several billion years younger than in our own halo), but what he also got was the deepest optical image of the Universe ever taken. Stars down to 31st magnitude can be seen in the data -- those are stars one ten-billionth as bright as what you can see with your unaided eye!

The image here shows different regions in that deep image. You can see faint background galaxies, stars in both Andromeda and the Milky Way, a densely-packed globular cluster, and much more.

The Moon is not too bright to see with Hubble.

A lot of people claim that some objects are simply too bright to observe with Hubble. For some limited cases this is true -- there's a camera on board Hubble sensitive to ultraviolet light, and at a 2500 Volt potential too many UV photons can fry the instrument.

But that's not true for most of Hubble's cameras. Actually, some of the brightest objects in the sky have been observed... including the Moon! The image shown here is of Copernicus, a 90 kilometer wide impact crater on the Moon. It wasn't actually Hubble's primary target; another camera (the Space Telescope Imaging Spectrograph, or STIS, a camera I worked on for many years) was observing reflected sunlight off the Moon's limb, and Hubble was rotated so that Wide Field/Planetary Camera 2 (WFPC2) would be able to take snapshots of the crater.

So while the Moon is not too bright to observe with Hubble, it is moving too rapidly across the sky for the 'scope to track it. So the observations were made in what's called "ambush mode": Hubble is pointed at a spot in the sky where the Moon is going to be, and when the right moment arrives the images are taken. It's a very difficult operation, which is one the reasons why there are so few observations of our nearest neighbor.

Back in 1999 I took part in a set of lunar observations using Hubble; we were hoping to get spectra of water ice splashed up from the Moon's south pole when the Lunar Prospector probe impacted there at high speed. Unfortunately, the spectra were screwed up; the pointing was off by a bit and we didn't see anything (it turns out no one saw anything using any telescope, so we didn't really miss much). Although it failed, that observation run was incredibly exciting, some of the most fun I've had using Hubble.

One problem with using digital detectors is knowing exactly what you're seeing. If a star looks brighter than another, is the star really brighter, or is the electronic chip just a little too sensitive right there? You have to calibrate the chip to know exactly what it's doing. There are several steps in that process, but one involves using a "flat field", observing a region of the sky that is perfectly evenly illuminated. That way, if one pixel or another is too sensitive, you can see it in the observation.

With Hubble, though, every patch of sky has some object in it, which would screw up the flat field. Some telescopes have internal illumination; little LEDs or some other method, but using them is notoriously difficult to get an evenly illuminated field. So what can you do when using Hubble?

One method is to observe the Earth! As Hubble orbits at 8 km/sec, the out-of-focus Earth screams by. If you observe for a while, objects will actually leave streaks in the image, and these can be treated mathematically to produce a flat field. The image shown here is just such a "streak flat". That's a Hubble observation of our home planet, with objects flying past. It's hard to say what they are, exactly. It depends on where Hubble was when the image was taken, and where it was pointed. They might be trees, hills, valleys, mountains, or even houses!

But don't worry, it can't see people. If the Moon is too fast to track, the Earth is certainly out of the question. But y'know, the company that made Hubble's mirror had an awful lot of those same sized mirrors lying around, and there are no other astronomical telescopes (you know, telescopes that point away from the Earth) with that same mirror. So what could those mirrors have been for?

Hmmm.

It observes the Earth... quite often!

If the Moon is not too bright to see, what about the Earth? On average, it's much more reflective and therefore much brighter. Well, it turns out Hubble not only has observed the Earth... it's done it thousands of times!

One problem with using digital detectors is knowing exactly what you're seeing. If a star looks brighter than another, is the star really brighter, or is the electronic chip just a little too sensitive right there? You have to calibrate the chip to know exactly what it's doing. There are several steps in that process, but one involves using a "flat field", observing a region of the sky that is perfectly evenly illuminated. That way, if one pixel or another is too sensitive, you can see it in the observation.

With Hubble, though, every patch of sky has some object in it, which would screw up the flat field. Some telescopes have internal illumination; little LEDs or some other method, but using them is notoriously difficult to get an evenly illuminated field. So what can you do when using Hubble?

One method is to observe the Earth! As Hubble orbits at 8 km/sec, the out-of-focus Earth screams by. If you observe for a while, objects will actually leave streaks in the image, and these can be treated mathematically to produce a flat field. The image shown here is just such a "streak flat". That's a Hubble observation of our home planet, with objects flying past. It's hard to say what they are, exactly. It depends on where Hubble was when the image was taken, and where it was pointed. They might be trees, hills, valleys, mountains, or even houses!

But don't worry, it can't see people. If the Moon is too fast to track, the Earth is certainly out of the question. But y'know, the company that made Hubble's mirror had an awful lot of those same sized mirrors lying around, and there are no other astronomical telescopes (you know, telescopes that point away from the Earth) with that same mirror. So what could those mirrors have been for?


Hubble once observed... wait for it... wait for it... THE SUN.

Oh, I got you with that one, didn't I? Admit it: you had no idea that Hubble actually and for real once observed the Sun, on purpose. I didn't know about it for a long time, until my friend and fellow astronomer Glenn Schneider clued me in. Glenn is a surprising guy in many ways -- he chases solar eclipses all over the planet, for example -- but this one was a doozy.

He has the whole story on his website. The short version is that some kinds of electronic detectors get extra electrons trapped in them, kinda like plaque in your arteries. One way to flush out these extra electrons is to flood the detector with ultraviolet light. The chips used in the original Wide Field/Planetary Camera launched with Hubble suffered from this, so they needed that UV flood. And it turns out there's a fairly bright source of UV light in space...

Maybe you see where this is going.

So the engineers rigged WFPC with a little mirror that stuck outside the camera. This part of the camera was actually mounted flush against Hubble's side, so the mirror stuck out from the 'scope like a wee periscope (there's a picture on Glenn's site that'll help). It faced backwards, towards Hubble's aft end. The great observatory was then pointed in the opposite direction of the Sun so the rear-view mirror was facing the Sun, and the sunlight was channeled right into WFPC.

The result is the image above: a bona-fide 100% actual image (well, mosaic) of the Sun taken by the Hubble Space Telescope.

How freaking cool is that?

HHubble cannot see the Apollo artifacts on the Moon.

This question is sent to me roughly once a month, and sometimes even more often: why don't we shut up the people who think the Apollo Moon landings were faked by pointing Hubble at the Moon and taking pictures of the Apollo sites?

Well, one reason is that, duh, NASA and astronomers have better things to do than try to prove something blaringly obvious to people who would just claim the resulting images are faked anyway.

But also, Hubble cannot see the artifacts on the Moon! They're way too small.

This surprises a lot of folks, since they're used to seeing razor-sharp images of nebulae and galaxies. However, remember that while those objects may be far away, they are also very, very big. Light years across, maybe thousands of light years across. The remains of the lunar landers are only 4 meters across. That's a tad smaller.

Sure, you say, but the Moon's a lot closer, right? Yeah, it is, but it turns out it's not close enough.

You can calculate how small an object a telescope can resolve (that is, see as more than just a one pixel wide dot) using really basic algebra. It depends mostly on the telescope's mirror size. When you do this for Hubble, you get an angular measure of about 0.1 arcseconds, a tiny amount to be sure. The Moon is 1800 arcseconds across, so 0.1 arcseconds corresponds to about 200 meters on the Moon! In other words, something has to be bigger than a football stadium on the Moon before Hubble can see it.

It's surprising, I know, but that's how the math works out. The lunar lander is about 0.002 arcseconds in size, well beyond the capabilities of any normal telescope (go to that link above for more info on ways this still might work).

So really, the only -- and best -- way to see the Apollo artifacts is to go back to the Moon. Of course, the Moon hoax believers will still deny it's real. Their refusal to see reality is cosmic in its proportions.



Even today, 19 years after Hubble's launch, it's not all that uncommon to hear a newscaster refer to "Hubble's lens". I once heard it used by an announcer on a science show produced by the Space Telescope Science Institute, the agency that runs Hubble!

The thing is, Hubble doesn't use a lens. It has a mirror.

Galileo used a telescope with a lens, as did everyone up until Isaac Newton. He was the genius who figured out that a properly shaped mirror could focus light as well, and has advantages over a lens: mirrors need only be ground on one side (lenses have two), and mirrors can be made larger than lenses because they can be supported all across their back side, while lenses must be supported around their circumference, where the glass is thinnest and most vulnerable.

Over a certain size, lenses are simply impractical, so mirrors are used. Hubble's primary mirror is 2.4 meters across, about 8 feet. Although it's the biggest mirror for astronomy ever lofted into space, it's considered small by ground-based standards; many telescopes today have mirrors 4 or more meters across. The mammoth twin Keck 'scopes in Hawaii have mirrors made of segments that total 10 meters across each!

It turns out the cameras on board Hubble use mirrors too. Why? Glass absorbs light. Not much, maybe 2% of the incident light, but that adds up. A lens has two surfaces, each of which reflect a little bit of light, so you lose more through a lens than with a mirror. Also, mirrors can be made to reflect light of different colors about the same, but lenses bend light at different colors differently. So all in all, doing it with mirrors makes a lot more sense.

However, there are lenses on board: they are used in the Fine Guidance Sensors, small telescopes that track stars with incredible accuracy and help keep Hubble locked onto to its targets.

Hubble doesn't use lenses. Sorta.

Even today, 19 years after Hubble's launch, it's not all that uncommon to hear a newscaster refer to "Hubble's lens". I once heard it used by an announcer on a science show produced by the Space Telescope Science Institute, the agency that runs Hubble!

The thing is, Hubble doesn't use a lens. It has a mirror.

Galileo used a telescope with a lens, as did everyone up until Isaac Newton. He was the genius who figured out that a properly shaped mirror could focus light as well, and has advantages over a lens: mirrors need only be ground on one side (lenses have two), and mirrors can be made larger than lenses because they can be supported all across their back side, while lenses must be supported around their circumference, where the glass is thinnest and most vulnerable.

Over a certain size, lenses are simply impractical, so mirrors are used. Hubble's primary mirror is 2.4 meters across, about 8 feet. Although it's the biggest mirror for astronomy ever lofted into space, it's considered small by ground-based standards; many telescopes today have mirrors 4 or more meters across. The mammoth twin Keck 'scopes in Hawaii have mirrors made of segments that total 10 meters across each!

It turns out the cameras on board Hubble use mirrors too. Why? Glass absorbs light. Not much, maybe 2% of the incident light, but that adds up. A lens has two surfaces, each of which reflect a little bit of light, so you lose more through a lens than with a mirror. Also, mirrors can be made to reflect light of different colors about the same, but lenses bend light at different colors differently. So all in all, doing it with mirrors makes a lot more sense.

However, there are lenses on board: they are used in the Fine Guidance Sensors, small telescopes that track stars with incredible accuracy and help keep Hubble locked onto to its targets.



You can see one of Hubble's cameras in the National Air and Space Museum.

Like I said, Hubble was designed to be periodically updated. When new tech makes for better cameras, old ones can be taken out and replaced with new ones. When STIS and the infrared camera NICMOS were inserted into Hubble in 1997, the Goddard Spectrograph and the Faint Object Spectrograph (FOS) were removed.

While I was still at Goddard Space Flight Center, I used to take a walk around the compound after lunch. I'd sometimes slip through one building that had a massive warehouse, and usually there was something cool to see in there. I saw satellites being constructed, the upper stage of a rocket (without fuel!) on a crane, and all sorts of odd and wonderful sights.

One day, from across the warehouse, I spot what looks like a big black telephone booth sitting on a pallet. Could it be...? I walked over, and yes! It was the FOS! I couldn't believe it. It was just sitting there, this camera which cost tens of millions of dollars to build. Two sides of it had been removed, and one had been replaced with clear thick plastic. I realized it must be going to a museum; the plastic would allow people to see inside it. But one panel was still removed, so the guts of the camera were exposed. Hmmm...

So of course I reached in and poked around. I had used the FOS for my PhD, analyzing spectra it had taken of an exploding star on two different dates. We wound up not using the data because we didn't know precisely where the telescope was pointed each time, and so I couldn't compare one spectrum with another. Still, I spent months learning how the camera worked, and seeing it in front of me was too tempting. It was amazing; I could see exactly how it worked, and all those diagrams I had pored over five years earlier suddenly came alive.

I convinced a friend to come with me the next day to see it, and he took the picture above of me pretending (Yes! Pretending! That's it!) to snip the wires with a wire cutter. Haha!


Years later, I was visiting DC. I went to the National Air and Space Museum, having completely forgotten the incident at Goddard. I rounded a corner, and there was my old friend. I smiled; I knew it would end up here. The second exterior panel had been replaced with plastic, and you could see into the camera. If you compare the picture above with the one here (click to embiggen) you can see it's the same beast.


It's the only piece of Hubble I ever physically touched. Well, besides the insulating blanket that flew on Hubble for years and was taken back to Earth after a servicing mission. Someone had draped the shiny silver blanket over a chair in a room we used to test STIS. When I saw it, I... hmmmm. No. That's a whole 'nuther story.

You can look at all the images it has ever taken, as long as they're over a year old.

Since its launch in 1990, Hubble has orbited the Earth over 100,000 times and taken something like a half million separate observations. Those figures alone are a bit staggering. But did you know that you (potentially) have access to those images? Well, most of them, anyway.

All the data taken by Hubble that is more than one year old is stored in an archive that the public can query. Want to know what Hubble was observing on your birthday two years ago, or at the moment your kid was born? Just ask the database! In many cases, when you search the database, you can also get a preview of the image; the above shot is of the spiral galaxy M51, also called the Whirlpool Galaxy. The preview shows the raw data right off the 'scope; it's not always particularly pretty. To beautify it you need to process it, which means subtracting a dark frame, a bias frame, dividing by the flat field, flagging bad pixels, combining multiple exposures to get rid of cosmic rays, performing a geometric correction... and if you want color, you have to do that for the other filters used in the observation, and then combining those using Photoshop or some other software.

Obviously, not everyone can do that (it's a lot harder even than it sounds). So not everyone is allowed to actually retrieve the data; that would strain the archive servers. To do that you have to justify the need and get an account. I used to have one, but I lost my password a long time ago. Probably all for the best; I'd just download gigabytes of cool images and get everyone at the archive ticked off at me.

Oh, about that "... as long as they're over a year old" thing: data is proprietary to the person who took it for the period of one year, so the scientists involved have time to look it over. It does take some time to process the data, and a lot more time to analyze it; if everyone had instant access to all the data, someone more experienced than you could scoop you on your own observations! However, it's also not fair to let people have the data forever. The compromise is the one year proprietary period; that gives scientists time to look things over, but still motivates them to get things done. I think this is a fine idea, and it even works in practice in the real world, amazingly. If a scientist wants, they can release the data early, too, so everyone wins.

In fact, I used old data quite a bit back in the day. If we found something interesting in our own data, we could go look for older observations to see if it had been seen before, or if there were related observations. And many times, even if the older data were still proprietary, the scientists involved were interested in collaborating. Funny thing about scientists: in lots of cases they are open, friendly, and interested in seeing what everyone else is doing. There were exceptions, of course, but that's what I found for the most part.

Maybe that's the thing that'll surprise you most in this article. But it's true.

Conclusion

Choosing just ten things for this article was, as usual, tough. I can think of lots more things to add: JWST won't replace Hubble, it succeeds it; Hubble isn't really a telescope, it's a whole observatory; it has flown the finest UV camera ever built, which was so sensitive that a massive and hot O-type star in the Andromeda Galaxy could have damaged it (and once I nearly blew it up); when there is a strong meteor shower, they point HST in the opposite direction.

There are tons of things about Hubble that I'm sure I don't know either; I worked on it for a decade, but in fact I haven't worked on it for nearly a decade since. It's a complicated and beautiful machine, and it changed the way we look at the Universe, maybe forever. It certainly changed the way scientists do astronomy... and I know that the best thing it did, the best thing it could do, was to let people see just how phenomenally gorgeous our Universe is.

And for that, I'm very grateful. And that's one thing I do know.