Why build the Hubble Space Telescope?

After I had started my post about the planning phase of the HST up to the start, I have noticed that the arguments (which I thought to be a small section in there) for building a Space Telescope in the first place took up quite some space, so I have decided to post this for now and delay the rest of the post into a different one (trying to keep the posts from getting too extensive as I promised last time), possibly the next.

At first sight it doesn’t sound like a terribly good idea to put an expensive telescope on an expensive rocket (which doesn’t mean it’s a successful start at all, usually expensive equipment and large quantities of explosives are kept well apart for a reason) and shoot it into orbit where again many things can (and did) go wrong and telescope maintenance is either impossible or at least very expensive. Why not simply build a ground-based telescope which is a lot cheaper but would still have a much bigger mirror (so can collect more light and potentially create sharper images, see below) and would still be much cheaper and easier to maintain? Where a new camera simply needs to be screwed on (a simplification about which many telescope engineers could rightfully complain), rather than having to employ the (again expensive) space shuttle to even get there in the first place and then having to work in unhandy space suits to get the new equipment in (with the risk to notice that a bolt might be too long and the camera doesn’t even fit in); and if it doesn’t work afterwards, you’re screwed and you cannot simply take it down again and fix it? Not even talking about potentially simple problems like cooling the camera chips for which you need liquid nitrogen or helium, which, on earth, you can simply refill with a tank, but which in space becomes a rather more complicated task altogether.

So at first sight, it looks like people need a very good reason to even think about Space Telescopes, an then start developing, maintaining and upgrading one with new cameras. So what are these advantages that made scientists built the HST (and many other Space Telescopes)?

Well, there are 2 main reasons, one of which solves a problem that was at least impossible to overcome back then and one that solves a fundamental problem alltogether:

Ausschnitt_COMBO2Ausschnitt_UDF2The first problem: In principle, the angular resolution of a telescope should be given by the wavelength of the observed light and the diameter of the mirror (or lens). This is simple optics that every physics student learns and it says that the bigger a mirror, the better the resolution. But there is one very large problem: This is pure theory for an ideal telescope in a vacuum. As in the old physicist joke: ‘Yes, I’ve got a solution, but it only works for spherical elephants in a vacuum’. But here, talking about telescopes instead of elephants, the problem really IS the atmosphere which changes the situation completely.

The atmosphere has its upsides for breathing, weather (sometimes we could do with a bit less atmosphere here in England 😉 ), airplanes and stuff, but for astronomers, it’s really just ‘in the way’. The air, or better to say the movement of the air, the turbulences of hot and cold air bubbles, bends the starlight on its way into the telescopes (very much like when you’re looking over the tarmac on a very hot summers day everything is flickering). This basically makes stars jump around very fast (Actually, you can see this effect by looking at the stars. Stars ‘flicker’ at night. Objects larger than this effect, e.g. planets, don’t flicker; That’s how you can tell them apart easily). As the apparent position of the stars jump around faster than the eye or a normal camera can detect, it doesn’t really make stars move in long-time exposures, but it leads to ‘blurring’ of the image (If you want to know what a stars picture actually looks like in extremely short exposures, read this article).

This effect is so big that it easily overpowers the resolution improvement due to a bigger mirror size. In fact, to get the maximum resolution of a telescope at sealevel, a telescope with something between 10 and 20 cm diameter is big enough. Any bigger than that does not increase the image resolution due to seeing. Groundbased telescopes can therefore not see details smaller than 0.5-1.0 arcsec in size, even at the best telescope sites (which are usually in very dry places very high up. Dry weather is usually less turbulent and building telescopes at high altitude avoids having to look through large parts of the atmosphere in the first place, so ‘seeing conditions’ are usually better). For comparison, the Hubble Space Telescope has a resolution of only 0.05 arcsec due to it’s position outside the atmosphere and its size of around 2.5 meter.

I show the effect of seeing in the images above. The top one is a picture taken by a 2.2 meter (so similar to HST) telescope in Chile from the Combo-17 survey. The bottom one is the same area of the sky taken by HST (to be precise, it’s a small bit of the Hubble Ultra Deep Field H-UDF, stay tuned for one of my later posts about this survey). You can clearly see that all the objects in the groundbased survey are blurred and some of the faint objects are actually blurred so much, that they cannot even be seen at all above the sky background. In the space based image, the details and features of the galaxies are visible much more clearly. This is the reason why these images were now chosen to be classified in Galaxy Zoo: Hubble. From groundbased telescopes a classification of galaxies at similar distances is simply not possible.

Of course, a big telescope has another obvious advantage: It collects a lot of light, so it enables us to see fainter objects. This is the main reason why telescopes in the past were built bigger and bigger and this trend continues even today.

Also, there are a few things that can be done to improve the image quality, but most of them go beyond the scope of this blog, but if you want to know about it, read e.g. this article and links therein. Basically, either interferometry (where several telescopes far apart are used, unfortunately, this does not produce a full ‘image’ of the object) or movable and  distortable mirrors can be used. The latter is called adaptive optics, a very complicated and expensive technique, in which the wavefront of the starlight that has been distorted by the atmosphere can be corrected by a mirror, which is generally speaking distorted exactly in the opposite way to make the wavefront flat again (The mirrors shape has to be adjusted around 200 times per second). A flat wavefront will create a sharp – ‘diffraction limited’ – image, using the complete power of the big mirror used. Adaptive optics was only developed in recent years on several telescopes. The Hooker telescope mentioned in my last blog runs one, for example, and most big telescopes like Gemini, Keck or the VLT run these facilities, too. Although I did call this technique ‘expensive’, it is, of course, a lot cheaper than building a Space Telescope.


In fact this is one of the reasons why the HSTs ‘successor’, the JWST telescope (also one of my later posts) will not be observing at the same optical wavelengths as HST but will rather concentrate on IR  wavelengths. Running adaptive optics facilities, groundbased telescopes now produce images of the same quality as the HST, although on a smaller field of view and only close to stars (they need bright-ish stars to compensate the effect of the atmosphere, although laser guide stars will avoid at least the second problem in the future). The 30-40 meter telescopes that are currently planned around the world will show much better resolution again and using very sophisticated adaptive optics might have a similar or even bigger field of view as HST currently has. In the image below (please click for full resolution), you can see an estimated example for a telecope called OWL (Overwhelmingly Large telescope, not kidding, the image above shows a simulation. The speck in the foreground is a car for size comparison). This was a planned telescope which has now been cut down to 42 meters, we now call it the E-ELT (the Extremely Large Telescope, yes, astronomers are not very creative when it comes to telescope names). It’s design is different (but OWL of course is more impressive and their website provided the images I was looking for), but very recently, its cite has been selected to be on a neighbouring mountain to Cerro Paranal, the cite of the VLT.


Resolution comparison for OWL (Click for better resoultion)

As you can see on the left, this telescope would, with perfect adaptive optics (diffraction limited) indeed produce much sharper images with much higher resolution then even HST can provide today. So in principle, the problem of the atmosphere can be overcome using clever techniques. The field of view on which this works today is still a lot smaller than what is covered by HST survey cameras, but this is a matter of technique and might be overcome in the future (if interested, google ‘multiconjugate adaptive optics’). In the past, when HST was planned, non of this existed, so a Space Telescope indeed sounded like a good idea.

g17b_atmosabsorbBut groundbased telescopes have an even more fundamental problem which is simply put impossible to overcome. At ground level, only certain wavelengths in the electromagnetic spektrum can be observed. (Far) Infrared, ultraviolet and gamma light cannot be observed at all from groundbased telescopes as the light is strongly absorbed by the atmosphere. The image on the right shows the height above ground at which light is basically absorbed by the atmosphere. As you can see, only optical and radio (and a bit near infrared) observatories make sense on earth, even on the highest mountains. For any other wavelength, you need a space telescope to be able to observe galaxies at all.

screengrab from Wikipedia

NGC1512 screengrab from Wikipedia

Different wavelengths can be very interesting, a galaxy looks completely different in optical than in X-ray or Radio and all these wavelengths show different physical parameters, e.g. star formation rates (X-ray) or the amount of ‘dust’ in the galaxy (in IR). A good and famous example for this is NGC 1512, a barred spiral galaxy whose center has been observed at different wavelengths by the same telescope. You can see the results on the right. Generally speaking, very blue light (shown in purple) shows very young stars, redder light (up to orange) older stars, red shows dust. Remember, these are all still more or less optical wavelengths, in X-ray or far infrared, this galaxy would look even more different.

The HST works at optical wavelengths, so this was not really a reason to build it, most things (although HST does have some IR filters) could indeed be observed by groundbased telescopes. But as I mentioned above, the HSTs successor, the JWST will exclusively be working in IR for exactly this reason. Optical cameras are not really needed in space anymore (although it’s of course a shame that we won’t have an optical space telescope in the future), but for IR observations it’s vital to be in space. Other famous Space Telescopes in other wavelengths include Spitzer (IR), Chandra (X-ray), GALEX (UV) and XMM-Newton (Xray), all of which are impossible to be replaced by earth-bound instruments.

There are also quite a few focused satellites for certain experiments in space, e.g. WMAP (to observe the afterglow of the Big Bang), Keppler (to detect exo-planets, planets around other stars than the sun) and others, but these are focused projects and not open telescopes everybody can apply to, which of course everyone can at the HST. I will talk about this and some big projects that successfully got their time on the HST in my future posts.

All in all, you can see, there are several good reasons to build a Space Telescope, scientists don’t do this because it’s ‘fun’ or ‘cool’. For observations in certain wavelengths it is still important today, for others it has at least been important in the past. Which brings us back on track: At some point, the decision was made to build the HST (although it wasn’t named Hubble then) and the planning began. But this will be my next post, so stay tuned. I am travelling a lot in the next month, so I might not be able to hold the 2-week schedule, but I’ll do my best.



Previous history of this series:

Tags: ,

About The Zooniverse

Online citizen science projects. The Zooniverse is doing real science online,.

6 responses to “Why build the Hubble Space Telescope?”

  1. Jean Tate says :

    The physicists I know refer to spherical cows, not elephants! 😉

    Great blog Boris, very informative and easy to read. One question though: doesn’t (or, perhaps, didn’t) the Hubble have a capability in the UV, at wavelengths shorter than those ground-based telescopes can observe at? Also, don’t space-based telescopes have the advantage of a darker sky, and no telluric lines?

  2. NGC3314 says :

    Though Boris hasn’t chimed in yet – indeed, HST has UV capability as short as 115 nm, compared to an atmospheric limit ~320 (the limit is essentially set by what coatings can make a mirror decently reflective all the way from UV to near-IR). The darker sky isn’t a huge issue in the visible (half as bright as the darkest sites on Earth in the V band, for example), but is a huge gain (thousands of times) as one goes into even the near-IR.

  3. Boris says :

    Indeed, Hubble does have (near-) UV filters, I have ignored them above. Sorry for that. I have just come across this website which shows this for NGC 1512, which I have mentioned above.


    UV really is the wavelength that shows young stars and star formation best, so what I said above about blue light, really means UV.

    The darker sky, as NGC3314 pointed out already, is not that much of a problem in optical wavelengths. It’s a big problem in IR, yes.

    I have ignored the telluric lines entirely, as I have ignored Hubbles spectroscopic ability entirely so far. Yes, Hubble has some instruments to take spectra on board, but as GalaxyZoo works with galaxy images and the surveys I will be talking about (which provide the data for GalaxyZoo: Hubble) are mainly imaging surveys as well, I didn’t want to dive into that too much. It’s also not where my expertise lies, I’m working on imaging data as well

    Spherical cows? How unrealistic 😀
    In Germany, it was always elephants

  4. Nathaniel Freedman says :

    The right-hand side of the diagram of atmospheric penetration of EM radiation shows that very long wavelength radio waves (>11 m) do not penetrate the atmosphere. This is the last completely unexplored area of astronomy. The few satellites that have looked at those wavelengths did not look at many frequencies, did not look at wavelengths very much longer than 11 m, and had poor resolution.

    Eventually an array of radio telescope satellites will need to be put into space to map and determine the spectra of very long wavelength radio sources. That will be very expensive, but it may give fascinating discoveries about the universe.

  5. Gail says :

    I just found this site that has lots of Hubble photos! It is great! http://hubblecollection.com/

Trackbacks / Pingbacks

  1. Spring comes to the planet Uranus - October 7, 2012

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: