Amateur Astronomy


Video from Hayabusa 2’s rover currently sitting on asteroid Ryugu.


The cost of the first repair mission was costed at $2.1B at the time. I remember this vividly because the Kecks had just signed up to finance the two telescopes in Hawaii and there was big issue about continuation financing of research. Its probably because so many astrophysics PhD’s end up in high tech, not really through choice, that I’ve got to hear updates over the decades on the current state of play in field, both politics and career wise.

Basically space science is very very expensive and only marginally productive in the bigger scheme of things. But because of the contractors involved, and the Beltway lobbyists, its far easier to get billions for a space borne experiment than a few million for 90%/95% equivalent effective non space borne experiment. Which is why I suppose so many researchers just chuck it in after a while.

To me the whole James Webb Telescope fiasco is just an indication of how bad things have got. How many billions by the time its finished. And for what.

I just found the early days a lot more honest and straight forward. We are doing this cool stuff to advance the technology and as a by product, lots of stories of derring do. Nowadays I find the interminable - this mission will answer some fundamental question of science - all a bit tedious. It never does. Lots more cool photos. And huge amounts of data that will never be properly analyzed. If at all. Another dirty little secret of these mission, huge data sets rarely inspected. Eventually often lost. Its not just the original Apollo 11 video down-link data they lost.

I used George Meuller as an example because he made thing happen, both getting things built and making sure the science was done as well. Something sorely lacking in the modern NASA. I happen to know this back story of inside NASA in the early days because I used to see him lurking around the offices of a company I worked for, he was some kind of outside director, and he was such a formidable presence (even though not very tall) that I dared not talk to him. I asked others who he was and they filled me in. It seems there are lots of George Meuller stories from the early days of NASA. Most of them true. Usually involving some bureaucratic bloodbath or other. I saw a very old guy on the bus recently and he had a full set of Mercury and Apollo mission patches on his jacket, probably a mission specialist. Unfortunately I did not get a chance to talk with him. I’m sure he had interesting George Meuller stories too.

One minor correction. The max NASA budget was around 5% of the Federal Budget. Which was less than 20% of GNP at the time. So maybe just over 1% of GNP max. So on the whole a very good ROI I think. For the early days.

I am all for the advancement of astronomy. I suppose its because I’ve known so many in the business over the years that I see the space science stuff as a huge distraction and vast misallocation of resources. I’d rather they would can just one space mission and use the money to support the lifetime research careers of several hundred astronomers. That would be a great day for science. But a bad one for Lockheed Martin Space System no doubt.


Any project to which large amounts of funds are going to be committed, whether ground-based or space-based, starts with a so-called “science book”. It’s typically compiled over a period of years, and is a compendium of possible science experiments along with the technical capabilities required to execute them. Nobody spends billions of extra bucks for 5% more bang. The JWST will be able to do broadband spectroscopy in the mid-infrared. That’s not going to happen on a ground-based telescope now, or ever.

Certainly, adaptive optics and huge mirror sizes mean that ground-based scopes can challenge the resolution of space-based ones. But they can’t make the atmosphere disappear. They are limited to looking at narrow wavelength windows that are relatively transparent, like the near-infrared H-band. That’s perfect for things like stellar and galactic evolution experiments where loads of molecular species produce absorption lines in the H-band, telling us how the local universe got chemically enriched. But the H-band is only 0.2μm wide. It’s useless for continuously sliding windows of high red shift galaxies and quasars. If we want to study the early universe in detail, it can only happen in space. The schedule for the JWST observing campaign will be full of things that only it can do.


More evidence for Planet Nine? A new dwarf planet with an incredible 40,000 year orbital period could be under the influence of the elusive ninth planet’s gravity. … oort-cloud


Years ago someone once told me that we could never travel at the speed of light because if we did, we would simple disappear. Is this true or was he pulling my leg?


There’s a grain of truth to it, but it’s also a misrepresentation for which we can probably blame pop science. Objects moving relative to an observer appear foreshortened in the direction of travel. If you raced past me at extremely high speed you’d be skinnier in the direction of travel, from my point of view. I’d also appear skinnier to you. In the limit of light speed, we’d be skinnified down to nothing. However, each observer will always see themselves as perfectly normal. And none of this is why we can’t travel at the speed of light.

The real reason we can’t do it is because that’s simply the way our world works. It may look as if we should be able to accelerate indefinitely as long as we could provide the energy to do so, but that’s an illusion that only works at relatively low speeds. In the real world there is no independent space through which you can move at an arbitrary speed. There is only a mixture of space and time. The faster you move through space, the slower you move through time. The consequences can be a bit baffling because they are different for observers in different states of motion. The common mistake is to think that their must be one single “true” statement of affairs. But just as space passes by at different rates for observers at different speeds, so does time.

So let’s pretend you could travel at the speed of light. You can’t because you’re made of massive particles, but massless particles like photons can – indeed, they must. Pretend you’re a photon. Here’s the weird thing. It would look to you as if you were travelling not at light speed but at infinite speed. Remember that skinnification thing? All of space in your direction of travel would be foreshortened down to nothing. Wherever you were going you would be there, instantly. Crossing the galaxy? You’re there already, in zero time. Crossing the universe? Same thing. So that’s kind of one answer to your question – you can’t go faster than light because you’d already be going infinitely fast.

This is amazing. There are countless photons that were emitted soon after the Big Bang and have been travelling through space ever since. Back in the day when we had analogue TV tuners and CRT screens, some of the “snow” on an out-of-tune channel was caused by these cosmic background photons. If you were one of those photons, you would have set off shortly after the Big Bang and arrived on Joe Bloggs’s TV screen instantly.

So how is this possible when we know that the Big Bang was 14 billion years ago and even light (or anything travelling at light speed) has taken that long to reach us from it? It’s just one of those differences in cosmic perspective. As stationary* observers light seems to travel at the speed of light. That’s just because that’s the fastest that anything can appear to a stationary observer. But “photon-you” can travel across the universe in less than the blink of an eye. If you could look around, though, you’d want to buy shares in cosmic anti-wrinkle cream as you’d notice that the rest of us had aged by 14 billion years.

( “stationary” in the context of an expanding universe has a particular technical meaning that we needn’t worry about for now)*


Not blowing my own trumpet 'cos it’s about the love of learning, but this feels like a milestone …


Well done ps. Physics is hard.


Well Done PS


The ESA’s mission to Mercury, BepiColumbo launched three days ago. Someone asked me why it needs such a complicated trajectory to get to an inner planet:

Shouldn’t it be easy to get there, since effectively you only need to fall in toward the Sun? Yes, that’s absolutely true – the problem is the speed you will be doing when you arrive. It’s way too fast to get into orbit around Mercury. To see why, we need to consider the orbital speeds of Earth and Mercury round the Sun.

A quick Google shows them to be 30 km/sec and 47 km/sec respectively. So we’re going to need 17 km/sec of extra speed to catch up to Mercury in its orbit. We’ll get that by falling. An object in freefall trades gravitational potential energy for kinetic energy. One neat thing about gravity, like all conservative forces, is that it doesn’t matter what trajectory we take. We only need to consider our starting point and end point and the difference in potential energy between them. It doesn’t matter how we get there. That’s why Galileo was able to do all his demonstrations by rolling balls down an inclined slope instead of lobbing them out of a leaning tower.

Ok, so we just need to consider the potentials due to the mass of the Sun at the distances of Earth and Mercury, and the difference will be converted into kinetic energy, which gives:\frac{1}{2}m\Delta%20v^2%3D-\Delta%20U%3D\frac{GM_\odot%20m}{R_E}-\frac{GM_\odot%20m}{R_M}

Here we can pull a neat trick by factoring out the orbital speed given by:\sqrt{\frac{GM}{R}}

We can rearrange the first equation to get our spacecraft’s increase in velocity by falling from Earth to Mercury, and we can express it in terms of the orbital velocities of those two planets:\Delta%20v%3D\sqrt{2\left%20(%20V_E^2-V_M^2\right)}

When we do the sums we get a delta V of 51 km/sec. That’s much more than the extra 17 km/sec we’ll need to match Mercury’s solar orbital speed. So we’re going to have shed 34 km/sec of speed. And that’s no easier than gaining speed to get to the outer planets. We need gravitational slingshots to reduce our speed, hence the total of nine planetary flybys over seven years before insertion into orbit around Mercury.

This has been done before, by the Mariner 10 mission to Mercury in 1974. It was the first spacecraft to use the gravitational slingshot technique so commonly used since. The guy who calculated this interplanetary gravity assist manoeuvre was a scientist at the University of Padua whose name now adorns this spacecraft … Bepi Columbo.


The earliest star catalogue that we explicitly know about comes down to us from 2,200 years ago. But Meton of Athens had his Heliotropion observatory on this hill adjacent to the Acropolis hundreds of years earlier, and the Babylonians and others were gazing skyward long before that.

Cataloguing the position of the stars is called astrometry, and the state of the art today is being carried out by the Gaia spacecraft. At night it is sitting a million miles above your head, pointing directly away from the Sun. It rotates slowly, and captures the positions and brightnesses of the stars as they drift past. Over the course of a year it – along with the Earth – will complete an orbit around the Sun, photographing stars from all corners of the galaxy. When it is finished, it will have captured an astonishing billion of them. In the course of five years, each one will be measured many times over, running to tens of billions of observations.

Apart from its huge extent, the incredible accuracy of the Gaia catalogue will revolutionise our understanding of our galaxy’s structure. Imagine trying to measure the width of a human hair from a thousand kilometres away – that’s the level of positional accuracy of some of the Gaia measurements. The science papers about Gaia’s construction and operation are available online for anyone who wants to dig deeper. Meton of Athens may be a more serious looking dude than the cutesy Gaia who adorned the rocket fairing for the spacecraft launch (and my Windows desktop ever since), but this is very serious business.

So now, as Gaia’s data starts to flow down to us from the heavens, the painstaking analysis work has begun in earnest. And it’s already adding to our knowledge of astrophysics, as it will continue to do for many years to come. Here’s an example: anyone interested in exoplanet discovery probably knows that we have discovered some nearby “earth like” planets. But some are orbiting red dwarf or “M type” stars which produce copious X rays. So even though the planets have suitable surface temperatures for life, they may still be exposed to very harsh radiation conditions. We believe the violence of emissions from M dwarfs are due to their convective envelopes.

A pot boils over when you heat it too strongly because the steep temperature gradient causes the hot lower material to be too buoyant compared to its surroundings, so it rises. Convective energy transport is more efficient than the alternative radiative transport. The radiators in your home should be called convectors, as the vast majority of their heat output is carried by rising air, not heat radiation. It’s a similar story with stars. Temperature transitions on the way out from the stellar core lead to some outer portion of the star being convective. It’s why the surface of our Sun looks like a pot of boiling porridge.

There is a certain minimum core temperature at which nuclear fusion occurs. Less than that and you cannot form a “normal” star. Since the minimum core temperature is roughly fixed, the smaller a star is the steeper the temperature gradient from core to surface where the fusion energy can escape into transparent space. So we believe small stars are more convective. Something new happens when the convection reaches all the way down to the core. Then it can dredge up materials being formed by fusion in the core, which would normally stay locked away in a less convective star. These materials change the opacity of the envelope, that is, the rate at which heat can escape by radiative diffusion. So they can create a feedback loop which makes the star even more convective.

For normal stars, which we call main sequence stars, the hotter stars are more luminous. If we graph temperature against luminosity for a population of stars on what’s called a Hertzsprung-Russell diagram, we get a continuous line for the main sequence. But now Gaia has detected a very small gap in the main sequence, seen in the imperfection at roughly magnitude 10 on the vertical axis below.

It’s believed to be a discontinuity corresponding to the onset of full convection in M dwarf stars. We’d never know about this without the exquisite astrometry and photometry of Gaia. The paper discussing this is here, with a simpler summary here.


Visualisations based on Gaia data release 2 are beginning to be put together. Here’s some from a presentation at the AMNH in New York. A bit long winded, but gives an idea what we’re going to be able to do with the Gaia data set.


Hubble Telescope’s Broken Gyroscope Seemingly Fixed After Engineers Try Turning It Off and On Again … 1829934018



Happy Birthday, Helium! … 150 years old last Friday.

Or at least, that’s by one accounting of the rather murky history of helium’s discovery. I wrote about this briefly a few years back. But on the occasion of helium’s birthday I’ve been reading a 2009 paper from Annals of Science entitled [*The Solar Element: A Reconsideration of Helium’s Early History * ( Our nomination for helium’s birthday – October 26th 1868 – was the day when, by coincidence, two letters arrived simultaneously at the French Academy from British and French astronomers Sir Norman Lockyer and Pierre Janssen.

Janssen had been dispatched to India to observe the solar eclipse of August 1868. But a couple of weeks later he figured out a way to take a spectrum of the Sun’s chromosphere without the occurrence of an eclipse. Lockyer had figured this out a couple of years previously but didn’t have the equipment to put it into practice. He finally cracked it at the same time as Janssen, and the two went down in history together. (The trick they discovered was to use high dispersion to spread the white light of the solar continuum so thinly that it faded by comparison to the narrow emission lines of hydrogen and helium from the chromosphere, which is otherwise normally drowned out except during an eclipse).

It was for this achievement that the French government struck a medal celebrating the two men some years later. But there was no mention of helium. And nor was there any mention of helium in the two men’s letters of October 1868. Indeed there was scarcely any mention of the D3 spectral line, a yellow emission line appearing in the solar spectrum near the sodium doublet lines. In fact, even the designation D3 was due, not to Lockyer or Janssen, but to the Italian Jesuit pioneer of astronomical spectroscopy, Fr. Angelo Secchi.

It was some time later that Lockyer coined the name helium, after the Greek sun god Helios. But he himself declared that he only invented it for purposes of discussion in the laboratory. He was not at all sure that it corresponded to some unknown element, and would have been very conscious of embarrassments such as the “discovery” of the element Jargonium ( :smiley: ) in 1869. The announcement of helium to the public was due to William Thompson in 1872, who credited Lockyer.

Fortuitously, Lockyer was consulted when Ramsay isolated helium by dissolving uranium salts in acid. But that wasn’t until a quarter of a century later in 1895. Spectroscopic analysis showed the same yellow line seen in the spectrum of the Sun all those years previously. Only in 1908 was it shown that alpha particles given off in radioactive decay of uranium were indeed helium nuclei, which explained the association of those two elements.

And around the same time, helium started to be retrieved from oil wells where it was trapped along with the hydrocarbons by impermeable cap rocks. Helium turned out to be not as rare on earth as had been supposed, though once it escapes to the atmosphere it is quickly lost to space, so forms only about five parts per million of air. Finally in 1925, Cecilia Payne showed in her doctoral thesis that stars were made almost entirely of hydrogen and helium, making helium the second most abundant element in the universe.

Lockyer had died a few years earlier, with his contributions to stellar spectroscopy lauded in [his obituary in Nature * (, the prestigious science journal which he himself founded. Having started as an amateur astronomer, he became the first person to be known as a professor of astrophysics. Payne was not to be so recognised, at least not initially. Having already had to emigrate from the UK to the US for the mere chance of being allowed to study astrophysics as a woman, she was informed that the conclusions in her thesis were wrong. In fact she had correctly realised that the different spectral types of stars corresponded to stellar temperatures, and that the strength of the spectral lines corresponded to these temperatures and not the actual abundance of the elements represented.

A few years later her thesis reviewer Henry Norris Russell of H-R diagram fame realised she had been right all along. He tried to set the matter to rights, but Russell is often wrongly credited for Payne’s discovery. I suspect, though, that in future years Payne will certainly be more famous than Lockyer, who has faded to relative obscurity (although he gets a brief mention in Payne’s thesis). Lockyer may have discovered helium, but Payne elevated it to among the most ubiquitous of elements.

*** As the story goes, Payne issued an apology along with her thesis, acknowledging that her conclusions were wrong. I’m not sure if she got to amend the thesis later. Certainly, the version I have linked tells Russell that he’s flat out wrong. His preconception was that the Sun had similar constituents to the Earth and the stony meteorites, reflecting their common origin in the presolar nebula. In her masterful conclusions, Payne looks beyond superficial appearances to a universe stranger than anyone had imagined, where the primordial materials had been fashioned very differently to those of Earth’s environs. She also closes with a wry-sounding remark that astrophysics has progressed by “mutual toleration of point of view” and a “lessening of the distrust of large numbers”.**


How long does it take the earth to spin once on its axis? A day, right? Well, maybe …

What we refer to as a day is the time it takes for the Sun to return to the same position in the sky after a rotation of the Earth. To a good approximation the Earth’s rotational axis is always pointing the same direction in space, and the planet spins at a constant speed. So one rotation should bring the Sun back to its starting position … except that the Earth is also moving in its orbit around the Sun. After one day, it has moved about one degree around its orbit so that Sun is one degree behind where it would otherwise be in the sky. Imagine you’re the stick figure standing on the equator in this picture:

The Earth has to rotate an extra degree to bring the Sun back above your head. That takes about four minutes. So the solar day is four minutes longer than a day as measured by the stars. The stars are so far away that, unlike the Sun, we can treat any given star as always lying in the same direction from Earth. We call a day measured by the stars a sidereal day. That’s a true rotational day, and is about four minutes shorter than a solar day.

Actually, the solar day has a little extra complication due to the fact that the Earth’s orbit around the Sun is not circular. That makes it speed up and slow down in its orbit so that the solar day can be half a minute longer or shorter than the average at different times of year. The time we measure on our clocks is the average or mean solar day, so that we don’t have to keep adjusting the time all year.

One of my homework assignments when studying undergraduate astronomy was calculating the length of the sidereal day. It’s a neat experiment that anyone can do with a bit of patience, and needs no equipment other than an accurate clock. Our dark winter nights are a good time for it. You’ll need to pick a particular star, and watch as it returns to the same position in the sky each night. You’ll note the clock time at which this happens. It’s not trivial to make accurate measurements but if you do this over a number of nights you can smooth out the errors by plotting a graph of the changing clock time. It doesn’t even matter if your observations aren’t on consecutive days due to weather or lack of dedication. The best-fit line graph will still work.

There are two important aspects to this experiment. First, you have to pick an appropriate star. Not any old star will do. It needs to be bright enough that you can reliably pick it up night after night, especially if you are viewing in marginal or light-polluted conditions. And of course, make sure you pick a star and not a planet. You also have to pick a star that moves! If you look at the North Star, it stays in the same spot hour after hour, night after night. All the stars rotate about it. The stars closest to it move in little circles. Stars further away move in bigger circles. Bigger circles translate into more movement on the sky, which makes your measurement task easier. You want a star as close as possible to the celestial equator for maximum movement. The celestial equator is an imaginary line in the sky that touches the horizon to your east and west and rises in an arc across your southern sky, reaching its highest point due south. From Ireland that highest point is 38 degrees above the horizon. But anywhere between a third and half way up the sky when facing south will do.

Now, having chosen your star, the second trick is to make stable measurements. There is no way you can make accurate measurements of star positions by eye alone. You need a marker, which might be the edge of a building, a chimney pot, a telephone pole etc. And you need to stand in precisely the same position each time you make an observation. The further away your marker, the more accurate your measurement will be, as the line from you to the marker will sweep through a smaller angle for a given amount of head movement. You need to practice this. When your star lines up with your marker, note the time. But also note how much the star moves when you move your head slightly. If it moves too much your marker may be too close to you for accurate measurements. As a rule of thumb, twenty metres should be the minimum. At this distance, one degree corresponds to about 30 cm of head movement. Also practice planting your feet a decent distance apart with equal pressure on each foot, so that your head is centred on the midpoint. Mark the position for future observations.

Here’s how my experiment went. The gable end of the neighbours’ house was conveniently arrow shaped, and at 25 metres distance was far enough away for accuracy. If you’re in a very dark location as I was, you can run into the problem that your marker is invisible in the dark. Mine was just close enough to illuminate with a torch so that it could be seen very dimly against a pitch black sky. Below is a twilight photo and a torch lit nighttime schematic.

My chosen star was Spica (α Virginis) the brightest star in Virgo. It is 11 degrees below the celestial equator which brings it conveniently close to terrestrial markers. By flexing my knees I was able to drop Spica right onto the point of my gable arrow and get a reasonably accurate time measurement. I observed on five different nights over two weeks between late April and early May. As expected, my star arrived back at the same point at an earlier clock time each night, as the sidereal day is shorter. Trying to keep your head steady while glancing back from your star to your clock is a bummer, so I noted the time to the nearest fifteen seconds – sufficient measurements should even out this source of error:

(Rookie mistake: there’s no such thing as a 4th quarter moon; it was ‘waxing gibbous’ :blush: )

Plotting the offset in seconds against the days of observation gave a straight line graph like this:

The error bars are based on wiggling my head a bit to see how much it threw the time off – not very scientific! But I managed to draw a best-fit line, and calculate a result with a margin of error of about 2.5%. I found that the the sidereal day was 242 ±6 seconds shorter than the solar day. The correct value is 236 seconds, or 3 minutes 56 seconds.

There’s a great satisfaction in having done some actual experimental astronomy. If you get this right you’ll have graduated to the level of your pre-scientific ancestors of many thousands of years ago! And while that should be a source of humility, it’s no mean feat as your ancestors knew the stars better than 99.9 % of people today. :smiley:


Official thread Anthem - Go!

Official soundtrack for this thread. :wink:


How much of the sky can you see? The short answer is half of it (assuming a clear view to the horizon all round). The other half is below your horizon. It’s not the whole answer though. The sky rotates in the course of a sidereal day which may bring more of it into view depending on where you are.

If you stand at the north pole you are out of luck. You can see all of the northern celestial hemisphere and none of the southern one, and that doesn’t change as the sky rotates. The celestial equator is always on your horizon. But if you stand on the terrestrial equator, the celestial poles are on your northern and southern horizon and the entire celestial sphere will rotate over your head in the course of one day. Let’s be parochial and ask how much of the sky can we see from Ireland, more specifically at 53°N latitude. Here’s our point of view at any given instant in time:

The green plane is bounded by our horizon which divides the celestial sphere in half. The blue spike in the sky is the north celestial pole which is 53° (equal to our latitude) above our northern horizon. Correspondingly, the celestial equator (yellow line) tilts 90°-53°=37° above our southern horizon, so at the horizon we can see 37° below the equator as long as we look directly south. But 12 hours later after half a sky rotation, the part that was below our northern horizon is now above our southern horizon. That means eventually we will see the whole sky down to 37° south of the celestial equator. That’s handy if you know the declination (“celestial latitude”) of a star. Sirius is 16° below the celestial equator, so it should get to a maximum of 21° above the southern horizon in Ireland. Go outside at 3.45am this morning and that’s exactly where you’ll see it.

But back to our question – how much of the sky can we see? Well, the long answer as we’ve stated is that we can see a truncated sphere lopped off below 37° south latitude. To measure that as a proportion of the whole celestial sphere, we need the concept of solid angle. The whole sky (or indeed any sphere) has a total solid angle of 4π steradians. Let’s calculate the area of the sky we can’t see. It will be the bit surrounding the south celestial pole, the blue area in this picture with θ=53° (equal to our northern latitude):

The solid angle in steradians is given by\pi(1-\cos\theta). So then the fraction of the sky we can see is:\frac{4\pi-2\pi(1-\cos\theta)}{4\pi}%3D\frac{1+\cos\theta}{2}

For Ireland that works out to almost exactly 80% of the total celestial sphere that we can see in the course of a day. Of course, if we want to look at the stars, the Sun gets in the way during daylight hours but as it moves around the sky in the course of a year we’ll get to see 80% of the whole sky in darkness over that period.


The Kepler spacecraft was finally turned off last week after more than nine years of operation. It was originally planned to last three, but Herculean efforts to keep it going after a reaction wheel broke down gave it a whole new lease of life. Kepler discovered more than 2,700 exoplanets and thousands more exoplanet candidates still to be confirmed. NASA posted this Kepler commemoration video:

Meanwhile, the recently launched TESS mission is already looking for exoplanets among our nearest neighbouring stars. TESS is surveying the southern and northern hemispheres in 13 strips each, at a rate of one strip per sidereal month of its two year mission. It is currently on strip 5 of the southern hemisphere.


Just came across a great physics/astrophysics podcast with nearly a hundred half-hour episodes so far. “Ask a Spaceman” is by Paul Sutter, a bona fide research astrophysicist. The tone is jovial and entertaining, but he gets into some quite deep stuff. Way better than anything you’ll come across on dumbed-down TV nowadays, in spite of being a one man show funded by donations. Recommend.


Everything you can see happened in the past. You can never see what is happening “now”. If you are talking to someone a metre from you, their moving lips appear as they were three nanoseconds ago. That’s how long it took light to travel from them to you. That’s way shorter than the smallest time interval your eyes or brain can resolve, so makes not the slightest difference to you. But the principle still stands, and becomes very relevant when we look out to cosmological distances.

We see the nearest stars as they were several years ago, the farthest stars in our galaxy as they were tens of thousands of years ago, the nearest galaxies as they were millions of years ago. Whatever we see is already old news by the time it reaches us. The faintest galaxies are many billions of years in our past. Bearing in mind that the Big Bang happened less than 14 billion years ago, what would happen if we could look back that far? What direction should we look, and what would it look like?

Here’s where we run into a conceptual difficulty, especially if we’ve watched pop science programmes that show us the Big Bang as a glorified firework explosion. It implies that we should, in principle at least, see that first spark of light somewhere in our sky as we gaze back into the past. We might wonder why some particular direction would be favoured over another. And we’d be right to wonder because the Big Bang isn’t like that at all. It is not an explosion into pre-existing space, but the origin of all matter, energy and space itself. That space has been expanding ever since, carrying the energy and matter with it. The Big Bang took place everywhere at the same time, remembering that “everywhere” used to be a very, very small place. So when we look for the Big Bang we need to look everywhere … literally all around us.

As it happens, we can’t see right back to the Big Bang. There is a fog in the way. When you heat matter hot enough, its atoms will eventually dissociate into a mixture of atomic nuclei and free electrons. Light cannot travel freely through this mixture because it interacts with all those electrically charged particles and scatters. The early universe was hot, but cooled as it expanded. When the temperature fell below 10,000 degrees the fog cleared and the universe became transparent to light for the first time. Cosmologists are quite specific about when this happened – 377,000 years after the Big Bang.

You’ve probably stood in the middle of a fog bank before. You can see sunlight, but it’s diffuse and coming from all around you. There is some limited distance over which you can see things, but they get fuzzier with distance, and then at some poorly defined edge it all fades away into a diffuse white glow. The whole universe is like that. We refer to the edge as the surface of last scattering. The light from the Big Bang diffused through the hot plasma fog until it suddenly broke free and started travelling through the universe relatively unimpeded. That light is all around us today, every direction we look. It is very faint and it has been stretched out to long (microwave) wavelengths with the expansion of the universe. We call it the Cosmic Microwave Background radiation, or CMB.

Now, while the CMB is perhaps the most important tool available to Big Bang cosmologists, that’s not what I intended to write about here. I just wanted to convey that events that occurred when the universe was a lot smaller can be seen all around us today, as photons reach our eyes from them. When the fog cleared after the Big Bang, space and matter rapidly became very dark and very cold. Hundreds of millions of years passed, but gradually little overdense knots of gas formed under the influence of gravity. These condensed to form the first stars, and the universe started to light up again for the first time since the Big Bang. The light from those first stars and from subsequent generations of them is sloshing around along with the CMB. We call it the Extragalactic Background Light, or EBL.

We can distinguish between the EBL and CMB quite easily. The CMB comes from a quite specific epoch and had a thermal spectrum characteristic of the 10,000 degrees at which primordial protons combined with electrons to form neutral hydrogen atoms. The EBL comes from stars at a wide range of temperatures, over a much longer period of time. So there was a wide wavelength range to begin with, and it has been stretched out by cosmic expansion for varying lengths of time, though always less time than the CMB. Thus we can see some of the EBL at optical wavelengths, and though it is very faint we can measure it. When I say “very faint”, I mean perhaps a thousandth of a trillionth of a watt falling on each square metre of a telescope mirror. But at least we can use large ground-based telescopes, whereas most of the microwaves of the CMB don’t make it through our atmosphere and require space-based instruments.

Another way we can measure the EBL is by looking at its effect on very intense sources of light. Photons don’t normally interact with each other, but high energy gamma rays from, say, an active galactic nucleus occasionally do so. A gamma ray can collide with an EBL photon and give rise to a pair of particles. (It’s called pair production and is the opposite process to when particles and antiparticles collide to give gamma rays). The result is that some of the energy of these bright sources is filtered out on its way to us across space and time. By modelling how bright they ought to be, we can calculate the intensity of the EBL (see here and here for examples).

So what can the EBL tell us? In effect it can tell us how many stars and galaxies have formed since the Big Bang. But because of its unique evolution with time, we can also infer the star formation rate and how it has changed in the cosmological past. Recently we’ve been refining our estimates of these numbers (see here and here for recent papers).

Then, this week we’ve had a paper published which claims to reconstruct the whole history of star formation rates. The second last sentence of the abstract reads “Our star-formation history is consistent with independent measurements from galaxy surveys, peaking at redshift z∼2”. Redshift is a measure of cosmic expansion and therefore z=2 corresponds to a particular age of the universe. So that single innocuous-sounding statement tells us our universe is dying! The star formation rate peaked around nine billion years ago and has been declining ever since. The myriad of galaxies are like glowing embers from a fire, gradually fading toward eternal blackness.

But don’t let it get you down. There’s plenty of life in the old dog yet! 8DD